Description
Groundwater inflows into tunnels constructed in fractured crystalline bedrock limits the advancement of the tunnel and, if not controlled, can seriously impact the overlying environment. As a consequence, prediction of groundwater inflow into tunnels is of importance. Among the most difficult problems in hydrogeology is describing the heterogeneity in fractured-rock systems. Accurate characterization of properties controlling fluid movement is problematic in fractured-rock, where orders-of-magnitude contrasts in hydraulic conductivity can occur over short distances. Identification of high hydraulic conductivity fracture zones is difficult, as direct hydraulic information is sparse. Groundwater research is dominated by papers describing the fractured rock system as one of decreasing fracture density and permeability with depth, with much of the literature suggesting that permeability decreases exponentially with depth. This study focused on two principle objectives: (1) to assess the cumulative frequency distribution of fractures and permeability of fractures with depth in fractured crystalline bedrock, and (2) to compare tunnel inflow and drawdown estimated by a numerical model to the most commonly used analytical solutions. Fracture and hydraulic conductivity data used in this research was obtained from a geotechnical investigation conducted by the URS Corporation (URS) in 2001 for the San Diego Water Authority (SDWA). The SDWA was evaluating construction of a water conveyance facility in fractured crystalline bedrock between the Imperial Valley and San Vicente Reservoir in San Diego County, California. The geotechnical investigation involved regional and site specific investigations to gain a better understanding of the soil, rock and groundwater conditions that would be encountered along the tunnel alignments. Statistical analysis of the data showed that fracture density in crystalline bedrock of San Diego County appeared to be relatively constant in all borings between 40 and 300 feet below ground surface (bgs). Two deeper borings showed similar results: boring C5A-9A showed a nearly constant fracture density from 40 to 1,000 feet bgs and boring C5A-13 showed a constant (but lower) fracture density from 200 to 1,750 feet bgs. Statistically, no change in the fracture frequency with depth was observed. Measured hydraulic conductivity also showed no significant decrease in permeability to a depth of 200 feet bgs. For the estimation of tunnel inflow portion of this study, a sixteen-layered conceptual model was designed to simulate transient discharge of fractured crystalline bedrock into a tunnel at a depth of 2,000 feet bgs intersecting a permeable fracture. A total of ten simulations were run using measured hydraulic conductivities that ranged from 0.485 feet/day to 6.72 feet/day for a single vertical fracture and 0.001 ft/day for background bedrock. Tunnel inflow was also evaluated using analogies to the transient and steady-state, confined aquifer well hydraulics equations. The results obtained from the numerical simulations and Cooper-Jacob solution showed that predictions of the transient hydrogeological responses induced by a tunnel in fractured crystalline bedrock are comparable. The results of the numerical model predicted that groundwater flow to such a tunnel would produce more than 7 feet (K=0.485 ft/day) to 35 feet (K=6.72 ft/day) of water table decline directly above the tunnel and would impact a substantial area around the fracture after 584 days of inflow. The steady-state and transient solutions were determined to be inapplicable to calculate drawdown of the water table due to assumptions that cannot be assessed. Therefore, the numerical model simulations were needed to effectively predict drawdown of the water table as the result of groundwater inflow into a tunnel.
