06.08.2013 Views

Rock Physics of Shale - Stanford University

Rock Physics of Shale - Stanford University

Rock Physics of Shale - Stanford University

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

<strong>Rock</strong> <strong>Physics</strong> <strong>of</strong> <strong>Shale</strong><br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory<br />

1


First Question: What is <strong>Shale</strong>?<br />

<strong>Shale</strong> -- a rock composed <strong>of</strong> mud-sized particles, such as silt and<br />

clay (Boggs, 2001). This most general classification is based on<br />

particle size, not composition.<br />

Variations in usage:<br />

• <strong>Shale</strong> is sometimes used to refer only to fissile rocks made <strong>of</strong><br />

mud-sized particles, while<br />

• Mudstone is sometimes used to refer to non-fissile rocks made<br />

<strong>of</strong> mud-sized particles, and<br />

• Siltstone is sometimes used for rock with mud-sized particles,<br />

but low clay fractions.<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory<br />

2


What is <strong>Shale</strong>?<br />

Wentworth Scale <strong>of</strong> grain size<br />

φ = −log 2 d<br />

€<br />

Mud<br />

( d < 30 µm)<br />

€<br />

Grain size<br />

(mm)<br />

φ<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory<br />

3


<strong>Shale</strong>s vary tremendously in composition!<br />

4<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Organic-Rich <strong>Shale</strong><br />

5<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Organic-Rich <strong>Shale</strong><br />

Organic-rich shales are composed <strong>of</strong> mixtures <strong>of</strong><br />

inorganic and organic materials. The inorganic<br />

fraction can include clay, silt, carbonate, pyrite,<br />

etc. The organic fraction (kerogen) appears as<br />

nano-size particles, called macerals, and<br />

hydrocarbons.<br />

In some cases organics appear as inclusions in<br />

the inorganic background, and other times, the<br />

inverse.<br />

Vanorio et al., June 2008, The Leading Edge.<br />

Confocal Laser Scanning Microscope image <strong>of</strong><br />

an organic-rich shale. Monterey Formation.<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory<br />

6


What the heck is shale porosity/permeability?<br />

• Network <strong>of</strong> permeable kerogen?<br />

• Inter-granular pores?<br />

• Micro/nano fractures?<br />

• Does hydrocarbon move via Darcy flow or molecular diffusion?<br />

Vanorio et al., June 2008, The Leading Edge.<br />

Wang, et a., SPE 124253<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory<br />

7


Porosity in Organic-rich <strong>Shale</strong><br />

Nano-scale image <strong>of</strong> pores within Kerogen.<br />

Wang, et a., SPE 124253<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory<br />

8


Porosity in Organic-rich <strong>Shale</strong><br />

Fossil fragment with micropores<br />

Nanopores around in-organic grains.<br />

Intergranular micro and nano-pores<br />

Nanopores in organic material.<br />

Loucks, et al.,, Journal <strong>of</strong> Sedimentary Research, 2009, v. 79, 848–861<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory<br />

9


Porosity in Organic-rich <strong>Shale</strong><br />

Nanopores in organic material.<br />

Loucks, et al.,, Journal <strong>of</strong> Sedimentary Research, 2009, v. 79, 848–861<br />

10<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Porosity in Organic-rich <strong>Shale</strong><br />

.<br />

Loucks, et al., Journal <strong>of</strong> Sedimentary Research, 2009, v. 79, 848–861<br />

Nanopores in organic matter.<br />

Loucks et al. observe that organic<br />

porosity increases with maturity.<br />

Diameter <strong>of</strong> CH 4 ~ 0.4 nm<br />

Methane<br />

11<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Organic-rich <strong>Shale</strong>s, Imaging and Nanoindentation<br />

Ahmadov, et al., The Leading Edge, January 2009.<br />

12<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Kerogen Moduli Inferred from Nanoindentation<br />

€<br />

Y<br />

µ<br />

Plausible<br />

Poisson’s Ratio<br />

K<br />

M<br />

Ahmadov, et al., The Leading Edge, January 2009.<br />

13<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Change <strong>of</strong> texture with maturity<br />

Mature Bakken Postmature Bakken<br />

Ahmadov, 2011, Ph.D. Dissertation, <strong>Stanford</strong> Univ.<br />

14<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Example Compositions<br />

