Lithofacies-controlled 3D geotechnical modeling for liquefaction potential assessment in a Holocene coastal system (Porto San Giorgio, Italy)

Liquefaction assessments based on Cone Penetration Tests (CPT/CPTU) provide point-scale estimates, yet their spatial representativeness is limited in heterogeneous coastal environments. We present a facies-controlled 3D geological-geotechnical workflow to extend Liquefaction Potential Index (LPI) estimation from discrete CPTU points to a model domain, using the Porto San Giorgio coastal plain (Italy) as a case study. The approach integrates 31 boreholes and 14 CPTU tests to reconstruct the stratigraphic architecture of Holocene coastal-alluvial deposits over a Pleistocene marine-clay substrate, and to populate a 3D geocellular model with cone resistance (qc) and sleeve friction (fs). Lithofacies were modeled geostatistically and used as hard conditioning domains, so qc and fs were interpolated within facies without mixing across facies boundaries. Property modeling used Gaussian Random Function Simulation (GRFS), Sequential Gaussian Simulation (SGS), and Kriging Interpolation (KI), combined with alternative lithofacies realizations. Two CPTUs were withheld for validation, comparing measured and synthetic qc–fs profiles; the 10 best realizations were then used for LPI computation. Results show that GRFS and SGS better reproduce measured CPTU behavior when conditioned by lithofacies models that preserve the dominant depositional architecture and a moderate degree of heterogeneity. Synthetic LPI values are closer to measured values in laterally continuous shoreface–beach deposits, whereas heterogeneous or poorly constrained lithofacies-body geometries produce wider LPI ranges and generally overestimate liquefaction potential. This highlights how uncertainty in sedimentary-body geometry, continuity, and thickness propagates into geotechnical property models and local LPI estimates. This work demonstrates the value of facies-controlled 3D geotechnical property modeling for liquefaction assessment by linking geological uncertainty, geotechnical property distributions, and LPI variability in Holocene coastal deposits.