Geometrical characterization of fracture networks: Core and borehole fracture surveys, density calculation, and spatial structure analyses

In hydrogeology, engineering and mining geology, fractures are one of the most common and most important geological structures, and have been the subject of considerable research. Three approaches have been taken in pursuit of fracture system study: mathematical and elastic theory, experimental rock deformation, and fracture characterization through field studies. The mathematical treatment of fracture theory (initiation, propagation and interaction of fractures) has proven valuable in increasing our understanding of this complex process. Experimental deformation of rocks and rock analogs is an important link between theory and natural fractures. Because of the complexity and limited accessibility of fracture mapping in nature, the last approach, field fracture mapping and fracture network characterization, has generally proven the most difficult. However, the characterization of natural fracture networks is more important because it serves as the basic evidence for the first two approaches. This research focuses on the last approach, and is initiated to: (1) develop fracture survey techniques by using readily available software to produce complete images of fractures from both cores and downhole video images; (2) study the relationship between fracture frequency and fracture density of the natural fracture networks obtained with rigorous fracture mapping techniques in the field. I will establish a method to calculate the 2-D and 3-D densities of a fracture network from a 1-D fracture survey (e.g., scanlines and borehole data); (3) detect the spatial structures of natural fracture networks through fractal and geostatistical methods.; Accomplishment of these research objectives provides the critical information for engineering geologists in mining design and especially in the characterization of subsurface flow and transportation systems in fractured rocks. It also provides geologists useful evidence to study regional structural phenomena (e.g., faulting and tectonic deformation history) associated with the formation of fractures, and to test theoretical models. Only with the detailed field data can the models be evaluated. To date, many numerical models have been developed to simulate fracture networks. Clearly, more field data and detailed study of the characteristics of the natural fracture networks are necessary to test and improve these models.; This doctoral dissertation consists of three parts. The first part develops two fracture image processing systems by using readily available software to produce complete images and analyses of the fractures from both cores and downhole video tapes. The second part investigates the statistical relationship between fracture frequency and fracture density {dollar}Dsb1.{dollar} Assuming that fracture length distribution follows an exponential or power law, the analytical solutions for converting dimensional density {dollar}(Dsb1){dollar} to non-dimensional densities {dollar}(Dsb2{dollar} and {dollar}Dsb3){dollar} are derived. The third part examines the spatial structures of natural fracture networks through fractal and geostatistical analyses.