Mineral Barnett (%) Marcellus (%)<br />

Quartz 35-50 10-60<br />

Clays (mostly illite) 10-50 10-35<br />

Calcite, dolomite,<br />

siderite<br />

0-30 3-50<br />

Feldspars 7 0-4<br />

Pyrite 5 5-13<br />

Phosphate, gypsum trace trace<br />

Mica 0 5-30<br />

A Comparative Study <strong>of</strong> the Mississippian Barnett <strong>Shale</strong>, Fort Worth Basin, and Devonian Marcellus <strong>Shale</strong>, Appalachian<br />

Basin, Bruner and Samosna, DOE/NETL-2011/1478.<br />

15<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Example Compositions<br />

16<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


What parameters are we interested in?<br />

Quantity: Two rocks with the same TOC …<br />

Dead<br />

Carbon<br />

TOC<br />

Live Carbon<br />

Live<br />

Carbon<br />

H/C = 1.15 H/C= 0.75<br />

Dead<br />

Carbon<br />

… but different Hydrocarbon-Generative Potential<br />

17<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


What parameters are we interested in?<br />

maturation<br />

gas window<br />

Quality; Maturity<br />

very oil prone<br />

oil prone<br />

inert<br />

gas prone<br />

oil window ca. 60–160 o C<br />

gas window ~ 150-200 o C<br />

Alginite: Fresh-water algae<br />

Exinite: Pollen, spores<br />

Cutinite: Land-plant cuticle<br />

Resinite: Land-plant resins<br />

Liptinite: land-plant lipids,<br />

marine algae<br />

Vitrinite: Land woody and<br />

cellulose materials<br />

Van Krevelen<br />

Diagram<br />

18<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


What parameters are we interested in?<br />

maturation<br />

Quality; Maturity<br />

very oil prone<br />

oil prone<br />

inert<br />

gas prone<br />

Type II marine shales give<br />

most production<br />

High maturity gives better<br />

gas delivery<br />

It has been suggested that<br />

shale in the gas window has<br />

more gas and improved gas<br />

mobility – modified pore<br />

structure? Reduced kerogen<br />

volume? Microcracks?<br />

Bob Cluff (SIPES 1832) SIPES 2009<br />

Annual Meeting, Hilton Head, S.C.<br />

19<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Vitrinite is a type <strong>of</strong> maceral with a shiny appearance resembling glass (vitreous). It is formed<br />

diagenetically by the thermal alteration <strong>of</strong> lignin and cellulose in plant cell walls. Chemically,<br />

it is composed <strong>of</strong> polymers, cellulose and lignin. It is therefore common in sedimentary rocks<br />

with a terriginous origin, containing plant matter.<br />

Vitrinite Reflectance, R o : percentage <strong>of</strong> light reflected from polished vitrinite samples,<br />

usually at wavelengths <strong>of</strong> 546 nm. R o is sensitive to temperature ranges that are responsible<br />

for hydrocarbon maturation; R o increases with thermal maturity <strong>of</strong> a source rock.<br />

20<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


€<br />

Passey’s 1 Method <strong>of</strong> TOC Estimation<br />

DlogR = log 10(Rd/Rd baseline)+0.02*(dtc − dtc baseline)<br />

(2.297−0.1688LOM )<br />

TOC = DlogR*10<br />

Rd = deep resisitivity (ohm-m)<br />

Rd baseline = deep resistivity in non-source shale<br />

dtc = sonic (usec/ft)<br />

dtcbaseline = sonic in non-source rock<br />

LOM = level <strong>of</strong> maturity<br />

TOC = total organic carbon (weight fraction)<br />

Sondergeld 2 et al find that Passey sometimes underpredicts TOC in<br />

mature and overmature gas sand<br />

1. Q. R. Passey, S. Creaney, J. B. Kulla, F. J. Morettand J. D. Stroud, AAPG Bulletin, V. 74, P 1777-1794, 1990 21<br />

2. Sondergeld et al., SPE Unconventional Gas Conference, 23-25 Feb, Pittsburg Pennsylvania<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Scale sonic and resistivity where 50 usec/ft<br />

equal 1 decade resistivity (ohm-m)<br />

22<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


The method employs the overlaying <strong>of</strong> a properly scaled porosity log<br />

