Aucun titre de diapositive

1
Étude in-situ des interactions hydromécaniques
entre fluides et failles Application au
laboratoire du rift de Corinthe
ÐOÀN Mai Linh Institut de Physique du Globe de
Paris
2
Fluid-fault interactions
Example of fluid-fault hydromechanical coupling
Fault-valve mechanism (Sibson70)
Fault closed
3
Motivations

• But field data
• altered outcrops
• after slip
• dynamical seismics
• indirect

After Matthai (1992)
4
Structure of the presentation
I Presentation of the Gulf of Corinth
and the DGLAB project II Characterization of the
hydraulic setting III A peculiar kind of
hydraulic transients Events triggered by
far earthquakes
I Presentation of the Gulf of Corinth
and the DGLAB project II Characterization of the
hydraulic setting III A peculiar kind of
hydraulic transients Events triggered by
far earthquakes
5
The Corinth Rift
Greece
Complicated geodynamics
Complex geology
From Jolivet (2005)
6
The Corinth Rift
After Bernard (1997)
7
Deep Geodynamic LABoratory
South
North
0.50.1MPa
8
Initial hydraulic knowledge of the Aigio fault
Impervious fault

• Difference in overpressure

9
Initial hydraulic knowledge of upper aquifer
Drawdown m
Hydraulic tests by GFZ July 2003
10
Initial hydraulic knowledge of the karst
k1-1.5 10-5 m/s No storativity
11
AIG10 permanent sensors
12
Pressure sensors
Tides
Log10(Pressure MPa)
2 absolute pressure gauges - high precision - low
precision 1 relative pressure gauge - hydrophone
Log10(Frequency Hz)
13
Structure of the presentation
I Presentation of the Gulf of Corinth
and the DGLAB project II Characterization of the
hydraulic setting III A peculiar kind of
hydraulic transients Events triggered by
far earthquakes
14
Quality of the pressure signal
Pressure
Pressure (Bar)
UT Time
Resolution better than 1
The pressure is similar to that of the karst
?The karst dominates the measured pressure
15
Strategy
Tidal calibration
Tidal calibration
Long-term fluctuations
How sensitive is the pressure signal to
deformation ?
What are the dimensions of the aquifers ?
How water flows through the aquifers ?
Thermal Regime
16
Tidal inversion
??
Triple origine

• Earth Tide
• (Prediction ETERNA 3.3)

Aigion
17
Analysis of the tidal signal
Linear regression on the input data
Ouput pressure in Aigio
Input Tide gage in Trizonia
Input Barometric pressure in Temeni
Input Theoretical tidal strain in Aigio
18
Analysis of the tidal signal
dP2.748 10-4 dhoc 1.784 10-4 d?ter
No offset
19
Barometric effect
Bad weather at the end of the year 2003 Observed
pressure (detided) Û Atmospheric pressure
20
Interpretation of the coefficients
Poroelastic model (large wavelengths)
B Skempton coefficient Ku Undrained bulk
modulus ?u Undrained Poisson ratio ?
Barometric efficiency
?0.30.1
B Ku171GPa
21
Oceanic load
S
Aig10
N
Oceanic load
Loading profile at a depth of 700m induced by a
unit load
sxxszz/2?gh
AIG10
Distance to southern shore (m)
The oceanic load should induce a phase lag !
22
Influence of boundaries
S
Aig10
N
Oceanic load
Aigio fault
Helike fault
Can the presence of impervious faults
explain this absence of phase lag ?
Analytical prediction of phase lag for a 1D
aquifer with impervious boundaries
23
Oceanic load
Map of semi-diurnal phase lag () for a
semi-infinite ocean
Phase lag -5 min 5min ß -2.5 2.5
L
x/L
24
Is Aigio fault impervious at all depths ?
25
Tidal information
Tidal calibration
Tidal calibration
Long-term fluctuations
How sensitive is the pressure signal to
deformation ?
Poroelastic parameters ? excellent  strain 
sensor
What are the dimensions of the aquifers ?
How water flows through the aquifers ?
Karst confined in a NS direction. By Aigio fault ?
Storativity ? Hydraulic diffusivity
Thermal regime
26
Long-term data
Pressure
Flow between the two previously independent
aquifers
No sharp seasonal variations
14 kPa
Pressure (bar)
Time
1 year
27
Analytical solution
Axisymmetric analytical solutions ? Finite
aquifers ? Transients controlled by the radii of
the aquifers and borehole radius
28
Development of the FEM2.1D method
1. Finite Element Method 2D to describe
flow in upper and lower aquifers
2. Manual coupling at a well node
(0.1D) Same pressure Mass
conservation of fluid

