Within the study area, a 2000 ft (600m) wide zone of faults defines a north-trending graben that disrupts the crystalline basement and overlying homoclinal Paleozoic strata, which dip northwesterly about 3 degrees (Figure, 62K) . Northward, the graben widens, changes to a NNE trend and the well-defined, bounding faults exhibit larger dip-slip displacements. Another N-to NE-trending graben occurs about two miles (3.2 km) due south and appears to be connected to the northern graben by a N-trending, right-lateral, oblique-slip fault. The two grabens are positioned on opposite sides of the N-striking, oblique-slip fault, forming a structural system that is compatible with an inferred NW-SE regional extension during Pennsylvanian time. The major faults that disrupt the Paleozoic strata in the Llano region developed during early to middle Pennsylvanian time and are inferred to be controlled primarily by re-activation of pre-existing structural weaknesses in the underlying Precambrian basement.

Within the study area, the graben is bound by north-trending faults (Geologic Map, 135K). In the northern and southern regions of the study area, the bounding faults are mapped with little ambiguity. Within the central region, however, only the western bounding fault is readily mapped, but high-resolution gravity surveys helped constrain some fault locations where exposures are lacking.

Within the graben, faults trending NNE and NE most commonly terminate against or progressively change strike to merge with either a N-trending fault bounding the graben or a N-trending internal fault (Geologic Map, 135K). This system of faults partitions the graben into a number of completely or partially fault-bound blocks.

Faults of small displacement and limited lateral extent are common in regions between larger displacement faults. Typically, a conjugate system of two sets of small faults is observed in an outcrop. Most commonly, one set strikes N to NNE and the other set strikes NE to ENE. Fault displacements range from 1 mm to 10 cm with associated gouge zones of 1 mm to several centimeters thick. The number of small faults per unit area varies with location and appears to be greatest in the central and northern regions. These small faults develop as a consequence of mechanical interaction between neighboring larger faults, as discussed below.

Fault displacement is predominantly oblique-slip. A strike-slip component of displacement is inferred from geometries of fault systems and associated secondary structures. Fault plane striations are observed infrequently. Many of the mapped NNE and NE faults are interpreted to form as a result of interaction and displacement transfer between overlapping northerly-trending faults with a component of right-lateral, strike-slip displacement. Details of fault geometries and outcrop-scale structures further support the existence of a component of right-lateral strike-slip. For example, at several places along the western bounding fault, the fault trace consists of a series of north-trending, en echelon, overlapping, left-stepping, fault segments connected by and bounding a system of NE striking, secondary faults. This fault geometry reflects a right-lateral sense of strike-slip displacement. The structure of the shear zone associated with a fault, as described later, also demonstrates the common occurrence of a strike-slip displacement component.

Generally, only estimates of actual fault displacements are possible because of the lack of suitable markers offset by the faults. The structural contour map on the top of the granite (Figure, 69K). shows the vertical offset of the basement across faults and provides an estimate of the dip-slip component, particularly where the dip of the basement surface is small. The structural contour map together with fault geometries and associated secondary structures suggest that relative magnitudes of dip-slip and strike-slip displacement vary from fault to fault. The major faults in the northern and southern parts of the study area appear to be dip-slip dominant. In the central region, the relative magnitude of strike-slip and dip-slip components is unknown, but the greater intensity of distributed small faults suggests some faults may either have comparable dip-slip and strike-slip components or be strike-slip dominant, as are some N-trending faults.

Vertical offset across the faults in the study area is less than the stratigraphic thickness of the Hickory, except along the southern parts of the faults bounding the graben. Consequently, along all faults, with the exception of the two noted fault segments, Hickory Sandstone is juxtaposed against Hickory Sandstone along a portion of a vertical section through the fault.

Lack of extensive outcrops allows only a generalized characterization of the structure of faults that juxtapose Hickory Sandstone against Hickory Sandstone. Outcrops of faults only show fault structure associated with sandstone intervals. There are no exposures of shear zones involving clay-rich interbeds.

Shear deformation associated with mappable faults is localized primarily in a well-defined zone. Within this shear zone, shearing-induced comminution of the sandstone typically is not uniform but is localized along a number of closely spaced small faults. Usually two sets of cross-cutting, small faults exist; one set is parallel to subparallel to the trend of the shear zone, whereas the other set is oriented at an angle of 30o to 60o to the trend of the zone. The occurrence of a conjugate system of small faults in a horizontal section reflects the existence of a strike-slip component of fault displacement. The width of the shear zone varies along an individual fault, as well as from fault to fault and ranges from 0.3 m (1 ft) to several meters. The average width of the shear zone of a fault appears to increase with increasing fault displacement, but this relationship could not be quantified.

Enhanced development of small faults is observed adjacent to the shear zone of a large fault as compared to a location more distant from the fault. The orientation and density of small faults in this zone reflects the local strain field, which can differ from areas further from the fault. The nature of this small-fault system can vary along a fault. For example, large faults that evolve by linkage of a set of en echelon, overlapping segments locally have greater densities of small faults in the displacement transfer zones between overlapping segments as compared to a position adjacent to a nonoverlapping segment. This case exemplifies the western bounding fault of the graben.

Shear deformation produces a fault gouge in the sandstone characterized by intense comminution of the quartz grains. The nature of the fault gouge in the Hickory Sandstone is similar to that described by other workers for faults in porous quartzose sandstones. The fault gouge has a marked decrease in grain size, sorting, porosity, and permeability as compared to undeformed Hickory Sandstone. Locally, in exposed faults, secondary quartz cementation further reduces the porosity and permeability of the gouge. Significant quartz cementation produces a gouge with the character of quartzite.

The net thickness of fault gouge associated with the shear zone of a large fault depends upon the width of the zone and the characteristics of the small faults that comprise the zone. Observed shear zones have small fault spacings ranging from 1 to 3 cm, with the individual faults having gouge zones from 0.1-1 cm thick. As noted earlier, shear zones have widths varying from 0.3 m to several meters.

The hydraulic conductance of a fault should reflect the net thickness and permeability of fault gouge associated with both the shear zone of the fault and the neighboring zone of associated small faults. Consequently, one would like to know the net thickness of gouge and its spatial variation along a fault. Unfortunately, at present, only crude estimates of net gouge thickness are available for only a few faults in the study area. These thickness estimates range from 0.5 ft (0.15 m) to about 2 feet (0.6 m).

For purposes of assessing potential hydraulic compartmentalization by faults, five particular regions within the fault-partitioned graben are highlighted and are designated by letters A through E (Figure, 33K). Each designated region contains two or more water wells and has bounding faults with estimated fault displacements of 50 ft (15 m) or greater. The estimate of fault displacement is based primarily on the vertical offset of the Precambrian basement across faults, except for faults along the southern portion of Region D. In the latter case, the faults have prominent shear zones (2 m wide) and appear to have a strike-slip dominant displacement. In addition, it is uncertain whether Regions B and E are bound completely by faults on their eastern sides. Areas outside the graben to the east and west are designated as the Eastern Region and Western Region, respectively. As noted earlier, the primary objective of the groundwater study was to assess the nature of the hydraulic communication between these regions and establish to what extent the faults act as hydraulic impedance structures to lateral flow of groundwater and subdivide the groundwater system into hydraulic compartments with differing short- and long-term hydraulic response characteristics.