(generally the sonic transit time curve) on a resistivity curve (preferably a<br />

deep-reading tool). In water-saturated, organic-lean rocks, the two curves<br />

parallel each other and can be overlain, since both curves respond to<br />

variations in formation porosity; however, in either hydrocarbon reservoir<br />

rocks or organic-rich non-reservoir rocks, a separation between the curves<br />

occurs. The separation in organic-rich intervals results from two effects: the<br />

porosity curve responds to the presence <strong>of</strong> low-density, low-velocity kerogen,<br />

and the resistivity curve responds to the formation fluid. In an immature<br />

organic-rich rock, where no hydrocarbons have been generated, the<br />

observed curve separation is due solely to the porosity curve response. In<br />

mature source rocks, in addition to the porosity curve response, the resistivity<br />

increases because <strong>of</strong> the presence <strong>of</strong> generated hydrocarbons. The<br />

magnitude <strong>of</strong> the curve separation in no-reservoirs is calibrated to total<br />

organic carbon and maturity, and allows for depth pr<strong>of</strong>iling <strong>of</strong> organic<br />

richness in the absence <strong>of</strong> sample data. This method allows organic richness<br />

to be accurately assess in a wide variety <strong>of</strong> lithologies and maturities using<br />

common well logs.”<br />

23<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


LOM (maturity) Estimation from R0<br />

24<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


What’s the <strong>Shale</strong>-Gas Geoscience Question?<br />

Porosity<br />

Matrix Perm<br />

Overall goal: high-flow gas wells in shale!<br />

Quartz,Calcite,Clay,TOC<br />

Content<br />

What determines a good (high-flow) well?<br />

Fractures<br />

Brittleness<br />

Tectonics<br />

Gas Content<br />

Maturity<br />

TOC<br />

Deposition<br />

Kerogen composition,<br />

Chemical & thermal history<br />

25<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


What’s the <strong>Shale</strong>-Gas Question?<br />

Basin<br />

Model<br />

What can we hope to detect from measurements?<br />

Porosity<br />

Matrix Perm<br />

Gas Content<br />

TOC<br />

Maturity<br />

Fractures<br />

Brittleness<br />

Azimuthal Anisotropy<br />

Low density<br />

High resistivity<br />

Low impedance<br />

High<br />

Impedance<br />

High Ur, Low K, Th<br />

Uplift<br />

Increased<br />

VTI anisotropy<br />

26<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


<strong>Shale</strong> Anisotropy<br />

27<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


What is Anisotropy?<br />

A rock is anisotropic when it has properties that depend on<br />

direction.<br />

• P- and S-wave velocity<br />

• P- and S-wave attenuation<br />

• Permeability<br />

• Electrical resistivity<br />

• Thermal conductivity<br />

Seismic waves see anisotropy, but in a more<br />

complex way than permeability or<br />

resistivity. Seismic propagation depends on<br />

the direction <strong>of</strong> propagation and the<br />

direction <strong>of</strong> polarization relative to the<br />

symmetry axes <strong>of</strong> the rock anisotropy.<br />

We define anisotropy at the scale <strong>of</strong> the measurement<br />

28<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Why Worry about Anisotropy?<br />

• It is everywhere.<br />

• It affects amplitudes, traveltimes, and ray paths<br />

• It changes fluid substitution procedures<br />

• It affects the seismic signature <strong>of</strong> pore pressure<br />

• It responds to stress<br />

It might also be a diagnostic <strong>of</strong> shale properties.<br />

29<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Practical Challenges<br />

• Predict / infer anisotropy from sparse data -- e.g. well logs<br />

• Incorporate anisotropy in seismic acquisition and imaging.<br />

• Understand seismic anisotropy for seismic interpretation<br />

(lithology, fluids, pressure, stress)<br />

• Map permeability anisotropy for reservoir management.<br />

30<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Sources <strong>of</strong> Anisotropy<br />

31<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Layering Anisotropy<br />

