PSI - Issue 78

Chiara Miglietta et al. / Procedia Structural Integrity 78 (2026) 309–316

310

1. Introduction The Seveso III Directive (Directive 2012/18/EU) is a key piece of European Union legislation aimed at preventing major industrial accidents involving dangerous substances and limiting their consequences for human health and the environment. It applies to industrial plants where hazardous chemicals are present in significant quantities, such as chemical plants, refineries, and storage facilities. Under Seveso III, operators of such establishments are required to implement stringent safety measures, conduct risk assessments, and develop emergency response plans in coordination with local authorities. In Italy, Directive 2012/18/EU dealing with the control of major-accident hazards involving dangerous substances is implemented through the Legislative Decree No. 105 of 26 June 2015. In this context, the seismic vulnerability of industrial facilities subjected to Seveso III regulations represents a critical aspect of risk assessment and management. Earthquakes can trigger cascading effects in these establishments, potentially leading to the release of hazardous substances and causing severe environmental and health impacts. Therefore, it is essential to integrate seismic risk into the safety reports and structural assessments required by the directive. This includes evaluating the resilience of process equipment, storage tanks, and pipelines to seismic loads and ensuring that emergency response systems remain operational in the aftermath of an earthquake (Marino et al., 2019). During a seismic event, a rupture anywhere in the system could release hazardous substances. The most dangerous ruptures are those occurring in locations where are not available safety valves leading to the partial or complete emptying of tanks. This issue could typically occur in the connection between the tank wall and the pipeline, at a point not equipped with valves. The experimental campaign presented in this paper is a part of an extensive experimental program aimed at investigating the complex behaviour of pipe-to-tank connections by analysing their damage modes and their hysteretic behaviour. This study has been developed in the framework of the MITPLANT research project funded by INAIL, the Italian National Institute for Insurance against Accidents at Work. Although pipe-to-tank connections are of paramount importance for industrial plant safety, few studies available in the literature investigate their seismic response. Wieschollek et al. (2013) conducted six monotonic and cyclic tests on three typical configurations of shell nozzle reinforcements, while nozzles installed at sidewall of a reduced-scale cylindrical tank were analysed by Aiba et al. (2000). Reza et al. (2014) and La Salandra et al. (2016) conducted monotonic and cyclic tests of non-standard bolted flange joints under pure bending and axial loading and under combined axial and shear loading. Varelis et al. (2013) focused on the behaviour of pipe elbows with internal pressure by investigating the in-plane bending capacity through cyclic tests. Tee pipe junctions were tested under strong cyclic loading by Papatheocharis et al. (2013). The experimental program presented in this work consisted of monotonic tests on steel specimens simulating pipe to-tank connections. Three different loading conditions were considered: axial, transverse, and a combination of both. The study aims to characterise the hysteretic behaviour of the joint with future cyclic tests and to identify the possible failure mechanisms due to seismic loads. Preliminary numerical analyses were carried out to predict the results of experimental tests and to calibrate loading protocols (Miglietta et al., 2025). 2. The experimental campaign The experimental program was developed to characterise the nonlinear cyclic behaviour of pipe-to-tank connections under different loading conditions, to consider all the possible scenarios induced by seismic loads. Tests were performed by applying loads in axial, transverse, and combined directions, following monotonic loading protocols. The complete experimental program includes three monotonic tests, one for loading direction, and six cyclic tests, two for each specimen. To simulate a pipe-to-tank connection, specimens were designed in accordance with the EN 14015:2006 standard, which specifies design requirements for nozzles. The specimens were made up of a perforated steel plate, simulating the tank wall, with a steel pipe welded into the opening to replicate the pipe-to-tank connection. Pipe and plate thicknesses were chosen considering the minimum values reported in EN 14015 for a tank diameter until 15 m. The material properties, selected according to the standard, are reported in Table 1.