EGU24-18376, updated on 13 Apr 2024
https://doi.org/10.5194/egusphere-egu24-18376
EGU General Assembly 2024
© Author(s) 2024. This work is distributed under
the Creative Commons Attribution 4.0 License.

Probabilistic modelling of soil properties variability for local seismic response analysis in NaTech events

Giorgia Berardo1, Leonardo Maria Giannini1,3, Alessandra Marino2, and Gabriele Scarascia Mugnozza1,3
Giorgia Berardo et al.
  • 1Earth Sciences Department, Sapienza University of Rome, Piazzale Aldo Moro 5 00185 Rome, Italy
  • 2INAIL Department of Technological Innovations and Safety of Plants, Via Del Torraccio di Torrenova 7 00133 Rome, Italy
  • 3CERI Research Center, Sapienza University of Rome, Piazzale Aldo Moro 5, 00185 Rome, Italy

NaTech events define the interaction between natural hazards and industrial accidents leading to major fires, explosions or toxic releases where hazardous substances are involved. Among NaTech events, earthquake is one of the most important, because it affects the entire plant and can cause simultaneous damage to different equipment. This study evaluates the local seismic response at the Bussi MHIP Chemical Company, located in Abruzzo region (Central Italy). It is a representative Major-Hazard Industrial Plant (MHIP) subject to Italian standard Decree (D.Lgs. 105/2015; Directive 2012/18/EC - Seveso III), which requires a multidisciplinary approach, given the high complexity of the problem and the numerous types of equipment. We focus on the EURECO plant, located inside the Bussi MHIP, which is featured by high seismicity with potential seismic amplification phenomena due to its complex geo-lithological setting. In the present framework, the influence of soil profile properties under the EURECO plant is investigated through a stochastic site response analysis. We aim to conduct a sensitivity analysis to assess the amplification factor (AF) variability with the randomness of soil properties and geological setting. The implementation of a geological reference model was supported by building a geo-database based on 125 collected boreholes stratigraphies, which identify the main lithological units and their spatial relationships. The limited availability of deep boreholes led to the adoption of other 110 virtual boreholes, where lithological column was extrapolated from stratigraphic sections, geological maps and information from the literature. This approach allowed us to integrate and combine actual with virtual subsurface data, through expert interpretation. Soil properties were collected from a review of relevant literature and previous Seismic Microzonation studies of geologically compatible areas. From this collection, we derived physical (e.g., g, Vs) and dynamic (e.g., shear modulus G/G0, damping D) properties for each soil type. In this study, the shear modulus and damping curves proposed by Darandeli (2001) and modified by Gaudiosi (2023) were applied to each soil type. The geotechnical properties and the variability associated with their distributions have a high impact on the seismic response of a site.  From the geological model, we defined the range of variability of the parameters associated with each seismic unit (e.g. Vs shear waves, bedrock depth, geomechanical properties). We assumed two scenarios in the ultimate conditions of the plant, the Safe Life State (SLV) and the Collapse Limit State (SLC) according to the National Building Code (NTC2018). The seismic inputs were selected using the Probabilistic Seismic Hazard Analysis (PSHA) approach. We performed numerical simulations of 1D – 1D stochastic -2D site response to take into account the influence of the variability of soil parameters and selected seismic input on amplification factor (AF). The development of a seismic response analysis for each simulation allowed us to calculate the AF for the EURECO plant within the estimated fundamental period of vibration of a specific H202 storage tank located inside it. The simulated seismic scenarios could involve the overturning of the storage tank, leading to fires or the release of toxic substances.

 

How to cite: Berardo, G., Giannini, L. M., Marino, A., and Scarascia Mugnozza, G.: Probabilistic modelling of soil properties variability for local seismic response analysis in NaTech events, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-18376, https://doi.org/10.5194/egusphere-egu24-18376, 2024.