Critical review and innovations in excavation practices for soft soil conditions

Rapid urban development increasingly requires deep excavations near existing structures, where soft soils often induce significant wall and ground movements. Accurate prediction and control of these deformations are essential for safety and serviceability. This study compiles an extensive database of excavation case histories with measured wall and ground responses, and systematically classifies geometry, boundary conditions, support systems, and shoring methods. Using this dataset, we perform empirical and statistical regression analyses to identify the dominant factors governing normalized wall displacement. Building on classical formulations, we introduce a composite system stiffness index (K system ) that explicitly combines structural rigidity, excavation geometry, and soil stiffness. The index is evaluated through regression analyses, external benchmarking, and selected FEM-based consistency checks. Results show that excavation geometry and soil stiffness are primary drivers of displacement, and that K system improves correlation relative to conventional indices, providing a more practical basis for design. The proposed framework enables more reliable, data-driven excavation design in soft soils and helps reduce excavation-induced risk in dense urban environments, with potential applicability across different soft soil conditions encountered internationally.