An Investigation of Probabilistic Stratification Models for Assessing Deep Excavations in Urban Environments

Despite the significant uncertainties inherent in geotechnical design and construction, the field predominantly relies on deterministic analyses. To address this limitation, the probabilistic-digital-twin framework has the potential to systematically integrate probabilistic methods into geotechnical design and construction. The framework can explicitly account for uncertainties which is relevantto ensure both safety and efficiency of design and construction.It does require realistic probabilistic models of the geological and geotechnical properties at a site. Recent probabilistic approaches in subsoil models have focused mostly on incorporating spatial variability of soil properties withoutaddressing the significant uncertainties associated with soil stratification. The objective of our work is to develop methods for incorporating uncertainty in soil stratification intogeotechnical analysis. In this paper, we investigate the effect of such an uncertainty assessment in a case study involving the construction of a subway station.We focus on the settlements induced by the deep excavations, which can impact neighboring buildings and are thus of particular interest in urban environments. The results demonstrate that stratification uncertainty does significantly impact the predicted settlements and hence the optimal design.

Keywords: Deep excavations in urban environments, Probabilistic modeling, Geotechnical design and construction.