• Efficient
• Keep the characteristic
• distance of the well radius

29
Dimensions of the aquifers ?
Can the decrease in pressure observed during the
first 3 months provide constraints on the
dimensions of the aquifers ?
Rectangular-shaped aquifers ? 4
unknowns Hydraulic properties of the upper
aquifer ? 1 unknown (storativity) 2 pieces of
information to fit amplitude and duration of
the drop
Try to find plausible configurations
30
Dimensions of the aquifers ?
Upper aquifer LNS1000m LWE200m Lower
aquifer LNS5000m LWE?
Pressure (bar)
Pertinence of The homogeneous Model for the karst
?
Time (days)
31
Long-term information
Tidal calibration
Tidal calibration
Long-term fluctuations
Poroelastic parameters ? excellent  strain 
sensor
Storativity ? Hydraulic diffusivity
Karst confined in a NS direction
Hydraulic diffusivity (Almost) no flow
Both aquifers are confined
Thermal Regime
32
Thermal profile
1 year after drilling
Temperature (C)
Depth (m)
33
Heat flow measurement
5010 mW/m2
22C/km
34
Karst convection
? qb 70mW/m² ? qb100mW/m² ? qb200mW/m²
Fault vertical offset150m ? ?zt-770m ? lt150m
Relation Tt?zt from extrapolation of Thermal
gradient
H gt 400 m
35
Thermal anomaly
Heat generated by fault slip does not explain
this anomaly
36
Thermal information
Tidal calibration
Tidal calibration
Long-term fluctuations
Poroelastic parameters ? excellent  strain 
sensor
Hydraulic diffusivity (Almost) no flow
Hydraulic diffusivity Internal advection
Both aquifers are confined
Both aquifers are confined Large vertical
extension
Thermal regime
Low heat flow 5010 mW/m²
37
Structure of the presentation
I Presentation of the Gulf of Corinth
and the DGLAB project II Characterization of the
hydraulic setting III A peculiar kind of
hydraulic transients Events triggered by
far earthquakes
38
A panel of hydraulic anomalies
2minute-long
10minute-long
10-200 Pa
10-200 Pa
100 events/yr
20 events/yr
Pressure
2minute-long
30minute-long
10-400 Pa
50-60 Pa
200 events/yr
2 events/yr
Only associated with teleseismic transients
Time
39
The Mw7.8 Rat Island Earthquake
November, 17th 2003 0643 UTC
Drop of 60 Pa (equivalent to 3.5nstr)
BKu17GPa determined from tidal analysis
30min
5min
40
Review of triggered hydraulic anomalies
Straingt10-8
2003 Rat Island Event
Magnitude
Strainlt10-8
Distance to epicenter (km)
After Montgomery and Manga (2003)
41
Comparison with other local sensors
Trizonia
Aigio
LF signal
0 10km
Sacks-Evertson Strainmeter
STS2 broad-band Seismometer (North component)
42
Validity of the pressure data
Comparison of seismic oscillations of both
deformation sensors
Nyquist frequency of the pressure sensor
P
Frequency
?h
- Strainmeter - Pressure
Good correlation of both sensors
Time
43
Response to a dislocation
One single wellhead value
Average of pressure anomaly along the borehole
Poroelastic response HETEROGENEOUS along the
borehole
Fault movement
44
Response to a dislocation
Average of pressure along the borehole induced
by a double-couple located at (x,y)
M0?DS
Map of Log10(Pressure anomaly) for DS1m3
x
y
S slip area D relative displacement
Dip direction
z
x
y
Distance from borehole v?hydraulictrelaxation
DS1m3
lt5000m3 (Trizonia data)
45
High-frequency hydrophone data
Hydrophone Close-up
Pressure
UTC Time 070702
0715
0710
0705
0705
0715
0710
Hydrophone
0.000

