SURFACE VS. SUBSURFACE FRACTURE CHARACTERISTICS ALONG CROSS STRIKE TRANSECT IN EASTERN MASSACHUSETTS

The trends of major fracture sets (dip >45°), their geographic distributions (domains), and their characteristics (spacing, trace length, and planarity) were measured in surface outcrops and in a 9 km section of the tunnel to determine how well fracture data collected at widely-spaced surface exposures can be extrapolated to a depth of 70 to 90 meters below grade. For the surface fracture data set, mean trends of fracture sets and domains were determined from 1513 measurements collected at 21 outcrops located within 3 km of the trace of the tunnel. Spacing, trace length, and planarity were determined from scanline measurements. For the tunnel data set, 413 fracture measurements were made to determine major sets and domains and a smaller subset was used to estimate fracture characteristics.

Five fracture sets (14, 38, 86, 117, and 171) were identified in the outcrops and seven sets (13, 29, 41, 62, 132, 159, and 175) in the tunnel. It was found that the most common fracture sets (14°, 38° and 171°) at the surface are also found in the subsurface. The 14 and 171 fracture sets correspond well with the 13 and 175 sets in the tunnel. The 38 set observed at the surface includes parts of the 29 and 41 sets in the tunnel. The 86 set does occur in the tunnel but is undersampled because it is aligned with the tunnel. The 62 and 159 sets occur in the tunnel but are not seen at the surface. Although large areas are devoid of outcrops, comparison of surface and subsurface fracture domains indicates that only the 14 and 171 sets show a reasonable spatial overlap with the 13 and 175 domains in the tunnel. Domain overlap analysis to predict subsurface fracture sets may be most reliable for the dominant sets. Furthermore, these latter sets are the fractures generating most of the groundwater inflow into the tunnel. The characteristics of surface and subsurface fracture sets were evaluated using non-parametric statistical tests to determine if statistically significant differences in fracture properties existed between different fracture sets. Median fracture spacing and trace lengths for the 13 and 175 sets in the tunnel are significantly wider and longer than the corresponding 14 and 171 sets at the surface. Fracture planarities showed no significant differences between any of the surface and subsurface fracture sets.