A method to determine the spacing and transmissivity statistics of ``conductive'' fractures in which the criterion disgtinguishing between conductive and nonconductive fractures is a sliding scale
Abstract
The design of cleanup applications for fractured rock strongly depends on the use of appropriate models and field measurements. Implicit in the use of discrete fracture models is that a subset of fractures have been distinguished from the remaining fractures and labelled as ``conductive''. Where injection tests have been used to measure permeability along boreholes, the dividing line between these two sets of fractures has traditionally been associated with investigators' ability to measure flow. For example, those sections of borehole into which the rate of injection is below the lower limit of measurement are often associated with the set of ``nonconductive'' fractures, while those sections into which flow can be measured have been assumed to contain at least one of the conductive set. Snow (1970) used this criteria along with the statistics of a set of transmissivity measurements to derive the mean and variance of the transmissvity of the conductive set. More recently, investigators have used this criteria along with a set of transmissivity magnitudes to derive maximum likelihood estimates of the parameters of a conductive fracture transmissivity set having an assumed gammadistribution. In this study, provided the distribution of fracture transmissivities is strongly positively skewed, this method of distinguishment is shown to be valid, even in the case where the lower limit of flow rate detection is artificially raised above that associated with equipment. This (upward) sliding detection limit is shown to be useful in the creation of a diminishing (in population) subset of fractures with increasing mean transmissivity and decreasing variance. Using the assumption that the lognormal distribution is stable during addition of small numbers of relatively low variance random variables, the lognormal distribution is shown to be a better model of fracture transmissivity than is the gamma distribution.
 Publication:

AGU Spring Meeting Abstracts
 Pub Date:
 May 2004
 Bibcode:
 2004AGUSM.H23D..01W
 Keywords:

 1055 Organic geochemistry;
 1803 Anthropogenic effects;
 1831 Groundwater quality;
 1832 Groundwater transport;
 1894 Instruments and techniques