Seismic Performance of Storage Steel Tanks during the May 2012 Emilia, Italy, Earthquakes Emanuele Brunesi 1 ; Roberto Nascimbene, Ph.D. 2 ; Marco Pagani, Ph.D. 3 ; Dumitru Beilic 4 Abstract: Field observations following the May 2012 Emilia, Italy, earthquakes have revealed the seismic vulnerability of storage steel tanks typical of the past Italian design practice, highlighting structural deficiencies observed during previous events in other areas and mostly related to lack of structural seismic design and detailing, lack of redundancy, and inadequate anchorage design and execution. Damage was disproportionately high, considering the moderate size of the events (M W ¼ 6.11 and M W ¼ 5.96 on 20th and 29th May, respectively). Failure modes observed are reported and classified. Comparison is provided with the results of high-definition finite element (FE) dynamic simulations, performed on tanks representative of those inspected. Numerical analyses, accounting for material and geometrical nonlinear- ities, at the element level, as well as for fluid-structure interaction through highly nonlinear methods, were able to capture the stress/strain concentrations that caused them to collapse. Elephant’ s foot and diamond buckling induced by hydrodynamic pressures as a result of inertial forces imparted during the earthquakes were reproduced, as well as shear-buckling failure of leg-supported tanks or sliding of unanchored systems. DOI: 10.1061/(ASCE)CF.1943-5509.0000628. © 2014 American Society of Civil Engineers. Author keywords: Earthquakes; Storage tanks; Thin shell structures; Failure modes; Buckling; Finite element models; Fluid-structure interaction. Introduction On May 20th and 29th, 2012, two earthquakes of magnitude (M W ) 6.11 and 5.96—according to the European and Mediterranean Regional Centroid Moment Tensor Solution (RCMT 1997)—hit the Emilia region in the Po Valley, one of the most industrialized zones of Northern Italy. The majority of structures severely dam- aged by the seismic events were industrial facilities: one-story pre- cast reinforced concrete structures (Magliulo et al. 2013; Bournas et al. 2013; Belleri et al. 2014) and nearby storage steel tanks, caus- ing an economic loss of approximately 5 billion Euros, mostly due to the industrial production interruption. The large amount of industrial facilities in the stricken area, in combination with their intrinsic deficiencies, makes several industries disproportionately vulnerable to these events, compared to their moderate seismic intensity. In the aftermath of the earthquakes, a large reconnaissance effort was undertaken. Field teams were deployed by the European Centre for Training and Research in Earthquake Engineering (Eucentre), and a clearinghouse (http://www.eqclearinghouse.org/ 2012-05-20-italy/) hosted by Eucentre and the Earthquake Engineering Research Institute (EERI) was prepared. Out of the significant number of the structures surveyed, this paper investi- gates the poor seismic performance provided by storage steel tanks, typical of the past Italian design practice, confirming vulnerabilities common for such early designs and already shown by other destructive earthquakes in other areas (Manos 1991; Niwa and Clough 1982; Swan et al. 1985; Stepp et al. 1990; Zareian et al. 2012; González et al. 2013). The most common types of failures observed were fracture of anchors and elephant’ s foot buckling near the base of the tanks. Generally, elephant’ s foot buckling was experienced in squat tanks, while some of the slender tanks surveyed developed diamond- shaped buckling. Total and partial collapse of legged tanks was an- other common occurrence, induced by shear failure and/or buckling of their legs due to axial forces resulting from tank overturning moment. In some cases, flat-bottomed steel cylindrical tanks, typ- ically larger than legged tanks, failed in tension at the bottom of the tank wall, in correspondence of the anchor rods or massive concrete pads. In the following, these mechanisms will be reported and dis- cussed in comparison with numerical observations, obtained from detailed FE analyses. Seismic Hazard and Ground Motions Observed during the 2012 Sequence The area stricken by the Emilia earthquake sequence can be con- sidered an area of intermediate seismicity in the Italian context (Meletti et al. 2012) as highlighted in Fig. 1, which depicts the of- ficial Italian hazard map (PGA with 10% probability of exceedance in 50 years) as well as the epicentral location of the main events of the sequence (data from INGV and ISIDe). The value of hazard—in terms of PGA with a return period of 475 years—computed for the epicentral area corresponds approximately to 0.14 g. According to the results of Stucchi et al. (2011) about half of the Italian territory (but Sardinia) has values of hazard equal to or larger than the epicentral area. 1 Ph.D. Student, ROSE Programme, UME School, IUSS Pavia, Institute for Advanced Study, Via Ferrata 1, 27100 Pavia, Italy (corresponding author). E-mail: emanuele.brunesi@eucentre.it 2 EUCENTRE, European Centre for Training and Research in Earth- quake Engineering, Via Ferrata 1, 27100 Pavia, Italy. E-mail: roberto .nascimbene@eucentre.it 3 Hazard Team, GEM Foundation, Via Ferrata 1, 27100 Pavia, Italy. E-mail: marco.pagani@globalquakemodel.org 4 M.Sc. Student, ROSE Programme, UME School, IUSS Pavia, Institute for Advanced Study, Via Ferrata 1, 27100 Pavia, Italy. E-mail: dumitru .beilic@umeschool.it Note. This manuscript was submitted on March 19, 2014; approved on May 12, 2014; published online on September 12, 2014. Discussion period open until February 12, 2015; separate discussions must be submitted for individual papers. This paper is part of the Journal of Performance of Constructed Facilities, © ASCE, ISSN 0887-3828/04014137(9)/$25.00. © ASCE 04014137-1 J. Perform. Constr. Facil. J. Perform. Constr. Facil. Downloaded from ascelibrary.org by Columbia University on 02/02/15. Copyright ASCE. For personal use only; all rights reserved.