CHALLENGES IN PREDICTING FRACTURE NETWORK EVOLUTION AND FLUID FLOW PATHWAYS IN LAYERED LIMESTONE FOLD SYSTEMS

The evolution of fracture networks during fault-propagation fold development provide an important control on the flow of subsurface fluids. We use detailed, outcrop-based documentation of fractures and later fill events to better inform our understanding of coupled fold-fracture systems in bedded carbonates. Our research reveals a chronology of fracture set formation within the NW-trending Stillwell anticline, a fault propagation fold to the northeast of Big Bend National Park in west Texas. Our measurements suggest 4 steeply-dipping fracture sets striking N (F1), NE (F2), ENE (F3), and NW (F4). Our observations and interpretations reveal that set F1 is dilatational and precedes fold formation, F2 and F3 are mixed mode conjugate sets formed early in fold formation, and dilatational set F4 is likely related to outer-arc bed flexure during anticline formation. The orientation of these sets are consistent with previous studies of fold-related fracture formation.

Outcrop-scale fracture maps reveal significant variability in fracture intensity for each fracture set not only when compared to different locations but also when compared to intensities measured at a given outcrop. In addition, very few plan-view exposures display all 4 sets. Based on these data, we hypothesize that fracture sets are distributed heterogeneously, both horizontally and vertically, within the well-bedded stratigraphy, with variations in inter-bed cohesion playing a role in the evolution of the overall fracture system.

Fracture fill characteristics support a strong correlation by sample location, with no correlation by fracture set. We therefore suggest that fractures at every location were open for some time before sealing by calcite precipitation, and that all fractures at a given location were filled in the same number of events (the number of events varies by location) by locally-sourced fluids with little O- or C-isotopic variation.

We suggest that: 1) fracture orientations and intensities vary significantly in 3D; 2) fluid flow rates and fracture connectivity were maximized during fold formation, when all fracture sets remained open; and 3) flow-inhibiting precipitation was localized and unpredictable. These findings reveal the challenges in applying rigorous fluid flow models to similar fold-related fracture systems.