Fracture Characterization

One of the primary motivations for studying fractures is that they commonly 
constitute conduits for fluid flow, such as this rooster-tail of water squirting 
out of a fracture in basement rocks of the Teton Mountains, Wyoming, 
presumably due to hydraulic connection with higher water level in a nearby 
stream.  Other fluids of interest are oil and gas in fractured reservoirs.

Studying fractures in the Cambrian Flathead Sandstone atop the Teton Mountains, Wyoming, with the help of PhD candidate Leonel Gomez.  Andy Gale, peaking around in back, had just returned from the South Teton summit at the time of this photo.  We have studied fractures in numerous stratigraphic units of the Rocky Mountains.  This summer, Masters candidate Tim Gibbons will begin research in Montana.

One problem for studying fractures in subsurface hydrocarbon reservoirs is that boreholes typically intersect few fractures.  This limitation can be overcome by using SEM-CL to study microfractures, such as the ones above in the Pennsylvanian Weber Sandstone from Rangely field, Colorado.  Microfractures are more abundant by orders of magnitude compared with macroscopic fractures.  Macrofractures are more important to fluid flow, but microfractures are more amenable to subsurface sampling.
Photomicrograph courtesy of Rob Reed.

The abundance of fractures in many sets follows a power law relative to fracture size.  Consequently, observations of microfractures can be used to predict the size and abundance of associated macrofractures.  The example above, from a subsurface core through the Permo-Pennsylvanian Ozona Sandstone, Texas, illustrates that the prediction was verified in this case.

Another important characteristic of fractures is the pattern of their distribution in space.  Some fractures, such as those in this bedding-plane pavement of Cretaceous Frontier Formation sandstone in Wyoming, have approximately regular spacing.

In other cases, such as this example from the Pennsylvanian Marble Falls Limestone, Texas, fractures are clustered.  Each of the erosional recesses in this bedding-plane pavement represent clusters of numerous closely spaced fractures, whereas intervening areas contain few fractures.  New techniques can quantify both regularly spaced and clustered patterns, and distinguish them from random fracture arrangement.  In the case above, the fracture clusters represent natural fractals across at least four orders of magnitude in length scale.

I have begun to study faults using the same techniques for analysis of spatial patterns.  Faults in Miocene sandstones and shales at A-Bomb canyon, Arizona, are well suited to this approach.  For example, the notebook in the photo above sits at the upper end of a relay ramp formed in the bedding-plane pavement of a gravel layer.