This will be present in all sedimentary<br />

basins. Anisotropy results from the<br />

contrasting layers, from the intrinsic<br />

anisotropy within layers, from<br />

superimposed fractures, and from stress<br />

differences.<br />

Intrinsic fabric anisotropy<br />

(e.g.shale) tends to dominate<br />

over the layering effect.<br />

The Backus Average is our<br />

primary rock physics tool for<br />

layered systems<br />

32<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Fabric Anisotropy<br />

Here we show a photo <strong>of</strong> beach sand, poured into a clear<br />

container. Even though it is relatively homogenous and<br />

well-sorted, there is a subtle fabric that translates into<br />

seismic anisotropy.<br />

If you can see a fabric, then there probably is anisotropy.<br />

Sand photograph<br />

Seismic Vp vs. confining pressure<br />

Vega, 2003, Ph.D. Dissertation, <strong>Stanford</strong> Univ.<br />

Horizontal<br />

Vertical<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Fabric Anisotropy<br />

Mylonite<br />

horizontal<br />

vertical<br />

Jones, 1983, Ph.D. Dissertation <strong>Stanford</strong><br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Fabric Anisotropy<br />

Gneiss<br />

horizontal<br />

vertical<br />

Jones, 1983, Ph.D. Dissertation <strong>Stanford</strong><br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Fabric Anisotropy<br />

<strong>Shale</strong><br />

horizontal<br />

vertical<br />

fast S<br />

slow S<br />

Tosaya, 1982, Ph.D. Dissertation <strong>Stanford</strong><br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


<strong>Shale</strong> Anisotropy<br />

Sources <strong>of</strong> shale anisotropy:<br />

• Alignment <strong>of</strong> platy clay minerals<br />

• Alignment <strong>of</strong> the domains (clusters <strong>of</strong> clay)<br />

• Alignment <strong>of</strong> non-spherical pores and microcracks<br />

• Alignment <strong>of</strong> fractures at scales much larger than the grains and<br />

pores<br />

• Fine-scale lamination <strong>of</strong> shaly materials with different stiffnesses<br />

• Lenses <strong>of</strong> kerogen<br />

<strong>Shale</strong> <strong>of</strong>ten has VTI symmetry<br />

37<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Aligned Fabric Anisotropy<br />

Bandyopadhyay, 2009, Ph.D. dissertation, <strong>Stanford</strong><br />

38<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Deposition/Compaction Hypotheses<br />

• <strong>Shale</strong> anisotropy should increase with effective stress/depth/compaction as clay grains<br />

become more and more aligned -- (Tosaya, 1982)<br />

• Dispersed sedimentation in a low energy environment leads to strong preferred<br />

orientation at the onset <strong>of</strong> sedimentation.<br />

• Bioturbation can randomize the texture<br />

• Flocculated clay aggregates have random orientation <strong>of</strong> clay particles. These aggregates<br />

might eventually collapse and align with compaction<br />

• Lab experiments seem to confirm some increase in anisotropy with compaction, but not<br />

to the extent <strong>of</strong> simple rock models.<br />

• Overpressure: disequilibrium compaction can slow alignment <strong>of</strong> grains, causing lower<br />

anisotropy. However, late-stage overpressure can open cracks in the fissile and increase<br />

anisotropy.<br />

Tosaya, 1982, Ph.D. Dissertation <strong>Stanford</strong><br />

Bandyopadhyay, 2009, Ph.D. dissertation, <strong>Stanford</strong><br />

39<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Organic-Rich <strong>Shale</strong> Anisotropy<br />

Presence <strong>of</strong> organic compounds requires a low energy<br />

environment, a rapid burial <strong>of</strong> organic materials, and anoxic<br />

bottom water for the preservation <strong>of</strong> the organic material – all<br />

favoring fabric alignments. Elastic anisotropy in organic-rich<br />

rocks is well established in Geophysics literature (Vernik and<br />

Landis, 1996; Vernik and Liu, 1997).<br />

Bandyopadhyay, 2009, Ph.D. dissertation, <strong>Stanford</strong><br />

40<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Velocity Anisotropy from Thinly Layered Kerogen<br />

Velocities in kerogen-rich Bakken shales (Vernik, 1990) and other low porosity<br />

argillaceous rocks (Lo et al., 1985; Tosaya, 1982; Vernik et al., 1987). Compiled<br />

by Vernik, 1990.<br />

<strong>Stanford</strong> Gary <strong>Rock</strong> Mavko <strong>Physics</strong> – <strong>Stanford</strong> Laboratory <strong>Rock</strong> <strong>Physics</strong> - Gary Laboratory Mavko<br />

