Introduction

Characterization of brittle fractures in outcrops or roadcuts is an important aspect of hydrogeologic assessments (Caine and Tomusiak, 2003; Manda and Mabee, 2010). Modern geologic maps are useful for this purpose where they contain sufficient brittle fracture data (Walsh and Clark, 1999, 2000).

In rock blasting, explosives create dense, random, radial oriented cracks in the immediate vicinity of boreholes (Kutter and Fairhurst, 1971; McKown, 1986; Fourney, 1993). The fracturing of the rock mass during blasting is influenced by preexisting discontinuities and geologic structure and such features might be extended during blasting but new geologic features are not created in the process (Kutter and Fairhurst, 1971).

Fractures in bedrock play an important role in fractured rock aquifers in crystalline metamorphic rocks because they increase the permeability and porosity of the rock (Caine and Tomusiak, 2003). Fractures in the bedrock near newly blasted roadcuts on Route 111 in Windham, New Hampshire, acted as fluid transport pathways for nitrate contamination from many sources including explosives used in construction of the highway bypass to drinking water wells (Degnan and others, 2016). The lithology and additional observations of mineralogy mapped during this effort also have many implications for water rock reactions affecting groundwater quality (Flanagan and others, 2018). In addition, when some rock types are reduced to rockfill through blasting and are used as roadway construction fill in wet areas, they can contribute to iron fouling of waterways (Lombard and others, 2020).

This report presents the results from a detailed survey of the fractures observed in the roadway rock cuts after blasting but prior to completion of the roadway in June 2014; data are presented in Walsh (2024). Orientation data were measured with an analog pocket transit and entered into the Fieldmove application running on an Apple iPad mini version 2. Coordinates were recorded with a Dual Electronics XPS150 global position system (GPS) receiver linked to the iPad via Bluetooth. Coordinate accuracy is ±5–10 meters (m).

Rock characterization followed traditional geologic mapping practices and findings presented in Walsh and Clark (1999). Fracture characterization followed the approach by Walsh and Clark (2000). The strike and dip of all fractures greater than 20 centimeters (cm) in length (Barton and others, 1993) between the observer's knees and chest were measured on traverses across two roadcuts. Data collected in this manner include all features and have been referred to as “objective” data (Spencer and Kozak, 1976) because there is no sampling bias on the part of the data collector. This most common method is also referred to as the “single scanline method” (Manda and Mabee, 2010). Roadcuts with two different trends were studied to limit directional sampling bias (Terzaghi, 1965; Park and West, 2002), but a statistical correction such as those used in rock mechanic and slope stability studies was not applied (Tang and others, 2018). Locality 1 trends at 80° and locality 2 trends at 30°. Data in Walsh (2024) include coordinate information, unlike traditional scanlines, but further processing or attempts to correct for directional bias are beyond the scope of this study.

Mineral identification was conducted by transmitted light microscopy on standard petrographic thin sections. Semiquantitative mineral identification of one sample of a vein was conducted on a polished slab by energy dispersive X-ray spectroscopy (EDS) on a Hitachi SU5000 scanning electron microscope (SEM) equipped with an Oxford Ultim Max EDS detector and AZtec software. Settings used for SEM analysis were 15 kilovolts (kV) accelerating voltage, 50 intensity (unit specific to Hitachi SEM software that corresponds to ~2.41 nanoampere beam current at 15 kV), and 11.2 millimeters (mm) working distance. Using the AZtec software’s built-in mapping and AutoPhaseMap tools, EDS spectra were collected from each pixel of a 512 x 384 pixel-area at a magnification of 50x (~4.98 microns/pixel). The AutoPhaseMap tool grouped each pixel by similar EDS spectra and assigned a color to each group to create a false color image. The AutoPhaseMap tool was run with a boundary tolerance of 10.0 and a grouping level of 2.0. A polished slab (cut oblique to the vein) was also analyzed using the Feature Analysis tool in AZtec, which collected grayscale backscattered electron image mosaics of large areas, identified features with grayscale values within a user-specified range, and collected EDS spectra from those features specifically. Features were then binned under appropriate mineral names and color coded by comparing their EDS spectra against user-defined criteria (for example, the presence of >5 percent arsenic (As), iron (Fe), and sulfur (S) in the EDS spectrum resulted in binning under the name “arsenopyrite” and coloring the feature orange), resulting in a large-area backscattered electron image with minerals of interest highlighted in color. Each identified feature was confirmed manually to ensure accuracy of mineral identification by the software.

