Due to rapid urban developments and limited land in urban areas, it becomes necessary to construct underground transport infrastructure at much greater depths and quite often in poor ground congested with building structures and utilities. Deep excavation inevitably alter the ground conditions and induced ground deformations. This may cause damage to adjacent building structures and utilities. Therefore, understanding and being able to predict the performance of deep excavations to minimise movements induced damages is an important issue and a challenge to geotechnical engineers. Traditionally, deep excavation analysis makes use of conventional soil mechanics and simple structural mechanics along with appropriate modifications according to field observed behaviours. Empirical data are normally used by designers to estimate excavations induced ground movements. Such approaches have been successful in the past especially for shallow excavations at only a few meters. Because of space constraints, present day excavation sometimes has to be made vertically, and the retention system used is constructed at close proximity to existing building structures and utilities. Under such circumstances, a conventional design approach has limited use and an advanced numerical method must be employed.