F.24


A negative correlation is <strong>of</strong>ten found between the clay fabric intensity and the<br />

quartz content in shale.<br />

Increasing<br />

alignment<br />

Data from Curtis et al., 1980.<br />

Clay orientation distribution from XRD.<br />

from: Bandyopadhyay, 2009, Ph.D. Dissertation, <strong>Stanford</strong><br />

Data from Johnston and Christensen, 1995.<br />

Anisotropy vs. quartz fraction.<br />

Johnston, J. E., and N. I., Christensen, 1995, Journal <strong>of</strong> Geophysical Research B, 100, 5991–6003.<br />

Curtis, C. D., S. R., Lipshie, G., Oertel, and M. J., Pearson, 1980, : Sedimentology, 27, 333-339.<br />

42<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Velocity Sensitivity to Effective Pressure<br />

Vertical Vp<br />

Increasing<br />

maturity<br />

Vernik and Liu, 1997, GEOPHYSICS, 62, 521–532.<br />

Vernik, GEOPHYSICS, 1994, 59, 555-563<br />

Vernik and Landis, 1996, AAPG Bulletin, 80, 531–544.<br />

Vanorio et al., June 2008, The Leading Edge.<br />

Pressure sensitivity indicates<br />

cracks. For these samples,<br />

higher maturity samples have<br />

more stress sensitivity.<br />

43<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Anisotropy Sensitivity to Pressure, Maturity<br />

Vp anisotropy<br />

Increasing<br />

Peff<br />

Vanorio et al., 2008<br />

Vernik and Liu, 1997, GEOPHYSICS, 62, 521–532.<br />

Vernik, GEOPHYSICS, 1994, 59, 555-563<br />

Vernik and Landis, 1996, AAPG Bulletin, 80, 531–544.<br />

Vanorio et al., June 2008, The Leading Edge.<br />

More crack anisotropy at<br />

higher maturity.<br />

Less “intrinsic” anisotropy at<br />

higher maturity.<br />

44<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Lab data: Anisotropy vs. Effective Pressure<br />

Colored by TOC<br />

Niobrara<br />

Vernik and Liu, 1997, GEOPHYSICS, 62, 521–532<br />

45<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Lab data: Anisotropy vs. Effective Pressure<br />

Colored by R0<br />

Niobrara<br />

Vernik and Liu, 1997, GEOPHYSICS, 62, 521–532<br />

46<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Vertical Fractures<br />

47<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Fracture Anisotropy<br />

48<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Fracture Permeability<br />

High Pp<br />

Natural or induced fractures can dominate shale perm; they are<br />

usually necessary for economically interesting flow in shale.<br />

These fractures also impact Vp, Vs, Vp/Vs ratio, attenuation,<br />

and anisotropy.<br />

49<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Fracture Permeability<br />

Low Pp<br />

Effective stress closes fractures, decreasing permeability and<br />

increasing seismic velocities.<br />

50<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Fractures/<br />

Brittleness<br />

51<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Predicting/measuring Brittleness<br />

Brittleness increases the chances <strong>of</strong> naturally occurring fractures, as<br />

well as success <strong>of</strong> hydr<strong>of</strong>racs. Brittle materials accommodate strain<br />

(deformation) by breaking. In constrast, ductile materials<br />

accommodate strain by “flowing.” Not only are ductile materials less<br />

likely to create permeable fractures, ductile materials will also allow<br />

man-made fractures to close or “heal.”<br />

Important practical issue is how to determine geomechanical<br />

properties from geophysical measurements.<br />

52<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Brittleness: Examples<br />

Low Clay (tight sandstone) High Clay Fraction<br />

Porosity .01<br />

Clay .12<br />

Porosity .05<br />

Clay .31<br />

Porosity .05<br />

Clay .63<br />

In these examples, ductility depends on (1) composition and (2)<br />

confining stress.<br />

Brittle failure has little or no plastic flow before failure.<br />

Jizba, 1991, Ph.D. Dissertation, <strong>Stanford</strong> <strong>University</strong>.<br />

Jizba, 1991, Ph.D.<br />

dissertation, <strong>Stanford</strong><br />

Univ.<br />

53<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Predicting/measuring Brittleness<br />

Brittleness is a complex function <strong>of</strong> lithology, composition, TOC,<br />

effective stress, temperature, diagenesis, thermal maturity,<br />

porosity, …<br />

54<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Fractures Usually Prefer Certain Lith<strong>of</strong>acies<br />

