Introduction
Chile has a long history of great subduction zone earthquakes and the local tsunamis produced by them. Since the 1500’s, there have been 14 documented earthquakes with Mw greater than 8.0, including the 1960 Mw 9.5 Valdivia earthquake, the largest recorded earthquake in recent human history. The Andes LNG project is planned to store and regassify liquefied natural gas (LNG) onboard a Floating Storage and Regasification Unit (FSRU) vessel and deliver gas via pipeline to an onshore power plant. Evaluation of tsunamis for moored vessels is not typically considered for terminal design due to the low probability of a vessel calling at a terminal simultaneously with a design-level tsunami event. However, in the case of a FSRU, the vessel is on site continuously for 20 years or more, greatly increasing the probability of the moored vessel occupying the berth during a tsunami event. In the immediate vicinity of the project site, the earthquake of 1922 generated the largest recent tsunami, with a likely amplitude of 3m near the proposed terminal. In this paper, we present a tsunami hazard assessment approach to account for the great tsunamis generated by these local earthquakes, and the potential effects of this hazard on the Andes LNG floating vessels and marine terminal.
Tsunami Stochastic Approach
To provide tsunami hazard information for the Front End Engineering and Design (FEED), a “stochastic scenario” approach is used which allows for the approximate expression of tsunami recurrence periods. With information found in the scientific literature (i.e. Dong et al., 2015), it is expected that a 1922-like earthquake (i.e Mw 8.5 with rupture immediately offshore of the site) occurs once every 100-250 years. However, the earthquake return period is a function of its magnitude only, and other parameters, such as focal depth and internal rupture angles, while controlling the initial tsunami properties, play no role in this return period. Therefore, to quantify a tsunami recurrence period, we must understand the range of potential tsunami impacts that might be caused by different configurations of a Mw 8.5 earthquake offshore of the site.
We simulate a set of 16 tsunami scenarios, wherein the earthquake epicenter, focal depth, strike angle, dip angle, and rake angle are varied. The range of over which these values vary is taken from recent large earthquakes in the region. For each of these scenarios, a high-resolution tsunami simulation is performed using both a shallow water tsunami model (MOST) and a Boussinesq-type hydrodynamic model (COULWAVE). Time series of ocean surface elevation and tsunami current velocity for all 32 simulations are recorded at the project site. The current field predicted by the numerical model is characterized by the presence of large eddies spinning in the horizontal plane (“whirlpools”). The presence of these eddies was a primary motivating factor for the statistical analysis discussed here, as a single deterministic simulation might not provide a complete description of the potential variability of this complex velocity field.
From the numerical output, it is possible to generate current-based exceedance curves as a function of direction. To perform this analysis, current heading was divided into 2-degree bins, and an exceedance curve developed for each bin. With this information, it becomes possible to express the relative likelihood of a specified current-direction pair, in terms of useful recurrence periods. For example, if the design earthquake for tsunami hazard is assumed to have a return period of 250 years, and a speed of 2 m/s (at a specific heading) is only exceeded in half of the tsunami simulations, then this information can be combined to state that a speed of 2 m/s will only be exceeded, on average, every 500 years. The speed-based exceedance curves are provided to the mooring analysis, such that a tsunami hazard level that is consistent with other considered hazards (e.g. seismic) may be used for design.
Mooring Evaluation
We evaluate the performance of the moored vessels by running a series of dynamic mooring analyses of the tsunami events. The design basis for the terminal specifies that the FSRU shall be capable of departing berth in an emergency. However, the warning time between earthquake generation event and tsunami arrival is too short to allow the vessel to be ready to depart. The dynamic mooring simulations are conducted for two scenarios: the FSRU at the berth by itself; and the FSRU and LNG carrier in ship-to-ship (STS) transfer operations. Performance results of both the LNGC and FSRU mooring are predicated on departing berth within 60 minutes and 90 minutes, respectively, of the earthquake-generated tsunami event.
We selected tsunami events for the mooring analyses based on the combined return period of the earthquake event and the current exceedance threshold acceptable for the design of the berth. The FSRU-only mooring arrangement is simulated for the first 90 minutes of the 10% exceedance tsunami (return period of 1000-2500 years). The STS mooring arrangement is simulated for the first 60 minutes of the 50% exceedance tsunami (return period of 200-500 years).
The response of the vessels during the design tsunami events is modeled using a dynamic analysis program (AQWA) used for the calculation of forces and motions of multiple floating bodies. The analysis predicted forces in mooring lines and fenders, as well as vessel motions throughout the tsunami event. The tsunami currents were simulated by applying the time series of current magnitude and direction at the vessel location, assuming the current field is uniform over the length of the vessels during the early part of the tsunami event, before eddies and other chaotic behaviors develop. Additional mooring line stresses due to fluctuations of the water surface elevation are calculated and added to the mooring response for each time-step of the AQWA simulation results.
By combining the tsunami probability evaluation with generation of mooring loads, we assess the mooring performance during tsunami events and develop an approach for incorporating tsunami mooring loads into the marine structural design.