Natural fracture characterization, Frontier Formation, Wyoming

Abstract

Fractures can increase the permeability and producability of reservoirs by acting as fluid and gas conduits to wells. Networks of fractures are most important in reservoirs where little to no matrix permeability exists such as tight gas sandstones. Two significant variables, fracture length and the abundance of fractures, are not readily measurable from subsurface observations such as those obtained from cores or well logs. Numerical models suggest natural fracture apertures and lengths follow systematic power-law (Marrett, 1996; Olson, 2007) and negative exponential distributions (Olson, 2004); fracture trace lengths are interrogated. This study tests those propositions through study of fractures in outcrop. Outcrops of the Cretaceous Frontier Formation at Oyster Ridge in southwest Wyoming and Oil Mountain near Casper, in central Wyoming provide evidence of reservoir scale fracture networks in sandstones. In the subsurface the Frontier Formation sandstones are reservoirs that produce gas and oil in several Wyoming basins. I mapped fracture patterns at twenty locations at Oyster Ridge and Oil Mountain and measured fracture trace length distributions and abundance (intensity). Fracture cumulative length distribution plots illustrate systematic length distributions. Trace length distributions of every fracture network follow negative exponential distributions regardless of the number of fractures (N = 39 to N = 394) or the size of the outcrop (1.3 to 710 m²). Results show that the fractures follow a negative exponential distribution over a range of lengths of a few centimeters to tens of meters. These trace length distributions are consistent with geomechanical model fracture pattern simulation results by Olson (2004) that suggests negative exponential trace length distribution result from fracture to fracture interaction during fracture formation. Length distributions from my field study are inconsistent with pattern simulation results by Marrett (1996) and Olson (2007) and others that produce power-law length distributions. This inconsistency suggests that the model assumptions of Olson (2004) best account for the patterns I observed. Two dimensional fracture intensity, defined as the total fracture trace length divided by the map area, was measured for each outcrop to determine how structural position affects fracture abundance patterns. Two-dimensional fracture intensity measurements collected at thirteen structural locations around Oil Mountain show higher values of fracture intensity near the fold-axial-trace compared to fold limbs. The difference is as much as 7.4 fractures per meter near fold hinges compared to 0.63 fractures per meter in fold limbs. Outcrops near small faults, with displacement of a few meters, show an increase in fracture intensity from background values around 4.8 fractures per meter to values nearly three times as high (13 fractures per meter) near faults. Values of fracture intensity that are more elevated near small tear faults imply that faulting has a greater influence on fracture intensity than folding.