Joint Sheared Joint<br />

Brittle facies <strong>of</strong>ten<br />

accommodate strain by<br />

fracturing, while shales<br />

might accommodate<br />

strain ductily.<br />

J-M. Florez, 2005. Ph.D.<br />

Dissertation, <strong>Stanford</strong> Univ.<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Quantifying Brittleness<br />

Because material failure is important in many technologies, there<br />

are many attempts to define or quantify a Brittleness Index, e.g.<br />

€<br />

B 1 = σ c<br />

σ t<br />

where<br />

€<br />

€<br />

€<br />

€<br />

B 2 = σ c − σ t<br />

σ c + σ t<br />

σ c<br />

σ t<br />

q<br />

Kahraman, 2003, Engineering Geology 65 (2002) 269–283<br />

B 2 = qσ c<br />

= Uniaxial compressive strength<br />

= Tensile € strength<br />

= Amount <strong>of</strong> fines in impact test<br />

56<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Predicting/measuring Brittleness<br />

‘Halliburton’ Brittleness Index<br />

€<br />

( ) /2<br />

B = E Brit +ν Brit<br />

E Brit =100( E static −1)<br />

/ ( 8−1)<br />

( ) / 0.15−0.4<br />

ν Brit =100 ν static −0.4<br />

Rickman et al., 2008, SPE 115258, SPE Annual Meeting<br />

( )<br />

57<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Quantifying Brittleness<br />

€<br />

€<br />

€<br />

Because material failure is important in many technologies, there<br />

are many attempts to define or quantify a Brittleness Index, e.g.<br />

B 1 = σ c<br />

σ t<br />

B 2 = σ c − σ t<br />

σ c + σ t<br />

B 2 = qσ c<br />

€<br />

€<br />

σ c<br />

σ t<br />

q<br />

Kahraman, 2003, Engineering Geology 65 (2002) 269–283<br />

Correlate with penetration rate <strong>of</strong> rotary<br />

drills<br />

Correlates with penetration rate <strong>of</strong> percussive<br />

drills<br />

= Uniaxial compressive strength<br />

= Tensile strength<br />

= Amount <strong>of</strong> fines in impact test<br />

58<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Quantifying Brittleness<br />

In terms <strong>of</strong> overconsolidation ratio<br />

€<br />

OCR = σ V max<br />

σ V<br />

(only valid for layered rocks and max principal stress is vertical)<br />

€<br />

€<br />

€<br />

Brittleness:<br />

( ) OC<br />

( ) NC<br />

B = σ c<br />

σ c<br />

σ V Vertical stress<br />

σ V max Vertical stress at max burial<br />

= OCR b<br />

Nygard et al., Marine and Petroleum Geology 23 (2006) 201–212<br />

59<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Brittleness: Composition<br />

Composition: There is anecdotal evidence that (1) silica<br />

(siltiness) and (2) calcite content increase brittleness. One<br />

index that is sometimes quoted:<br />

€<br />

B ( % ) =<br />

An intuitive extension to calcite:<br />

€<br />

B ( % ) =<br />

Q<br />

Q + Carbon + Clay<br />

Q + Calcite<br />

Q + Calcite + Carbon + Clay<br />

60<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Quartz and Calcite are more brittle than clay and TOC<br />

Dolomite (dolostone, dolomicrite) more brittle than calcite<br />

61<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Brittleness: Composition<br />

⎛⎛<br />

⎜⎜<br />

⎝⎝<br />

Issue: How to measure brittleness from logs? Calcite and quartz each have<br />

distinctly different Vp/Vs than shale. However when added, they might cancel<br />

changes in Vp/Vs.<br />

VP VS ⎞⎞<br />

⎟⎟ ≤<br />

⎠⎠ sand<br />

V ⎛⎛ ⎞⎞ P<br />

⎜⎜ ⎟⎟ ≤<br />

⎝⎝ VS ⎠⎠ shale<br />

V ⎛⎛ ⎞⎞ P<br />

⎜⎜ ⎟⎟<br />

⎝⎝ VS ⎠⎠ limestone<br />

Increasing<br />

stiffness<br />

Increasing Vp/Vs<br />

Increasing gas<br />

Increasing<br />

porosity<br />

62<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Poisson’s ratio Young’s Modulus (GPa)<br />

