Transcript PowerPoint-presentasjon

Microseismic emissions – Heartbeats of a reservoir
Programseminar for Norges forskningsråds Olje og gass program
3. – 4. April 2003, Stavanger
Michael Roth, NORSAR
Introduction
Processing
Applications
Outlook
Related projects
Introduction
Projects:
• Internal strategic institute program
• Research project (NFR, Read Well Sevices, Statoil, TFE)
• PhD - project (NFR)
Objectives:
• Development of an automatic microseismic monitoring system
• Real-time processing and localization
• Visualization of seismic events with subsurface structural model
• Analysis and interpretation of microseismic data
Prototype monitoring software for microseismic data as
recorded with 3C-geophones installed in an observation well
Introduction
Production of hydrocarbon reservoirs
Hydrofracturing
Geothermal energy production
Subsurface gas storage
Mining activity
Changes in stress field,
pore pressure and load
Microseismic events
Analyses of microseismic data has the
potential to:
• image the internal structure of the
subsurface (faults, fractures)
• monitor fluid pressure front movements
• identify sealed-off reservoir volumes
• identify regions of reservoir compaction
• provide input for reservoir management
• provide feedback during hydrofracturing
• map thermal fronts
• provide input for hazard mitigation
Introduction
Typical features of microseismic data
Signal frequency:
”Fault plane radius”:
Seismogram duration:
Magnitude:
Source–receiver distance:
Frequency of occurrence:
100 - 1000 Hz
10 – 1 m
<1s
–3 – 0 Mb
up to 1000 m
10 –1000 per hour
Observation with:
3C-geophones receiver
downhole installation
Processing
Calibration
Detection
P-onset
Polarization
S-onset
P-wave onset, azimuth, incidence, S-wave onset
Velocity model,
Receiver coordinates
Localization
Hypocenter
Processing
Z
Calibration
H1
H2
H1
H2
H2
H1
H2
H1
arbitrary
orientation
Processing
Z
Z
Calibration
H2
H1
polarization
analysis
of P-wave signal
H2
H1
shot position,
geophone position,
velocity model
H2
H1
H2
H1
measured azimuths (am)i
theoretical azimuths (ath)i
H2
H2
H1
H1
differences Dai = (am - ath)i
H2
H1
arbitrary
orientation
H2
H1
rotation by Dai
consistent
orientation
Processing
Processing
Calibration
Detection
P-onset
Polarization
S-onset
P-wave onset, azimuth, incidence, S-wave onset
Velocity model,
Receiver coordinates
Localization
Hypocenter
Processing
Detection
based on the evaluation of signal-to-noise ratio
Processing
P-onset determination
• compute AR model
• error prediction filtering
• compute AIC function
• find AIC minimum
Processing
Detection
P-onset
Polarization
S-onset
P-wave onset, azimuth, incidence, S-wave onset
Velocity model,
Receiver coordinates
Localization
Hypocenter
Processing
Localization (3D velocity model)
3D raytracing for specific
receiver geometry and
3D velocity model
Tabulated travel
times and angles
Directed grid search
Applications
Hydrofracturing Data Set
• 12 3C geophones deployed in a vertical observation well
• gimbal mounted, i.e. vertical components are oriented
• non-rigid connection, i.e. horizontal orientation is inconsistent
• 10 m receiver spacing
• ~ 100 m distance to injection well
• ~ 1h contineous recording, sampling interval 0.5 msec
• ~ 1000 microseismic events
• homogeneous velocity model
Applications
Data Example
Left:
4 s time window
4 microseismic events
~ 0.2 s signal duration
Right:
Close up of the first event
(channel 4 – 36)
Applications
Data Example
Left:
4 s time window
4 microseismic events
~ 0.2 s signal duration
Right:
Close up of the first event
(channel 4 – 6)
Applications
Data Example
Left:
4 s time window
4 microseismic events
~ 0.2 s signal duration
Right:
Close up of the last event
(channel 31 – 36)
Applications
P-wave polarization
Applications
Hypocenters of microseismic events
View from SW
Applications
View from SE
Color-coded origin time
(early = blue, late = red)
Applications
Ekofisk Data Set
• 6 3C geophones deployed in a
vertical observation well
• passive listening in 3 km depth
close to the hydrocarbon reservoir
• 20 m receiver spacing
• event-triggered recording for 18
days, 1 ms sampling intervall
• ~ 4000 events
Applications
3D raytracing for EKOFISK velocity model
Velocity model has a strong
vertical gradient at the receiver
depth
Model volume: (2 km)3
413 potential sources on a
regular grid with 50 m spacing
6 receivers
Application of reciprocity:
6 sources and 413 receivers
Applications
Applications
Applications
Applications
Radius = magnitude
Color = location error
Outlook
Future plans for monitoring software
• Restructuring for heterogenous network
(3C-, 1C-geophones, hydrophones)
• Restructuring for arbitrary receiver deployment
(wells, surface, sea floor)
• Facilitate interactive processing for selected events
• Waveform analyses
• Source mechanisms
Related Projects
Related Projects
San Andreas Fault Observatory at Depth (SAFOD)
Cooperation with:
Duke University, North Carolina
US Geological Survey, Menlo Park, California
• penetrate fault zone
• directly sample fault zone materials
• directly measure physical and
chemical fault zone properties
• monitor an active fault at depth
•pilot hole to identify and localize
earthquake targets
•1.8 km distance to San Andreas Fault
•32 3C geophones
• depth 850 – 2090 m
• 40 m spacing
• 1 ms sampling
Related Projects
Pyhaesalmi Ore mine, Finland
Research cooperation with
Lawrence Livermore National Laboratory
• Monitoring of explosions and rockbursts
• 4 3C and 12 1C geophones deployed in
drifts in the lower part of the mine
• 3D velocity model
• 3D receiver distribution
• Determination of hypocenters and
radiation patterns
Rock-slope failures: Models and risk assessment
Related Projects
Monitoring of potential
rockslide sites in Norway
• Extensometers
• Photogrametry
• GPS
• Synthetic Aperture Radar
• Laser measurements
• Microseismic monitoring
Preliminary regional rock-avalanche hazard zones in Møre & Romsdal