0.100
0715
0710
0705
Time
46
Angle of slip
z
x
x
y
y
Fault plane
Slickensides
47
The Mw9 Sumatra event
Data acquisition problem ? Irregular sampling
Pressure in Aigio
December, 26th 2004 0058 UTC
P
S
Strain in Trizonia
48
Conclusion

• Hydraulic characterisation of AIG10
• We measure the pressure of the bottom karst
• Poroelastic response to both Earth tides and
  ocean load
• Þ Sensitive strain sensor
• Aquifers are confined with almost no flow at the
  boundaries
• and internal convection within the karst
• Aigio fault is impervious at the intersection
  with the borehole
• but is it the case below the Pindos nappe
• Low heat flow

• Hydraulic characterisation of AIG10
• It is now possible to model
• the wellhead pressure response to
• fault movement
• within an homogeneous poroelastic framework

Hydraulic anomalies The DGLAB project provides
the opportunity to study dynamic fluid-fault
interactions

• Hydraulic anomalies
• A large set of hydraulic anomalies.
• An anomalous hydraulic anomaly
• dynamically triggered by S waves from a
  teleseism,
• with a concomitant local microseismic event

49
Perspectives
Interpretation of the remaining hydraulic events

• Better knowledge of the surrounding seismicity
• Better interpretation of the hydrophone signal
• Better understanding of the aquifer and its
  heterogeneities

• But we monitor fluids around a fault
• rather than fluids inside a fault
• But no independent evaluation
• of fluid evolution and fault movement

50
Perspectives
Expected full instrumentation
0 m
Hydrophone
Installation of the whole instrumentation schedule
d in March 2006
Hydrophone
700 m
High-precision pressure gage
750 m
High-precision pressure gage
870 m
3C Seismometer
1000 m
51
(No Transcript)
52
Link between storativities
S? uniaxial storativity ?11 ?220, d?330 S?
unstrained storativity ?0 S? strained
storativity d?0
S? ? S? ?S?
S? (1-aB) S?
53
The AIG10 borehole
0.50.1MPa
54
Age of the karst water
Simple optimistic model Cermák model
Tgt1000 yr (In accordance with the absence of
Tritium in water)
55
The lower aquifer is karstic
900
800
56
Are the aquifers well confined ?
57
All the three studies were necessary
Tidal calibration
Tidal calibration
Long-term fluctuations
Thermal regime
58
Pressure sensors
Tides
Log10(Pressure MPa)
2 absolute pressure gauges - high precision - low
precision 1 relative pressure gauge - hydrophone
Log10(Frequency Hz)
59
Development of the FEM2.1D method
Analytical axisymmetric solutions shows that the
transitory regime is partly controlled by the
borehole radius
1. Finite Element Method 2D on each
aquifer 2. Manual coupling at a well
node (0.1D)

• Efficient
• Keep the characteristic
• distance of the well radius

60
Conclusion

• Hydraulic characterisation of AIG10
• We measure the pressure of the bottom karst
• Poroelastic response to both Earth tides and
  ocean load
• Þ Sensitive strain sensor
• Aquifers are confined with almost no flow at the
  boundaries
• but internal convection within the karst
• Aigio fault is impervious at the intersection
  with the borehole
• but is it the case below ?
• Low heat flow and rigid block. Not a process
  zone.

• Hydraulic anomalies
• A large set of hydraulic anomalies.
• A anomalous hydraulic anomaly
• dynamically triggered by S waves from a
  teleseism,
• with concomitant a local seismic event
• Borehole instrumentation provides tools
• to understand the triggering mechanism