quartz .07 94<br />

calcite .31 79<br />

dolomite .23 121<br />

clay .35 36<br />

rubber ~0.48 .01<br />

copper 0.34 117<br />

steel 0.30 210<br />

glass 0.2-0.3 50-90<br />

cork ~0 0.03<br />

63<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


€<br />

Beware <strong>of</strong> Poor Legacy Assumptions<br />

€<br />

σ v<br />

σ h<br />

• Sediments deform plastically<br />

during burial – not elastically<br />

• Therefore the horizontal stress is<br />

NOT the elastic expression:<br />

σ h ≠σ v<br />

σ h ≠σ v<br />

⎡⎡ ν ⎤⎤<br />

⎣⎣ ⎢⎢ 1−ν ⎦⎦ ⎥⎥<br />

⎡⎡ C ⎤⎤ 13<br />

⎢⎢<br />

⎣⎣ C ⎥⎥<br />

⎦⎦ 33<br />

64<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Modeling<br />

65<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


€<br />

€<br />

Modeling Composition<br />

Density:<br />

ρ = ∑ xiρ i<br />

ρ = ( 1− k −φ)ρ<br />

mineral + kρker +φρ fluid<br />

k = ρb −( 1− φ)<br />

ρmineral − φρfluid ρk − ρmineral k ≈ ρ b − ρ mineral<br />

ρ k − ρ mineral<br />

; when φ ≤ .03<br />

66<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


€<br />

Modeling Composition<br />

Elastic Modulus: Voigt-Reuss-Hill average<br />

⎡⎡<br />

K = 0.5 ∑ xiK i + ∑ xi /Ki ⎣⎣ ⎢⎢<br />

⎡⎡<br />

µ = 0.5 ∑ xiµ i + ∑ xi /µ i<br />

⎣⎣ ⎢⎢<br />

( ) −1<br />

( ) −1<br />

⎤⎤<br />

⎦⎦ ⎥⎥<br />

⎤⎤<br />

⎦⎦ ⎥⎥<br />

67<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Mineral Properties<br />

density Bulk<br />

modulus<br />

Shear<br />

modulus<br />

Young’s<br />

modulus<br />

Poisson’s<br />

ratio<br />

quartz 2.65 37 44 94 .074<br />

feldspar 2.62 37 15 40 .324<br />

calcite 2.71 70 30 79 .313<br />

dolomite 2.87 75 49 120 .232<br />

siderite 3.96 124 51 135 .32<br />

clay 2.55 25 9 36 .279<br />

pyrite 4.93 147 133 306 .15<br />

kerogen 1.30 6 5 12 .19<br />

A Comparative Study <strong>of</strong> the Mississippian Barnett <strong>Shale</strong>, Fort Worth Basin, and Devonian Marcellus <strong>Shale</strong>, Appalachian<br />

Basin, Bruner and Samosna, DOE/NETL-2011/1478.<br />

68<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Examples<br />

Quartz Calcite Dolomite Clay Kerogen Pyrite E Poisson’s<br />

Ratio<br />

40 10 5 30 10 5 62 0.19 2.65<br />

20 10 5 50 10 5 48 0.22 2.66<br />

10 10 5 60 10 5 49 0.24 2.66<br />

10 20 5 45 15 5 50 0.24 2.60<br />

10 30 5 40 10 5 56 0.25 2.67<br />

10 20 5 50 10 5 52 0.24 2.67<br />

10 30 5 40 10 5 56 0.25 2.67<br />

10 40 5 30 10 5 59 0.25 2.67<br />

10 40 5 25 15 5 56 0.25 2.60<br />

Density<br />

69<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Example computed trends<br />

70<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Example computed trends<br />

71<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Density vs. Kerogen<br />

Lab data (Vernik and Landis, 1996), plotted as bulk density vs. kerogen volume, from samples <strong>of</strong> Bakken shale.<br />

The four sets <strong>of</strong> lines are labeled with the assumed kerogen density, ranging from 1.0-1.6 g/cm 3 . The four black<br />

lines assume the measured porosity ~0.01, and oil in the pores. The dashed red lines assume the porosity is zero.<br />

The dashed blue lines assume that the porosity is saturated with gas.<br />

72<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Bandyopadhyay, 2009, Ph.D. dissertation, <strong>Stanford</strong><br />

