Software

FracSize

FracSize is a Windows 95/NT program to derive fracture size (radius or area) from fracture trace length measurements on tunnel walls, surface exposures, and lineament maps.

FracSize uses forward modeling of the process of fracture trace data collection to determine the fracture size distribution most consistent with measured fracture trace length distributions. This takes into account:

  • Sampling biases due to fracture size and orientation
  • Censoring due to data collection procedures
  • Truncation due to limited fracture exposure on outcrops and tunnel walls

FracSize compares simulated and measured fracture trace length distributions, and adjusts the assumed fracture radius distribution to match the measured fracture trace length distribution.

FracSize

FracSize has been used successfully on data at scales from 100 km lineament maps to 1 m joint maps. FracSize is an important member of the FracSys data analysis system.

FracSize Features

Feature Description
Platform DOS (FracSys Version 2.601)

Windows 95/NT (FracSys Version 1.1)
Approach Simulation of Fracture Traces by Intersection of Stochastic Fracture Realizations with Sampling Structures
Radius Distribution Assumptions Constant, Uniform, Exponential, Lognormal, Normal, Power Law (Pareto)
Orientation Distribution Assumptions Constant, Normal, Fisher, Bingham, Bootstrap
Goodness of Fit Criteria Kolmogorov-Smirnov or Chi-Squeared
Search Algorithms Conjugate Gradient, Simulated Annealing, Interactive Visual Optimization
Statistical Reporting Mean, Std Dev. Skewness, Kurtosis, Panel Percentages
Graphical Reporting Histogram (PDF) and Cumulative Distributions (CDF)
   

Algorithm Summary

The input for the FracSize algorithm consists of

  1. The geometry of the tunnel or surface on which the data was collected, and
  2. The measured trace lengths.
  3. Optionally, the fracture trace length file can also include the fracture orientation and the percentages of length on each of the tunnel walls, if available.

Based on the fracture trace information, FracSize calculates the following fracture trace statistics:

  • Trace length mean, standard deviation, skewness, and kurtosis
  • Percentage of fractures spanning one panel, two panels, etc. up to the number of panels defined on the tunnel
  • Fracture trace intensity (P21) on the mapped panels

The user then specifies the assumed fracture population orientation (Fisher, Bingham, bootstrap, etc) and size distribution (log normal, power law, exponential, etc.).

The following parameters are used to control the FracSize simulation:

  1. The minimum fracture trace length mapped
  2. The search space maximum, minimum, and step size for fracture radius radius mean and standard deviation.
  3. The number of realizations to be combined to calculate the trace length and panel percentage statistics for a given fracture radius assumption
  4. The type of search (single point, grid search, conjugate gradient search, or simulated annealing search).
  5. The search criteria (Kolmogorov-Smirnov or Chi-Squared Statistic)
  6. The maximum number of iterations, parameter temperature, and jumping scale (for a simulated annealing search), or the damping factor (for a conjugate gradient search).

Based on the parameters provided, FracSize generates stochastic fracture realizations, and calculates the trace statistics and panel percentages for intersection with the specified traceplane or tunnel. FracSize generates fractures in each realization up to the calculated value of intensity P21. This ensures that the standard deviation and panel percentage statistics of simulations and field measurements are strictly comparable.

FracSize compares simulated fracture trace distributions against measured fracture trace statistics using either Kolmogorov-Smirnov or Chi-Squared statistics. FracSize can search for the fracture radius distribution with the best goodness of fit measures using either a conjugate gradient or simulated annealing search.

FracMan Program Modules