Results

Fracture characterization at two new roadcuts along the New Hampshire Route 111 bypass in Windham, New Hampshire, shows that the outcrops consist of the metasedimentary Berwick Formation intruded by multiple phases of foliated to nonfoliated granite to granitic pegmatite assigned to the New Hampshire Plutonic Suite (Walsh and Clark, 1999, 2000). The Silurian age of the Berwick Formation is constrained by detrital zircon geochronology (Wintsch and others, 2007). The age of the New Hampshire Plutonic Suite is considered Devonian (Zartman and Naylor, 1984; Lyons and others, 1997; Walsh and others, 2013). Post-metamorphic diabase and lamprophyre dikes of the White Mountain Plutonic-Volcanic Suite occur in nearby outcrops (Lyons and others, 1997; Walsh and Clark, 1999), but were not observed at the two localities in this report. Several of the black dikes were observed nearby on the new roadcuts on Interstate 93 (I–93) 0.8 kilometers to the southeast (Walsh, 2024). The dikes were previously exposed at an abandoned quarry (Walsh and Clark, 1999). The mafic dikes are considered Mesozoic age (Foland and Faul, 1977).

On June 10, 2014, bedrock fracture data were collected at two roadcuts in the vicinity of Indian Rock, located west of I–93 and north of Cobbetts Pond (localities 1 and 2, fig. 1). The collection of fracture data followed the objective method employed by Walsh and Clark (2000), including the measurement of fractures over a distance of approximately 225 and 85 m at localities 1 and 2, respectively (fig. 1). Fracture data are available in Walsh (2024).

The Berwick Formation consists of medium-gray biotite-plagioclase-quartz granofels, biotite schist, and lesser calc-silicate rock. At the two localities, the Berwick Formation consists mostly of granofels. At the time of this study, the Berwick Formation was locally sulfidic, but the fresh, unweathered roadcuts exhibited only locally rusty weathering (10 of 72 fractures at locality 1 are rusty weathering). Abundant masses of light-gray to white, medium- to coarse-grained, muscovite-biotite-quartz-albite-microcline granite and granitic pegmatite occur as foliated irregular masses and crosscutting tabular, weakly foliated and folded dikes. For the purposes of this report, the igneous rocks are simply referred to as “granites.”

The Berwick Formation is polydeformed and locally contains the following four planar fabrics: S₀ = bedding, S₁ = oldest foliation, S₂ = second oldest foliation, and S₃ = youngest foliation (Walsh and Clark, 1999). The granites post-date S₀ and S₁ (Walsh and Clark, 1999). Because the Berwick Formation is largely a granofels, foliation is not readily apparent in many places, especially at the hand-sample scale. Layering of the granofels is apparent at the scale of the roadcuts, however, and the layering is largely tectonic. Two high-strain, locally subparallel, ductile foliations (S₁ and S₂) are present in the Berwick Formation (Walsh and Clark, 1999), but for the purposes of this fracture study, the two foliations were not considered separately. The most conspicuous planar fabric is a foliation in the area that is steeply northwest dipping to subvertical and northeast-southwest striking (fig. 2). Regionally, the strike of the foliation is consistently to the northeast-southwest, but the dip is locally variable to both the southeast and northwest (Walsh and Clark, 1999). The foliation exhibits limited parting with 9 of 107 fractures (approximately 8 percent) observed parallel to the foliation at the two localities. The limited degree of parting is in agreement with observations from other exposures of the Berwick Formation within the garnet zone elsewhere in the Windham 7.5-minute quadrangle (Walsh and Clark, 2000). Parting fractures along the northeast-striking foliation are more apparent at locality 2 where the data produce a subordinate trend (28°±4°, fig. 3) in the steeply dipping (dip >59°) fracture data. The youngest ductile fabric is a subhorizontal spaced cleavage (S₃) with associated symmetrical, low amplitude (5–10 cm) and long wavelength (50–100 cm) folds. This fabric is widespread throughout the area (Walsh and Clark, 1999), but is only visible on vertical exposures due to the shallow dips. This “accordion-like” fabric could be related to a weak D₃ vertical-loading event. In the quadrangle, at large quarry and roadcut exposures, the rocks locally exhibit a weak exfoliation parallel to the subhorizontal S₃ cleavage. Much of the exfoliation, particularly in the granitic rocks, is most likely related to glacial rebound or post-Mesozoic unroofing, but appears to have utilized the subhorizontal ductile fabric where present.

