Return to search results
Hybrid machine learning model to predict 3D in-situ permeability evolution
Enhanced geothermal systems (EGS) can provide a sustainable and renewable solution to the new energy transition. Its potential relies on the ability to create a reservoir and to accurately evaluate its evolving hydraulic properties to predict fluid flow and estimate ultimate thermal recovery. Here we develop a hybrid machine learning (ML) model to predict permeability evolution of intermediate-scale (~10 m) hydraulic stimulation experiments at the Sanford Underground Research Facility (the EGS Collab project). We present a 3D map in situ permeability evolution for two of stimulation episodes in this project using microearthquakes (MEQs) data and injection histories of wellhead pressure and flow rate. This map includes both average reservoir permeability evolution over time and local fracture permeability distribution within the evolved reservoir. Compared with the ground truth of average permeability calculated from the well data, our predicted average permeability for these two episodes has a MSE value less than 2.9E-4 and R2 higher than 0.93, indicating that average permeability predicted by machine learning is consistent with good agreement with field observation. Additionally, distributed fracture permeability calculated by empirical equation over time shows the process of fracture propagation and identify the potential fluid path for geothermal reservoir.
Complete Metadata
| @type | dcat:Dataset |
|---|---|
| accessLevel | public |
| bureauCode |
[
"019:20"
]
|
| contactPoint |
{
"fn": "ziyan Li",
"@type": "vcard:Contact",
"hasEmail": "mailto:liziyan1992@gmail.com"
}
|
| dataQuality |
true
|
| description | Enhanced geothermal systems (EGS) can provide a sustainable and renewable solution to the new energy transition. Its potential relies on the ability to create a reservoir and to accurately evaluate its evolving hydraulic properties to predict fluid flow and estimate ultimate thermal recovery. Here we develop a hybrid machine learning (ML) model to predict permeability evolution of intermediate-scale (~10 m) hydraulic stimulation experiments at the Sanford Underground Research Facility (the EGS Collab project). We present a 3D map in situ permeability evolution for two of stimulation episodes in this project using microearthquakes (MEQs) data and injection histories of wellhead pressure and flow rate. This map includes both average reservoir permeability evolution over time and local fracture permeability distribution within the evolved reservoir. Compared with the ground truth of average permeability calculated from the well data, our predicted average permeability for these two episodes has a MSE value less than 2.9E-4 and R2 higher than 0.93, indicating that average permeability predicted by machine learning is consistent with good agreement with field observation. Additionally, distributed fracture permeability calculated by empirical equation over time shows the process of fracture propagation and identify the potential fluid path for geothermal reservoir. |
| distribution |
[
{
"@type": "dcat:Distribution",
"title": "Induced microearthquakes predict permeability creation in the brittle crust.pdf",
"format": "pdf",
"accessURL": "https://gdr.openei.org/files/1311/Induced%20microearthquakes%20predict%20permeability%20creation%20in%20the%20brittle%20crust.pdf",
"mediaType": "application/pdf",
"description": "In this paper, we develop a hybrid machine learning (ML) model to visualize in situ permeability evolution for an intermediate-scale (~10 m) hydraulic stimulation experiment. This model includes an ML model that was trained using the well history of flow rate and wellhead pressure and MEQ (microearthquake) data from the first three stimulation episodes to predict average permeability from the statistical features of the MEQs alone for later episodes. Moreover, a physics-inspired model is integrated to estimate in situ fracture permeability spatially. This method relates fracture permeability to fracture dilation and scales dilation to the equivalent MEQ magnitude, according to laboratory observations. The seismic data are then applied to define incremental changes in permeability in both space and time. Our results confirm the excellent agreement between the ground truth and model- predicted permeability evolution. The resulting permeability map defines and quantifies flow paths in the reservoir with the averaged permeability comparing favorably with the ground truth of permeability."
},
{
"@type": "dcat:Distribution",
"title": "EGS Collab Experiment 1 Stimulation Data",
"format": "HTML",
"accessURL": "https://gdr.openei.org/submissions/1229",
"mediaType": "text/html",
"description": "Stimulation data from Experiment 1 of EGS Collab, which occurred on the 4850 ft level of the Sanford Underground Research Facility (SURF)."
},
{
"@type": "dcat:Distribution",
"title": "Newberry Volcano EGS Demonstration Well 55-29 Stimulation Data",
"format": "HTML",
"accessURL": "https://gdr.openei.org/submissions/271",
"mediaType": "text/html",
"description": "The Newberry Volcano EGS Demonstration in central Oregon, a 5 year project started in 2010, tests recent technological advances designed to reduce the cost of power generated by EGS in a hot, dry well (NWG 55-29) drilled in 2008. This submission includes all of the files and reports associated with the geophysical exploration, stimulation, and monitoring included in the scope of the project."
},
{
"@type": "dcat:Distribution",
"title": "Newberry Well 55-29 Stimulation Data 2014",
"format": "HTML",
"accessURL": "https://gdr.openei.org/submissions/508",
"mediaType": "text/html",
"description": "The Newberry Volcano EGS Demonstration in central Oregon, a 5 year project started in 2010, tests recent technological advances designed to reduce the cost of power generated by EGS in a hot, dry well (NWG 55-29) drilled in 2008. This submission includes all of the files and reports associated with the stimulation, pressure testing, and monitoring included in the scope of the project."
}
]
|
| identifier | https://data.openei.org/submissions/7429 |
| issued | 2022-11-22T07:00:00Z |
| keyword |
[
"EGS",
"EGS collab",
"Newberry",
"energy",
"enhanced geothermal systems",
"flow rate",
"fracture permeability",
"geothermal",
"hydraulic",
"hydraulic fracturing",
"induced seismicity",
"machine learning",
"microearthquake",
"permeability evolution",
"processed data",
"seismic data analysis",
"stimulation",
"wellhead pressure"
]
|
| landingPage | https://gdr.openei.org/submissions/1311 |
| license | https://creativecommons.org/licenses/by/4.0/ |
| modified | 2023-10-04T19:45:01Z |
| programCode |
[
"019:006"
]
|
| projectLead | Mike Weathers |
| projectNumber | EE0008763 |
| projectTitle | Machine Learning Approaches to Predicting Induced Seismicity and Imaging Geothermal Reservoir Properties |
| publisher |
{
"name": "Pennsylvania State University",
"@type": "org:Organization"
}
|
| spatial |
"{"type":"Polygon","coordinates":[[[-121.4333333,43.51666667],[-121.0333333,43.51666667],[-121.0333333,43.91666667],[-121.4333333,43.91666667],[-121.4333333,43.51666667]]]}"
|
| title | Hybrid machine learning model to predict 3D in-situ permeability evolution |
