The Electrical Properties of Saturated Crustal Rocks: Measurement and Implications. Paul Glover University Montpellier II

Transcription

1 The Electrical Properties of Saturated Crustal Rocks: Measurement and Implications Paul Glover University Montpellier II 1

2 Structure Basics Ignoring the difficult mathematics Past Experiments Benchtop Multisalinity Confining Pressure Deformation Developments in Montpellier Application to the Pyrennees 2

3 Basic Theory 3

4 Causes in Rocks Matrix Fluid Surface Melt Rock Forming Minerals Minor Minerals Electrolyte Double Layer Molten Rock Hydrated Amphiboles & Clay Minerals Electronic Metal Oxides & Sulphides, & Graphite Electronic Natural Brines Ionic +ve ions on mineral surfaces Ionic ions in melt Ionic 4

5 Fundamentals of Measurement Electrical properties are incredibly sensitive to changes in the rock microstructure Therefore, ideally suited as a probe In-phase and out-of-phase components Vary with frequency (Impedance Spectroscopy) Can vary with AC amplitude (AC Voltammetry) 5

6 Related Measured Variables Measured In-phase impedance, Z Out of phase impedance, Z In-phase admittance, Y Out of phase admittance, Y Scaled real permittivity, e Scaled imaginary permittivity, e Length, L Diameter, D Permeability, k Porosity, f Electrical properties of fluids Specific In-phase resistivity, ρ Out of phase resistivity, ρ In-phase conductivity, σ Out of phase conductivity, σ Real permittivity, ε Imaginary permittivity, ε Real realtive permittivity, K Imag. relative permittivity, K Phase angle, θ Electrical properties of fluids 6

7 Basic Equations I * Z = Z' + i Z " * " * 1 Y ' i Y where Y * Z Y = + = σ ρ * " * 1 = σ ' + i σ where σ = and σ = * ρ * " * 1 = ρ' + i ρ where ρ = and ρ = * σ * * Y L A Z * * L A 7

8 8 Basic Equations II ω σ ε ε ε ε i where i * * " * ' = + = = + = ε ε * * " * ' K where K i K K 2 " ' tan 1 π σ σ θ = σ e μ = n *

9 Argand Plot - ρ 1 Hz Response 1 MHz Functions: ρ Bode Plot Debye ρ ρ ρ ρ ρ o = iωτ - ρ 1 Hz log (Frequency) 1 MHz 9

10 Argand Plot - ρ 1 Hz Response Functions: 1 MHz α ρ Bode Plot Cole & Cole ρ ρ = ρ + ρ o 1 + i ρ ( ) ( 1 ωτ α ) - ρ 1 Hz log (Frequency) 1 MHz 10

11 Rocks Basic Constant Frequency Archie s Law (Archie, 1942) m eff = σ f χ f hence σ τ = ( ) 1 m χ Simple mixing laws to account for only the conductivity of the fluid saturating the pores of the rocks Modified Archie s Law (Glover et al., 2000) (log (1 χ ) log (1 χ )) σ eff = σ mat (1 χ f ) m f f + σ f χ m f 11

12 1 Conventional: m=1.0 Archie s Law Old & New Effective Conductivity (S/m) Conventional: m=1.5 Conventional: m=2.0 Conventional: m=2.5 Modified: m=1.0 Modified: m=1.5 Modified: m=2.0 Modified: m= Volume Fraction of Phase B 12

13 Rocks Added Constant Frequency Bussian (1983) 1 σ eff = σ f + m F 1 F ( ( ) σ ) s Complex theory containing surface conduction BUT still no frequency dependence Revil & Glover (1998) σ f σ = Fξ + ξ ξ + ξ 4, for Fξ ξ F ( ) ( ), 13

14 Fluid Flow in Rocks There exists: An undisturbed central zone of laminar flow, and A surface boundary layer of turbulent flow, and Zero flow at the rock surface 14

15 Electrical Conduction in Rocks There exists: A -ve charged rock surface, and A layer of +ve adsorbed ions, and A net ve diffuse layer [thickness f(salinity)] and Net neutral bulk fluid 15

16 Electrical Conduction in Rocks Boundary of moveable fluids is in diffuse layer Flow separates ve charges to the right and +ve charges are left behind this generates a potential difference called the STREAMING POTENTIAL 16

17 Challenges 17

18 Electrical Properties in Geosciences Theory of bulk and surface conduction in saturated rocks Generalisation to multi-frequency space Improvement of interpretation and analysis of MT data Electrical precursor signals associated with earthquakes Electrical signals associated with volcanic activity Fluid flow mapping in the crust using remote electrical tomography Improved borehole and remote tools for the oil and water industries Characterisation of sites for nuclear waste storage 18