The most prominent fracture trend is subvertical to steeply northeast-dipping and northwest-southeast striking. Contoured poles to all fractures at each locality show a principal strike and dip of 290°, 78° (locality 2) and 295°, 80° (locality 1) (figs. 3 and 4A). The peak trend of steeply dipping (dip >59°) fractures is 289°±6° (locality 2) and 295°±12° (locality 1) (figs. 3 and 4A). Three fractures with this trend at locality 1 occur along thin (1–3 mm) black veins in granite (fig. 4B). The veins locally show very minor (<1 mm) displacement of mineral grains as microscale fault surfaces or dilation features and were observed in the granite. The veins primarily consist of quartz, tourmaline, and ankerite identified by EDS under a Hitachi SU5000 SEM (fig. 4C–F). Other minerals in the veins include minor amounts of sulfides (identified as arsenopyrite, galena, and rare sphalerite) and apatite. Sulfides were observed in the analyzed vein but not in the host granite. The 289°–295° fracture trend closely agrees with the most prominent fracture trend recognized within the Windham quadrangle, and at the nearby fracture station 6007 (fig. 1) on Indian Rock Road (Walsh and Clark, 2000). Independently, northwest-striking lineaments in the Windham quadrangle show a higher probability of having a high-yielding well nearby indicating that these fractures likely have a higher transmissivity in the fractured bedrock aquifer (Moore and others, 2002).

Gently south- to southeast-dipping and east- to northeast-striking fractures occur as joints and as parting fractures along a weak cleavage (called S₃ by Walsh and Clark, 1999, 2000). These gently dipping fractures show a principal strike and dip of 125°, 15° (locality 2) and 17°, 31° (locality 1) (figs. 3 and 4A).

Water-bearing fractures were observed at locality 1 along joints and gently dipping contacts between the Berwick Formation and the granite to granitic pegmatite (figs. 5 and 6). Six of 72 fractures are water-bearing and 4 of the 6 dip gently to the southeast (fig. 4A). A throughgoing, northeast-striking, southeast-dipping water-bearing fracture zone is exposed at locality 1 (fig. 5).

For the purposes of this study, fracture data (Walsh, 2024) were divided into generalized rock type including either the Berwick Formation or granite. Fracture data separated by generalized rock type (fig. 7) shows a similar distribution for steeply dipping northwest-striking trends, but with much fewer observed steeply north-dipping fractures in the granitic rocks. Both rock types show a cluster of gently south-dipping fractures. The granites show far fewer steeply dipping northeast-striking fractures, which reflects a greater degree of parting along the foliation in the metasedimentary rocks than in the granites. No foliation-parallel fractures were observed in the granites, but some contacts between granites and the Berwick Formation do exhibit parting (fig. 6).

Using the fracture termination classification of Barton and others (1993), data from both localities combined yield 3 percent abutting, 76 percent dead end (or blind), and 21 percent crossing (or throughgoing) fractures (fig. 7).

Six brittle faults were observed at localities 1 and 2, and none were water-bearing. The faults strike northeast and most dip steeply to the northwest (fig. 8). Calculated paleostress tensors for the faults (fig. 8) show an average stress field that is consistent with Late Triassic to Early Jurassic northwest-southeast extension associated with rifting of the New England crust during the initial opening of the Atlantic basin (Foland and Faul, 1977; McHone, 1978; Kaye, 1983; McHone and Butler, 1984; de Boer and Clifford, 1988; Manning and de Boer, 1989). Fault data are consistent with brittle reactivation of the northeast-striking and northwest-dipping dominant foliation (compare figs. 2 and 8).