Vernik and Liu, 1997, GEOPHYSICS, 62, 521–532.<br />

74<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Bandyopadhyay, 2009, Ph.D. dissertation, <strong>Stanford</strong><br />

Vernik and Liu, 1997, GEOPHYSICS, 62, 521–532.<br />

75<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Bedding-Plane Microcracks<br />

Adding horizontal cracks tends to decrease velocities – depends on fluid.<br />

Wet cracks<br />

Gas-filled cracks<br />

76<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Bedding-Plane Microcracks<br />

Adding horizontal cracks modifies Vp/Vs – depends on fluid.<br />

Wet cracks<br />

Gas-filled cracks<br />

77<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Bedding-Plane Microcracks<br />

Cracks modify intrinsic anisotropy.<br />

Wet cracks<br />

Gas-filled cracks<br />

78<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Bedding-Plane Microcracks<br />

Velocities vs. angle <strong>of</strong> incidence.<br />

Wet cracks<br />

Gas-filled cracks<br />

79<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Vertical Fracture Modeling<br />

Observed:<br />

Strong travel time anisotropy,<br />

not much amplitude variation<br />

Reduced travel time anisotropy,<br />

subtle amplitude variation<br />

small travel time anisotropy,<br />

more amplitude variation,<br />

Early troughs are stretched<br />

Azimuth From Fracture Normal<br />

80<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Vertical Fracture Modeling<br />

Interpretation<br />

Heavily<br />

fractured,<br />

brine<br />

Rotated<br />

minimal<br />

fractures<br />

No<br />

fractures<br />

Heavily<br />

fractured,<br />

brine<br />

Brine<br />

Gas<br />

Brine<br />

Fracture<br />

strike<br />

Brine<br />

Fracture<br />

strike<br />

Fracture<br />

strike<br />

Azimuth From Fracture Normal<br />

81<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Vertical Fracture Modeling -- Amplitude<br />

Amplitude<br />

Brine<br />

Gas<br />

Brine<br />

Fracture<br />

strike<br />

Brine<br />

Fracture<br />

strike<br />

Fracture<br />

strike<br />

Azimuth From Fracture Normal<br />

S08bSM07bM-0bBR0E-09b.tif<br />

82<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Vertical Fracture Modeling -- NMO<br />

Travel time<br />

Brine<br />

Gas<br />

Brine<br />

Fracture<br />

strike<br />

Brine<br />

Fracture<br />

strike<br />

Fracture<br />

strike<br />

Azimuth From Fracture Normal<br />

S08bSM07bM-0bBR0E-09b.tif<br />

83<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Summary<br />

• <strong>Shale</strong> is defined by particle size.<br />

• <strong>Shale</strong> can have a very large range <strong>of</strong> compositions.<br />

• <strong>Shale</strong> can have a large range <strong>of</strong> P- and S-wave velocities<br />

- Composition<br />

- Porosity<br />

- Effective stress<br />

- Compaction<br />

• <strong>Shale</strong> Vp/Vs depends on composition, especially relative<br />

amounts <strong>of</strong> clay, silt, organics, and carbonate<br />

• <strong>Shale</strong> can have a large range <strong>of</strong> anisotropies<br />

- Small if bioturbated<br />

- Large if a pronounced fabric<br />

- Silt and cementation can reduce anisotropy<br />

• In kerogen-rich shales, properties depend on composition, TOC,<br />

and maturity, and fracatures.<br />

84<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Issues<br />

• <strong>Shale</strong> lab data are sparse, compared with sandstone and<br />

carbonate.<br />

• Logs are also more common in reservoirs than shales.<br />

• Other than models like s<strong>of</strong>t-sediment and Raymer, we don’t<br />

have any comprehensive shale models.<br />

• <strong>Shale</strong> anisotropy depends on many factors and is difficult to<br />

predict.<br />

• Organic shales (oil shale and gas shale) can have a range <strong>of</strong><br />

properties, depending on composition, TOC, maturity.<br />

• For gas and oil shales, it is not clear what the geophysical<br />

questions are:<br />

- TOC?<br />

- Maturity?<br />

- Geomechanical?<br />

85<br />

Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory


Gary Mavko – <strong>Stanford</strong> <strong>Rock</strong> <strong>Physics</strong> Laboratory

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!