Mapping Geothermal Potential in the Western United States

1
Mapping Geothermal Potential in the Western
United States

• Colin F. Williams, Jacob DeAngelo
• USGS, Menlo Park, CA

• http//energy.usgs.gov/other/geothermal/

2
Outline

• Introduction
• Weights of Evidence Approach
• Evidence Layers
• Modeling Results
• Other Components of Undiscovered Resource
  Estimates
• Summary

3
Relation to the National Resource Assessment
Identified Systems - Mean 9,057
MWe Undiscovered Resources Mean 30,033
MWe Enhanced Geothermal Systems - Mean
517,800 MWe
4
Geothermal Assessment Concepts

• Resource Geothermal energy that is technically
  recoverable and can be added to Reserves at some
  future time - Identified and Undiscovered
• Undiscovered Resource Components
• Identified but Mischaracterized
• Identified Resource Growth System size larger
  than estimate
• True Undiscovered

5
Outline

• Introduction
• Weights of Evidence Approach
• Evidence Layers
• Modeling Results
• Other Components of Undiscovered Resource
  Estimates
• Summary

6
Weights of Evidence
The Weights-of-Evidence approach employs
probability to determine the degree of
correlation among spatial databases. This measure
of correlation can be used to produce maps of
favorability for the occurrence of features of
interest. The technique has been used to study
the spatial relationship of deposits to evidence
layers through the use of Geographic Information
Systems (GIS). Prior probability Derived from
knowledge that is present before a particular
observation is made. Posterior probability
Derived from knowledge developed once the outcome
of the observation is taken into
account. Positive and negative weights are
assigned which represent the degree of
correlation between evidence layers and
occurences.
7
www.unr.edu/geothermal
8
Outline

• Introduction
• Weights of Evidence Approach
• Evidence Layers
• Modeling Results
• Other Components of Undiscovered Resource
  Estimates
• Summary

9
Idealized Geothermal Power Plant
Need Heat (Temperature) and Fluid (Permeability)
10
Evidence Layers Magmatism and Faulting
Geothermal systems correlate with 10km distance
of magmatic activity, felsic weights slightly
greater than mafic. Systems also correlate within
4km distance of Quaternary faults.
11
Evidence Layers Heat Flow
All geothermal systems show strong positive
correlation with heat flow.
12
Evidence Layers State of Stress
Extensional stress correlates with amagmatic
geothermal systems. Note dominance of extensional
and strike-slip stress in regions of known
geothermal potential.
13
Evidence Layers EQ Density for Mgt3.0 since 1965
Correlation with seismicity for all types of
geothermal systems. Warmer colors represent
greater number of earthquakes per km2 since 1965.
14
Evidence Layers Summed Moment Release
Correlation with seismicity for all types of
geothermal systems. Warmer colors represent
greater moment release since 1965.
15
Outline

• Introduction
• Weights of Evidence Approach
• Evidence Layers
• Modeling Results
• Other Components of Undiscovered Resource
  Estimates
• Summary

16
Weights of Evidence Model Results
Distribution of WofE results for Model 1, which
correlates faults, magmatic activity, seismicity
and heat flow with the presence of all geothermal
systems.
17
Weights of Evidence Model Results
Distribution of WofE results for Model 7, which
correlates faults, magmatic activity, seismicity
and heat flow with the presence of magmatic
geothermal systems.
18
Weights of Evidence Model Results
Distribution of WofE results for Model 13, which
correlates faults, stress, seismicity and heat
flow with the presence of amagmatic geothermal
systems.
19
Logistic Regression Model Results
Logistic regression is not affected by
conditional dependence and shows less extreme
variation in posterior probability across the
study area.
20
Undiscovered Resources Effect of Closed Public
Lands
21
Outline

• Introduction
• Weights of Evidence Approach
• Evidence Layers
• Modeling Results
• Other Components of Undiscovered Resource
  Estimates
• Spatial Distribution by WofE and Logistic
  Regression
• Size Distribution from Identified Systems
• Magnitude of Resource from Evaluation of Geologic
  Constraints and Historical Exploration
  Effectiveness
• Summary

22
Power Distribution of Identified Systems
Identified Systems Potential Mean 9057
MWe F95 3675 MWe F5 16457 MWe
80 of power potential from the 20 of systems
with TR gt150oC
2 systems
23
Summary

• Weights-of-evidence and logistic regression
  models used to produced maps of geothermal
  favorability.
• Analyses show that presence of geothermal
  systems correlated with heat flow, magmatic
  activity, faulting, seismicity, and tectonic
  stress.
• The results highlight regions with significant
  geothermal potential but few identified
  geothermal systems.
• The results also allow for comparison of
  predictions with observations in regions of
  varying degrees of geothermal exploration.
• When combined with digital land status maps, the
  model results can also be used to help estimate
  the spatial distribution of undiscovered
  resources.