Spatial and correlation analysis of engineering-geological survey data for logistics center construction

Abstract

Introduction: at present, the geostatistical methodology is broadly used abroad for constructing spatial-correlation and spatial-stochastic models of lithoengineering systems, including description and analysis of soil body heterogeneity. The main goal of this work is an attempt to evaluate the possibility of generating a spatial-correlation model of lithoengineering space based on survey data which could be used for subsequent simulation and deterministic-stochastic analysis of geotechnical structures as well as when designing bases and foundations. Materials and methods: main input parameters for spatial analysis were geological survey report and cone-penetration test (CPT) data. The data were analyzed using descriptive statistic methods, with the calculation of particular and ambiguous values and using statistical software STATISTICA. Later use of the results of the statistical analysis aims at the application of geostatistical interpolation models (kriging) for generating spatial structures of deformation features. The article elaborates experimental correlation functions (variograms) to validate the developed spatial structures. The functions show not only the validity of the spatial simulation but also a level of heterogeneity of the researched soil body. Results: the statistical analysis of the initial data showed relative homogeneity of the researched soil body properties what is substantiated by GOST 20522 instructions. Nevertheless, the geostatistical analysis resulted in the opposite conclusion. Principally, this is due to the lack of information on the researched geological medium. Conclusions: requirements of current Russian standard GOST 20522 do not contain a sufficient amount of information for geostatistical analysis and simulation of a soil body based on modern geostatistical methods. As a result, data of standard engineering-geological researches cannot be used for developing digital foundation models as well as for using probabilistic approaches in geotechnics.