Introduction to Passive Seismic Surveys and Seismic Imaging in Geothermal Reservoirs

1
Introduction to Passive Seismic Surveys and
Seismic Imaging in Geothermal Reservoirs

• Geophysical Techniques in Geothermal Exploration
• Lawrence Hutchings
• Lawrence Berkeley National Laboratory
• September 28, 2007

2

• Uses of microearthquake data
• Seismicty location of earthquakes
• High resolution locations can identify fractures
• Tomography from wave propagation can image
  reservoir for seismic velocity and attenuation,
  geologic structure, and lithology?
• Interpretation of tomography results (rock
  physics) can identify altered or permeable zones,
  phase states of fluids, crack density, and
  saturation
• Improve resolution for drilling targets
• Improve success estimates for drilling targets
  Monte Carlo Coupled inversion
• Monitor existing wells and production

3
Roadmap

• locations of microearthquakes
• instrumentation and data processing
• high resolution locations for mapping
• rock physics interpretations
• tomographic inversion
• shear-wave splitting
• Joint inversion techniques

4
Tectonics of Salton Sea Region
5
Geophysical Setting Salton Sea
6
Reservoir Recent Seismicity
Depth - km
7
Roadmap

• locations of microearthquakes
• instrumentation and data processing
• high resolution locations for mapping
• rock physics interpretations
• tomographic inversion
• shear-wave splitting
• Joint inversion techniques

8
Instrumentation

• Three-component microearthquake recorders
  (high-frequency)?
• Records continuously at 500 samples per second,
  more
• GPS location and timing information

9
Automated processing to identify events
10
Two events from the same small area
11
Roadmap

• locations of microearthquakes
• instrumentation and data processing
• high resolution locations for mapping
• rock physics interpretations
• tomographic inversion
• shear-wave splitting
• Joint inversion techniques

12
Regionally Recorded Microearthquakes
13
Microearthquake Locations
CalEnergy Data
14
Hypoinverse locations
15
HypoDD locations
16
Roadmap

• locations of microearthquakes
• instrumentation and data processing
• high resolution locations for mapping
• rock physics interpretations
• tomographic inversion
• shear-wave splitting
• Joint inversion techniques

17

• Rock Physics Observations
• Variations in lithology observed in elastic
  constants
• Increase of velocity and decrease in attenuation
  with depth due to closing of small cracks due to
  pressure
• Decrease in velocity and increase in attenuation
  due to fracturing
• Decrease in velocity due to chemical alteration
• Extreme temperature gradient works to decrease
  velocity with depth
• Fluid saturation acts to stiffen the pores to
  deformation affects P-waves, but not shear-waves
• Saturation increases the density of the material
  and increases both compressional- and shear-wave
  velocity, and increases attenuation
• Dilatency can cause expansion and permeability

18
State 2-14 Borehole Data
Typical of Inversion Results
Daley et al., 1988 Tarif et al., 1988
19
Poissons Ratio from Vp and Vs
Daley et al., 1988
Monotonic decrease of PR due to compaction and
lithification of sediments
Permeable zone 915 m 960 m loss of
circulation Paillet et al., 1986
Hydrothermal alteration sulfide,
chlorite,epidote 1220 m Elders and Sass, 1986
20
Roadmap

• locations of microearthquakes
• instrumentation and data processing
• high resolution locations for mapping
• rock physics interpretations
• tomographic inversion
• shear wave splitting
• Joint inversion techniques

21
Velocity Variations from Expected Geysers'
graywacke
22
Attenuation Variations from Expected

• l

23
Anomalous Zones
a
c
b

• (a) Qp gt 50 and Vp gt 5 from expected model
• (b) Qp gt 50 and Vp lt 5 from expected model
• (c) Qp lt 50 and Vp lt 5 from expected model
• Felsite formation (blue shading) and pressure
  contours also shown.

24
Anomalous Zones

• Qp gt 50 and Vp gt 5 (Red - felsite, Green -
  graywacke) Consistent with reduced fracture
  density.
• Qp gt 50 and Vp lt 5 (Blue - graywacke) Effect
  of fluid and compositional changes.
• Qp lt 50 and Vp lt 5 (Brown - graywacke)
  Consistent with a high fracture density.

25
Roadmap

• locations of microearthquakes
• instrumentation and data processing
• high resolution locations for mapping
• rock physics interpretations
• tomographic inversion
• shear-wave splitting
• Joint inversion techniques

26
Shear-wave Splitting
27
Shear-wave Polarization Directions

• Seismic anisotropy deduce information about the
  fracture orientation and fracture intensity.
• from Stroujkova and Malin, 2000, Long Valley

28
Roadmap

• locations of microearthquakes
• instrumentation and data processing
• high resolution locations for mapping
• rock physics interpretations
• tomographic inversion
• shear-wave splitting
• Joint inversion techniques

29
joint inversion techniques

• yields model solutions that are consistent with
  all of the data
• obtain a probabilistic approximation to the
  solution of a problem by using statistical
  sampling techniques, in such a way that the
  probability of a model (proposed solution)
  depends only upon the immediately preceding model
  and the available data.
• make an educated guess at the solution to a
  difficult inverse problem, runs a forward code to
  determine the output data corresponding to the
  model, then checks the model data against
  actual data to determine whether to accept the
  current model as a better approximation, or back
  up to the previous model and guess again.

30
Example Dixie Valley (Foxall et al.)?

• Drilling Requires multiple, steeply dipping
  faults
• Reflection allows single range-front fault or
  complex faulting
• Gravity main density contrasts displaced
  valleyward of valley-range contact
• Temperature outflow of hot fluid into valley
  fill, possibly from faults within valley basement
• Well-flow/tracer tests heterogeneous
  permeability along-strike

Blackwell (2002)?
31
Bayesian inference is implemented with a Monte
Carlo-Markov Chain (MCMC) search of the model
space

• Staged inversion produces solutions that are
  consistent with all of the data

32
The Stochastic Engine produced better results,
but at a larger computational cost?

• 107 degrees of freedom
• 103 - 104 forward calculations
• highly non-linear (multiBH code)?
• 12 hours computation time on 128- node
  cluster

MCMC Inversion
Deterministic Inversion

• 10 -20 iterations/forward calculations

33
Impact of Work

• More rigorous and accurate analysis of existing
  data for a more complete description of resource
• better informed well siting decisions
• Quantitative assessments of uncertainty and
  sensitivity
• probability distribution of alternative models
• new exploration - risk-based strategies
• Modest commercial computing capabilities will
  soon be adequate to meet demands of MCMC