Motivation
The trend in recent years is to locate moored (LNG) terminals more in shallow and intermediate water depths (15-40m) in unsheltered conditions, instead of in the harbor. This led to problems due to the fact that the wave field around the terminals was influenced by interactions with the shore. These interactions produce low frequency (LF) waves (in the frequency range of 0.005 - 0.05Hz) coming from different directions, which are in the range of the mooring system eigenfrequency (and thus induce substantial excitations). Usually, ship design conditions assume deep open water, where these LF wave components are negligible. This led to a substantial underprediction of the ship motions for the first nearshore terminal designs. Coastal engineering methods model the interactions with the coastline, but do not consider the impact of the resulting waves on an offshore moored structure. These observations were reason for MARIN, Single Buoy Moorings (SBM), Bureau Veritas (BV) and WL | Delft Hydraulics (currently Deltares) to start the Joint Industry Project (JIP) sHAllow WAter Initiative (HAWAI), which had the aim to 'To improve the reliability of offshore (LNG) terminals in shallow water by using the combined expertise of offshore hydrodynamics and coastal engineering'. One of the most important conclusions drawn after 3 years of JIP-HAWAI was that LF (free) waves can dominate the overall response of an LNG carrier in shallow water. The project brought more insight into shallow water wave mechanisms and the influence of LF free waves on the motion response of a vessel, but a number of research questions remained unresolved. This was the reason to start a follow-up JIP HAWAII in 2009, which is a collaboration between MARIN, Deltares, Shell and BV. The aim of the second project is 'To develop a consistent design methodology for offshore terminals in a nearshore wave climate'. The ultimate goal is to make the knowledge gained during JIP-HAWAI practically applicable for the JIP participants.
Problem description
In a design process, several steps need to be taken. Started is with preliminary rough methods, which are step by step refined until at the end of the process completely detailed calculations are performed. An important step in the LNG terminal design is the coupling between a wave model (from coastal engineering) and a ship motion response model (from hydrodynamic engineering). The most detailed version of this coupling is almost established, by applying time series of surface elevations, pressures and velocities at specified locations from the wave model directly to the panels on a ship's hull in a diffraction method. This is not a procedure recommended to use during a preliminary design stage however, it is very time consuming. If for instance only the main dimensions of the vessel need to be determined yet, a whole computation including all possible details is a waste of time.
For this reason, a spectral approach is desirable. In this theory, the wave field from a wave model is analyzed to produce a wave spectrum. This spectrum serves as input for the vessel motion calculation subsequently, superimposing the wave forces on the vessel resulting from each wave component with an amplitude, frequency and direction (and possibly phase angle in case of a phase-resolving model). These wave motions in turn serve as input for the motion calculation. The used wave model might even be a less sophisticated model in such a first step, as long as it calculates LF waves and shore interactions correctly.
The problem with this approach is that there is no validated method available to analyze the directional spectrum of a wave field consisting of LF waves in 2DH. There is a range of methods available for determining directions of primary high frequency waves, but these are not directly applicable to LF waves. Problems encountered trying this are the required size of the measuring array, the lenght of the time series (long enough to measure enough waves but short enough to assume stationary conditions) and the distinction between velocities of bound and free LF waves (the first category consists of second order difference frequency waves travelling with the group speed, while the second catorogy travels with the phase velocity). Other considerations concern interactions of waves with the shore. A lot of the analysis methods use a random-phase assumption, which is breached for phase-locked reflecting waves. Another constraint is that the simpler methods assume one or a couple of principal directions, and are therefore not really suitable to analyze wave fields in coastal areas (where directional spreading is high).
Graduation project
The aim of my graduation project is to evaluate the use of existing multidirectional wave spectrum analysis tools applied to wave fields consisting of LF waves in coastal areas. This will be done analyzing artificially manufactured wave spectra and comparing outcomes. The complexity of the wave fields will gradually be increased if this is succesful. If a method is found that is promising even for fairly complex wave fields, a complete calculation including resulting vessel motions will be performed. The outcome of this exercise will be compared with the outcome of a direct time-domain coupling of wave and vessel models, and hopefully deliver satisfactory results.
Info:
