Seismic performance of multiply excited, dynamically interactive energy infrastructure systems

Abstract

Given the high complexity of modern energy infrastructure systems, seismically induced dynamic interaction phenomena are common and have prominent practical implications due to the key role these assets play in community resilience. While design codes typically allow these systems to be analysed independently and tend to consider various interactions via simplified approaches, infrastructure systems, in reality, situate in complex environments where they not only interact with adjacent soil and with each other, but can be affected by simultaneous asynchronous excitations or multiple hazards too. The problems studied in this dissertation thus focus on the exposure of interactive energy infrastructure systems to coupled multiple excitations and/or coupled multiple hazards that are located in seismically active regions and supported on soft soils.
Concerning coupled structures, a realistic scenario of structure-pipe-structure interaction within a natural gas processing plant is investigated by means of numerical parametric study and pseudo-dynamic hybrid simulation. Critical seismic demands on the interconnected pipe elbows are proven probable due to the different dynamic characteristics of the two supporting structures hence their resultant out-of-phase oscillation during earthquakes. Key parameters governing the dynamics of structure-pipe-structure interaction are identified and their critical combination explored.
Concerning coupled multi-hazard, the performance of healthy and ageing monopile-supported offshore wind turbines subjected to simultaneous excitations of wind, wave and earthquake is investigated by means of Latin Hypercube sampled, cloud-based fragility analysis. The significance of considering joint excitations from multiple hazards is probabilistically highlighted throughout the entire design life of an OWT, during which the structural demand is found nonnegligible and the structural capacity worsen with age. Recommendations are made on the appropriate selection of statistical regression methods and intensity measure pairs for the purpose of deriving multi-hazard fragility surfaces for operating OWTs on its entire range of allowable inflow wind speed, considering the active blade-pitch control mechanism typical for most modern utility-scale wind turbines.
The gaps addressed in this dissertation highlight the importance of assessing multiply excited, dynamically interactive infrastructure systems with appropriate sophistication; and that the applied methodologies are interchangeable between and beyond the discussed scenarios. It is noted that the dynamics of most infrastructure systems are rarely, if ever, decoupled from their surroundings, hence the appropriate consideration of all boundary conditions and relevant hazards is likely to be necessary so to design, maintain or retrofit such systems more safely and economically.

Date of Award	6 Dec 2022
Original language	English
Awarding Institution	University of Bristol
Supervisor	Anastasios Sextos (Supervisor) & Adam J Crewe (Supervisor)