Adverse effects of pore-pressure drainage on stress measurements performed in deep tunnels: an example from the Lower Kihansi hydroelectric power project, Tanzania

Abstract

We describe a campaign to measure stress within a gneissic ridge near the edge of the Tanzanian craton . The measurements were conducted to determine whether the level of minimum principal stress along the trajectory of a hydropower pressure tunnel attained the level of 10 MPa, as required to leave the tunnel unlined. Extrapolation of early hydrofracture and overcoring measurements conducted mostly in boreholes drilled behind face suggested that the requisite level was not quite met. However, the minimum principal stress estimates from the instantaneous shut-in pressures (ISIPs) of the hydrofracture tests generally showed large scatter and were sometimes less than the estimated pre-disturbance pore pressure . It was suspected that the stress tests were affected by stress alteration around the tunnel due to pore pressure drainage. To test this, hydrofracture and hydro-jacking tests were conducted in two long, horizontal holes drilled ahead of the excavation face into relatively undrained rock, and a third vertical hole drilled behind the face. All were near the critical location where the pressure tunnel enters the powerhouse. The minimum principal stress estimates in the ahead-of-face holes were found to be much higher than those from the behind-face hole, consistent with the hypothesis. Estimates of the drawdown in pore pressure about the tunnel and test holes suggested a coupling factor between drawdown and minimum principal stress magnitude of approximately unity. This is slightly higher than predicted from an hydromechanical model that explains the coupling between the pore-pressure and stress fields in terms of fracture compliance and poro-elasticity of intact blocks. Nonetheless, both results demonstrate the importance of accounting for drawdown when conducting stress measurements in boreholes drilled from tunnels. We also note some instances where hydrofracture and hydro-jacking tests yielded different estimates for the minimum principal stress magnitude.