Leakage prediction and post-grouting assessment in headrace tunnel of a hydropower project

Abstract

In the Himalayan region, tunnels are often constructed through complex and varying geological formations having rock mass with higher degree of jointing, faulting, folding, and weakness/shear zones. Such rock mass condition significantly increases the rock mass permeability which enables a higher possibility of water leakage into and out of the headrace tunnels built for hydropower projects and is a challenging situation for tunnel stability. Therefore, comprehensive leakage assessment and effective pre- and post-grouting application are essen­tial in hydropower tunnels. In this research, the water leakage was predicted by using three machine learning approaches such as Support Vector Regression (SVR), Decision Tree (DT) regression, and K-Nearest Neighbors (KNN) models. The water leakage/inflow was predicted in one of the hydropower tunnels based on the geological condition of rock mass, rock mass quality, and hydro-geological conditions. The effective post-grouting method was applied to mitigate the potential water leakage and to enhance the rock mass quality and stability of the hydro­power tunnel. It was observed that the injection grouting technique helps to make tunnels less permeable, reduces instability conditions, and ensures the long-term safety and structural integrity of the hydropower tunnels.