Analysis of grouting design based on water control for rock tunnels

Abstract

Water inflow into rock tunnels is an adverse but unavoidable problem in projects under ground water table. In fractured rock masses, discontinuities are pathways for water inflow when they are conductive. When water inflow takes place, fillings in persistent discontinuities will be flew out, which may reduce the strength of the fractured rock mass and may even lead to tunnel instability and unsafe working conditions. Moreover, the excavation speed is probably reduced due to water inflow which may result in an increase in construction time and cost. Rock grouting is a conventional and commonly used water control strategy. Grouting efficiency and cost-effectiveness are two key aspects in grouting engineering. The prediction of the water inflow into rock tunnels, the investigation of the influence from grouting on water inflow and the estimation of grout take are of interest for grouting engineers. However, theoretical analyses developed so far have limited use in grouting design due to the complexity of site geological and hydro-geological conditions, and the commonly used empirical based grouting designs heavily depend on engineers’ personal experiences. Therefore, the present work tries to fill several gaps in theoretical analysis and practical grouting design based on onsite data analysis. Firstly, the relationships between the various water inflow parameters, including the water inflows into probe holes, tunnels, check holes and grouted tunnels, are obtained for the preliminary design of water control strategy. In addition, an analytical solution for water inflow into a grouted tunnel considering exponential decay of hydraulic conductivity with depth is proposed. This analytical solution reveals that when the decay gradient of hydraulic conductivity is relatively large, its influence should be considered in grouting design. To assist a cost-effective grouting design for water control, the distributions of hydrogeological properties in a real project are investigated. It is found that both the water inflow in grout holes and hydraulic conductivity at the drilling location of grout holes follow a logarithmic relationship. Besides, the influence from grouting overlaps should be taken into consideration in the calculation of field hydraulic conductivity. To assist data analysis, an App for practical grouting visualization, analysis, and prediction is proposed. Finally, the Artificial Neural Network (ANN) is used for developing predictive models for grout take forecasting to assist a cost-effective grouting design. The present study shows that it is difficult to derive a predictive model with a high degree of reliability for estimating grout take at individual grout hole. However, the reliability of the predictive models will increase after data are categorized properly. Besides, the grout take at individual station can be well predicted via geological and hydrogeological data classified by individual stations. It is hopefully that with more project data added into the project database, the accuracy of these predictive models can be further improved.