Seismic fragility analysis of LNG sub-plant accounting for component dynamic interaction

Earthquakes can cause significant damage to Liquefied Gas terminals, a critical part of lifeline facilities of energy supply networks, failure of which may lead to loss of hazardous material, explosion, environmental contamination, loss of functionality and disruption of business. To date, seismic r...

Πλήρης περιγραφή

Λεπτομέρειες βιβλιογραφικής εγγραφής
Κύριοι συγγραφείς: Farhan, Muhammad, Bousias, Stathis
Άλλοι συγγραφείς: Μπούσιας, Ευστάθιος
Γλώσσα:English
Έκδοση: Springer 2020
Θέματα:
Διαθέσιμο Online:http://hdl.handle.net/10889/13519
Περιγραφή
Περίληψη:Earthquakes can cause significant damage to Liquefied Gas terminals, a critical part of lifeline facilities of energy supply networks, failure of which may lead to loss of hazardous material, explosion, environmental contamination, loss of functionality and disruption of business. To date, seismic risk analysis of such facilities mainly focuses at component level, with the inherent dynamic interaction between the supporting structure and non-structural components not receiving the merited attention. In the present study, the seismic performance of an actual facility comprising of a piping system and a supporting structure is analyzed through a detailed 3D finite element model in the nonlinear regime, both as coupled and decoupled cases. Plastic strains are used as Engineering Demand Parameters (EDP) to define leakage limit state for pipes. Since the reinforced concrete (RC) pipe rack supporting structure is designed for low seismic loads, shear is recognized as the predominant failure mode. The same components are then analyzed in unison, considering coupling due to dynamic interaction. Fragility functions are evaluated for both cases using Multiple Stripe Analysis. A set of 250 strong ground motions artificially generated using the Specific Barrier Model, are employed for developing fragility curves. They are expressed with peak ground acceleration (PGA) as an intensity measure. Statistical estimation of the parameters of fragility functions are based on maximum likelihood method. It is inferred that in the decoupled case, pipelines show higher vulnerability at lower PGA, at higher PGA pipe rack can fail abruptly resulting in total failure of the system. Moreover, in coupled case the fragility of the pipes and RC rack changes substantially because of the piping system boundary conditions. Thus, concluding that the risk estimation could be erroneous if dynamic interaction is neglected.