OpenPetro AI · Learning Tools

Shale Brittleness from Logs

From sonic, density and mineralogy logs to dynamic elastic moduli and brittleness indices. Compare the mineral and elastic definitions side by side, see where they disagree, and pick candidate frac intervals.

BIRickman = ½[(E − Emin)/(Emax − Emin) + (νmax − ν)/(νmax − νmin)]
BIWang–Gale = (Qz + Dol)/(Qz + Dol + Cal + Clay + TOC)

Sonic and density

Δtc µs/ft
Δts µs/ft
ρb g/cm³
Static/dynamic E ×

Mineralogy (weight %, ECS or XRD)

Dynamic elastic properties

Rickman normalization Mineral BI in composite Weight mineral : elastic Brittle ≥ductile <

Log view

A synthetic shale section to start, with limestone and clay-rich intervals. Load your own LAS 2.0 or CSV; it stays in your browser. Click a depth to load it into the calculator.

Sample data (synthetic): 07_brittleness_shale_full.las (DTS and mineralogy) 07_brittleness_no_shear_metric.las (no DTS: tick “Estimate DTS”; metric)

Lab brittleness from stress–strain tests

Not from logs, but the reference that log indices should be calibrated against. Enter triaxial or strength results for one plug.

Energy method (post-peak)
DBI = M / (M − E)
E (GPa)
M (GPa) < 0
–
Peak–residual strength
PBI = (τp − τr) / τp
τp (MPa)
τr (MPa)
–
Compressive vs tensile strength
B₁ = σc/σt · B₂ = (σc − σt)/(σc + σt)
σc (MPa)
σt (MPa)
–

Rickman crossplot: Young's modulus vs Poisson's ratio

Brittle rock plots toward high E and low ν (upper left). Lines are constant Rickman brittleness for the current normalization; dots are log samples coloured by mineral brittleness.

Mineral ternary: QFP – carbonate – clay

Quartz + feldspar + pyrite, calcite + dolomite, and clay, normalized to 100%. Siliceous and carbonate-rich mudstones tend to be brittle; argillaceous ones ductile.

Do the methods agree?

Mineral brittleness against elastic brittleness for every log sample. Scatter here is the inconsistency reported in the literature: composition and stiffness are related but not the same thing.

Methods and formulas

Values at the current point.

Method and assumptions