Understanding and representing subsurface heterogeneity is highly relevant yet remains a significant challenge in developing reliable groundwater flow models. In this work, a novel stochastic subsurface modelling approach, tailored to unconsolidated sediments, which are characterized by high complexity and heterogeneity in terms of its geological structure and hydraulic properties, based on ArchPy framework for hierarchical 3D geological modelling, is proposed and tested for the drinking water supply area of the city of Görlitz, in framework of the project CRossWATER, which is funded by the EU-Interreg-Poland-Saxony-Programme. A large borehole dataset served as the basis for developing a geologically informed underground model. After pre-processing and reinterpreting this dataset, an ensemble of stochastic realizations is generated to represent different possible configurations of geological units, lithofacies distributions, and associated hydraulic property fields by applying the ArchPy framework. These ensembles collectively account for both structural and parametric uncertainty in the modelled domain. Subsequent analyses explored probability-based and information-theoretic measures, such as entropy metrics, to assess spatial uncertainty and evaluate the degree of confidence in subsurface representations. Ultimately, this work aims to enhance understanding of spatial hydrogeological uncertainty arising from sedimentary processes, heterogeneity, and anisotropy, and to evaluate how these factors influence not only geological interpretations but also subsequent analyses of groundwater flow and transport processes. The approach and workflow presented here primarily serves as a foundation for the future exploration and reduction of uncertainty in hydrogeological modelling for the groundwater resource management in Görlitz, which can easily be transferred to other study areas with similar characteristics and data availability.