19 Some Progress 19

20 MultiSalinity Experiments (CoCw) 10 Prepare a range of solutions of different salinities Measure the conductivities of each of the solutions Saturate the rock with solution 1 Measure the conductivity of the rock Replace with solution 2 Measure the conductivity of the rock again, and so on Conductivity of Rock (S/m) Clean SST (Berea) Clayey SST (Darley Dale) Conductivity of Solution (S/m) 20

21 Confining Pressure Saturated rocks confined in a hysrostatic oil pressure vessel 4 electrode Pt-blacked Pt system used 0 to 50 and 25 to 400 MPa ranges (2 vessels) Frequency sweeps carried out Rock containing low salinity fluids Size of dispersion curve inflates as pore structure collapses with pressure Shape of dispersion curve does not change 21

22 The Bode Plots 1600 In Phase Resistivity ( ohm.m ) MPa 10 MPa 20 MPa 30 MPa 40 MPa 50 MPa 1.E+01 1.E+02 1.E+03 1.E+04 1.E+05 1.E+06 Frequency (Hz) Out of Phase Resistivity ( ohm.m ) MPa 10 MPa 20 MPa 30 MPa 40 MPa 50 MPa 0 1.E+01 1.E+02 1.E+03 1.E+04 1.E+05 1.E+06 Frequency (Hz) 22

23 The Argand Plot Out of Phase Resistivity (ohm.m) MPa 10 MPa 20 MPa 30 MPa 40 MPa 50 MPa Confining Pressure Increasing In Phase Resitivity (ohm.m) 23

24 Uniaxial Deformation Saturated rocks confined in a load frame 4 electrode Pt-blacked Pt system used No confining pressure, 0 to 1.5% strain at /s Electrical properties measured at 1000 Hz Rock containing low salinity fluids Shape of conductivity measurements indicate that they mirror two sets of microcracks closing and opening 24

25 See 1 slide 25

26 Triaxial Deformation I Saturated rocks confined in a hydrostatic oil pressure vessel and load frame 4 electrode Pt-blacked Pt system used 50 MPa confining pressure, 0 to 2% strain at /s Electrical properties measured at 1000 Hz and frequency sweeps Rock containing low salinity fluids 26

27 Triaxial Deformation II Size of dispersion curve changes with deformation Initially bigger as cracks close Then smaller as new cracks form and link Shape is similar indicating that the dispersion mechanism is not changing Single frequency measurements indicate that the conductivity is a direction sensitive crack damage parameter This has been used to successfully reconstruct the measured stress-strain curve 27

28 See 4 slides 28

29 Developments at the Université Montpellier II 29

30 The New Measurement Cell Voltage Connexion 2 Fluid Inlet Current Connexion 2 Polyolefin Sleeving PTFE End-piece Sample Perspex Pressure Vessel Perspex Spacers Perspex End Closure BNC Coaxial Connectors Voltage Connexion 1 Fluid Outlet Current Connexion 1 Confining Pressure Inlet Current Electrode Blacked Pt Gauze Filter Paper Voltage Electrode Blacked Pt Gauze Pt100 Sensor Porous Disc SMA Gold-Plated Coaxial Connectors Temperature Lead-through Swagelock Fittings (Drilled-out) 30

31 The Cell Fully Assembled 31

32 The Cell in its Retaining Box Non-conducting materials 4 or 2 electrodes Pt-blacked Pt Gauze Heat-shrink sleeving with silicone 50 Ω miniature coaxial leads with SMC connectors 32

33 P Fluid resur e Solartr Impe dance on129 4 Tempe raturec ontrol er Loger Inte rface 5 0TerΩ minatio ns Fans (x2) S tirers Re servoirs Solartr F RA on1260 Plex iglas Cabine Heater PreTrans ducersure Transdu Pres ure cer ReturnT Pump 1 Reservo (Syrin Pump 2ge) Mea sureme ntcel (Perista ltic) Pt10 1 0BarGa CP resure onfining Schematic Diagram of the Rig Rig in temperatures controlled enclosure Heater, fans and special control box Measurements made from 10 μhz to 32 MHz Solartron 1260 FRA Logging of fluid pressures, with back pressure 33

34 The Assembled Rig Entire fluid process in temperature controlled environment All measurements logged to PC Shielded, low electrical noise apparatus 34

