Hydrogeology of the Blue Mountains, NSW: Simulating impacts of bore abstraction and sewer tunnel inflows on stream base-flow

Abstract

The Blue Mountains sandstone aquifer is an important water resource for domestic, rural and industrial use, and also hosts numerous ecosystems that are highly or entirely dependent on groundwater. In this study, conceptual and numerical models of the Blue Mountains aquifer system have been developed in order to assess the impact of bore use and tunnel drainage on groundwater levels and resultant base-flow in streams and groundwater dependent ecosystems in the Katoomba - Wentworth Falls area. The aquifer system comprises gently east-dipping sandstone and claystone units of the Triassic Narrabeen Group, which form prominent cliffs that are deeply incised by stream gullies. The permeability in the rock mass is largely controlled by fractures and hydraulic conductivity decreases with depth from -0.1 mid near the surface to -0.005 mid (at 80 m depth). Specific storage was estimated at - 10-6 m L from barometric fluctuations in piezometers and specific yield is estimated at approximately 0.02 (2%). Groundwater recharge is between -4 and -9 % of rainfall. Finite difference models covering the Katoomba-Wentworth Falls areas were developed using MODFLOW -2000 and calibrated against bore water levels and initial tunnel inflows. Water balance calculations indicate that bore use represents between 10% and 30% of rainfall recharge over the Katoomba-Wentworth Falls area, in excess of the suggested sustainable yield (5% of rainfall recharge). Model simulations indicate that stream base-flow is reduced by between 2 and 6% for an estimated current licensed pumping rate of 500 m3/d (10% of rainfall recharge). Inflow into the unlined sewerage tunnel system equates to 17- 30% of rainfall recharge in the model area. A significant drawdown effect is predicted in both the upper and lower aquifers. leading to predicted decreases in local stream base-flow by up to 50%. Transient simulations suggest that most of the drawdown and impact on stream and cliff seepage should have occurred in the first 8 - 10 years after the construction of the tunnel (i.e. by 2005). No systematic study was undertaken to continuously monitor the impact of the tunnel system on the local water table or stream base-flow. However, declines in water levels of between 0.7 and 7.7 m following completion of tunnelling were observed in 5 boreholes located within 100 m of the tunnel. Continuous groundwater monitoring undertaken since 1997 (2 years after tunnel completion) has not detected continued drawdown associated with the tunnel. Verification of base-flow impacts was not possible due to the lack of stream monitoring stations in the study area. The lack of observed drawdown in piezometers is likely due to drawdown being controlled by, and largely restricted to fracture zones and highlights limitations in using bulk (effective) hydraulic parameters in fractured rock systems to predict drawdown at the scale of individual monitor bores. However, the modelling shows that unlined tunnels, while relatively small in diameter and volume, have a large regional influence. Because of this, tunnels need to be incorporated into sustainable yield estimates.