In the brittle crust, the distribution of natural rock fractures and their failure modes are a function of rock strength and its interactions between overburden pressure, pore-fluid pressure, and tectonic loading. The characterization of variability in rock strength and the associated changes in subsurface strain distribution is especially important for modeling the response of low-permeability rocks to changes in effective stress. This paper documents the effect variations in elastic mechanical properties have on the nature and distribution of fractures in the subsurface. Outcrop and geophysical wireline log evaluation of the Jurassic Carmel Formation and Navajo Sandstone was used to identify mechano-stratigraphic units and model subsurface strain distribution within sedimentary successions and across sedimentary interfaces.

Two finite element models were constructed and populated with elastic moduli derived from geophysical wireline data in order to understand where natural fractures form in rocks with varying layer thickness and elastic properties. Strain distribution results from a 3-layer and a 5-layer model are compared to the natural deformation response visible in outcrop. Model results show that more fractures are expected in high strain regions and fewer fractures in low strain regions. Strain variations are observed in both model scenarios and occur at material interfaces. The simple 3-layer model results in a smoothing of strain variations, while the 5-layer model captures strain variations that more closely match the fracture density observed in outcrop. Results from the 5-layer model suggests an interplay between Young’s modulus and Poisson’s ratio and that high strain regions form in thin (1-m thick) layers with moderate Young modulus (17.2 GPa) and Poisson ratio (0.26) values.

Outcrop observations and modeling results indicate that the potential for subsurface failure and fluid flow would not be restricted to the low fracture strength units but can cut vertically across interfaces of varying mechanical strength. Results from this work indicates that these types of models can be used to identify stratigraphic layers that are more prone to mechanical failure or identify layers that have more natural fractures or are more likely to form induced fractures.

You do not currently have access to this article.