Grouting in rock cavern by data mining approach

• Main Author: Teng, Keith Chi Han
• Other Authors: Zhao Zhiye
• Format: Final Year Project
• Language: English
• Published: 2019
• Subjects: DRNTU::Engineering::Civil engineering
• Online Access: http://hdl.handle.net/10356/77798
• Institution: Nanyang Technological University

In tandem with the increasing need for Singapore to venture underground in view of its dwindling available surface land, the Jurong Rock Caverns was constructed 130 metres below the Banyan Basin as a petrochemical storage facility. Its sheer depth resulted in it being susceptible to water seepage, which undermined structural stability and operational safety. To mitigate the water seepage, grouting was administered to reduce the hydraulic conductivity of rock strata through the injection of grout into its discontinuities. This study concerns itself with determining relationships between the site investigation parameters and the eventual grout take of the Jurong Rock Caverns’ tunnels by means of Data Mining and Artificial Neural Networks. The methodology is focused on the processing of data from obtained during onsite investigations, which were mined and sieved to determine suitable input parameters in the formulation of predictive models for grout take. Two separate analyses were executed– Individual Grout Hole analysis and Station-based analysis. The former aims to predict the Grout Take and Grout Pressure required for each individual hole while the latter focuses on predicting the Cumulative Grout Take along the tunnel length. The predictive models generated will aid in grouting design, as well as provide a reliable basis for economical quantity surveying of grout required in the construction of future subterranean projects.