Grouting in rock cavern : a case study

• Main Author: Indera, Reka Putra
• Other Authors: Zhao Zhiye
• Format: Final Year Project
• Language: English
• Published: 2019
• Subjects: DRNTU::Engineering::Civil engineering
• Online Access: http://hdl.handle.net/10356/78081
• Institution: Nanyang Technological University

The construction of deep subsea cavern is very risky and costly. One of the major risks is water inflow into the cavern. Rock pre-grouting is used to seal the tunnels and caverns from excessive water inflow. This method has been considered to be notoriously unpredictable and relied heavily on the engineers’ experiences. This research did a case study with the grouting data from tunnel OT01-c provided by the Jurong Rock Cavern (JRC) owner to predict the grouting volume needed to seal the tunnel. Large amount of grouting data was extracted, pre-processed and transformed to be a clean dataset fit for Artificial Neural Network (ANN) input. The ANN model developed has reasonable capacity (accuracy and reliability) and could be used to predict grouting volume for different tunnel section. In addition of using machine learning tool like ANN, the research also explored the data using conventional techniques such as regression and statistical graphic methods. These methods have revealed many insights regarding the relationship between parameters. For example, it was found that Rock Quality Designation (RQD) has little importance in influencing the grout take volume. This research has provided a broader understanding and perspective about rock grouting especially about parameters that influence the grouting volume.