35 Dummy Test Measurements 10 μhz to 32 MHz Noise-free measurements Single dispersion curve for 501 nf element In Phase Impedance (Ω) = log10 (1/2πRC) = 3.18 khz Out of Phase Impedance (Ω) log10 (Frequency (Hz)) 35

36 New Fluid Change Experiments Single Frequency 200 Hz Changing Fluids from Distilled Water to 1 M NaCl Porosity = 0.12 NaCl (M) m Distilled In Phase Resistivity (ohm.m) Distilled Water to 0.01 M NaCl 0.01 M NaCl to 0.1 M NaCl 0.1 M NaCl to 1 M NaCl Time (s) 36

37 New CoCw Measurements 1.E+00 1.E-01 Effective Conductivity (S/m) 1.E-02 1.E-03 1.E-04 1.E-05 Gres de Fontainebleau Gres de Clashach 1.E-06 1.E-04 1.E-03 1.E-02 1.E-01 1.E+00 1.E+01 Fluid Conductivity (S/m) 37

38 New Electro-Kinetic Measurements Gres de Fontainebleau 0 0 Zeta Potential (mv) Streaming Potential (mv) Pressure Difference (MPa)

39 New Impedance Measurements Gres de Fontainebleau Bode Diagram Argand Diagram No Flow 200 ml/hr 39

40 Spectroscopy of Carrara Marble Flow causes charge separation that takes 2 days to relax Now known to be ARTIFACT In Phase Resistivity (ohm.m) 150 Initial No Flow 130 Flow Stopped 0.5 hrs Flow Stopped 1.33 hrs 110 Flow Stopped 4.23 hrs Flow Stopped hrs Flow Stopped hrs E-01 1.E+00 1.E+01 1.E+02 1.E+03 1.E+04 1.E+05 1.E+06 1.E+07 Out of Phase Resistivity (ohm.m) Initial No Flow Flow Stopped 0.5 hrs Flow Stopped 1.33 hrs Flow Stopped 4.23 hrs Flow Stopped hrs Flow Stopped hrs Out of Phase Resistivity (ohm.m) Initial No Flow Flow Stopped 0.5 hrs Flow Stopped 1.33 hrs Flow Stopped 4.23 hrs Flow Stopped hrs Flow Stopped hrs Frequency (Hz) In Phase Resistivity (ohm.m) 5 1.E-01 1.E+00 1.E+01 1.E+02 1.E+03 1.E+04 1.E+05 1.E+06 1.E+07 Frequency (Hz) 40

41 Crustal Conductivity Modelling 41

42 Location of the Study 42

43 Questions What are the mechanisms of the conductivities in the crust and mantle? What is the mechanism of the high conductivities in the slab? If the slab high conductivities are caused by partial melting, what is the partial melt fraction and what is the melt connectivity? Why is there no surface volcanism in the Pyrenees? 43

44 Fluid Salinity & Conductivity Lab. Rock Conductivities Elevation Model Structure Pressure Predicted Conductivitiesor Melt Fractions Parallel Layer Thermal Properties Gravity Seismic Structure Lithology Density Rock Conductivity Prescribed Melt Fraction Melt Conductivity Waff/HS Upper Bound Random Areas Modified Archie s Law HS Lower Bound Lab. Melt Conductivities Temperature Perp. Layer MT Conductivity For Comparison 44

45 Mixing Models 6 mixing Models have been used: Parallel model (arithmetic mean) Hashin-Shtrikman upper bound Waff s model Random model (geometric mean) Modified Archie s law Hashin-Shtrikman lower bound Perpendicular model (harmonic mean) Well Connected Melt Moderately Connected Melt Badly Connected Melt 45

46 Effective Conductivity MT Observed Conductivities 46

47 Effective Conductivity Hashin- Shtrikman Upper Bound/Waff s Model 47

48 Effective Conductivity Hashin- Shtrikman Lower Bound 48

49 All Effective Conductivity Models 49

50 Melt Fraction Hashin- Shtrikman Upper Bound/Waff s Model 50

51 Melt Fraction Hashin- Shtrikman Lower Bound 51

52 Summary I A two-dimensional conductivity model for the Pyrenees has been constructed A good match to the conductivities observed by MT is possible Aqueous fluids alone can explain the conductivity in most of the profile Aqueous fluids cannot explain the conductivity of the subducting slab 52

53 Summary II Partial melting is likely to be the cause of the very high slab conductivities A partial melt fraction of at least 4.7% is necessary This is consistent with geochemical melting models The melt must be well connected The absence of surface volcanism is partly due to its compressive tectonic regime, and volcanism is likely in the Pyrenees if the area becomes extensive 53