--> --> Abstract: Geostatistical Analysis of Fault and Joint Measurements in Austin Chalk, Superconducting Super Collider Site, Texas, by R. E. Mace, H. S. Nance, and S. E. Laubach; #90960 (1995).

Datapages, Inc.Print this page

Abstract: Geostatistical Analysis of Fault and Joint Measurements in Austin Chalk, Superconducting Super Collider Site, Texas

R. E. Mace, H. S. Nance, S. E. Laubach

Faults and joints are conduits for ground-water flow and targets for horizontal drilling in the petroleum industry. Spacing and size distribution are rarely predicted accurately by current structural models or documented adequately by conventional borehole or outcrop samples. Tunnel excavations present opportunities to measure fracture attributes in continuous subsurface exposures. These fracture measurements can be used to improve structural models, guide interpretation of conventional borehole and outcrop data, and geostatistically quantify spatial and spacing characteristics for comparison to outcrop data or for generating distributions of fracture for numerical flow and transport modeling.

Structure maps of over 9 mi of nearly continuous tunnel excavations in Austin Chalk at the Superconducting Super Collider (SSC) site in Ellis County, Texas, provide a unique database of fault and joint populations for geostatistical analysis. Observationally, small faults (<10 ft. throw) occur in clusters or swarms that have as many as 24 faults, fault swarms are as much as 2,000 ft. wide and appear to be on average 1,000 ft. apart, and joints are in swarms spaced 500 to more than 21,000 ft. apart. Mean fault and joint spacing range between 25 and 50 ft. for the different tunnel segments. Semi-variograms show varying degrees of spatial correlation. The best correlations and spatial relationships are obtained using semi-variograms generated with respect to fault frequency. These var ograms have structured sills that correlate directly to highs and lows in fracture frequency observed in the tunnel. Semi-variograms generated with respect to fracture spacing and number also have structured sills, but tend to not show any near-field correlation. The distribution of fault spacing can be described with a negative exponential, which suggests a random distribution. However, there is clearly some structure and clustering in the spacing data as shown by running average and variograms, which implies that a number of different methods should be utilized to characterize fracture spacing.

AAPG Search and Discovery Article #90960©1995 AAPG Southwest Section Meeting, Dallas, Texas