Extreme wave height for design of an offshore wind turbine installed in shallow water zone is calculated based on Goda's theoretical wave deformation model (Goda, 1975; Goda, 2010) in accordance with requirement of exceedance probability specified in IEC61400-3 (2009). Based on the results, Goda's approximation formula for the maximum wave height is adjusted to be consistent with IEC61400-3. Comparisons are made between Battjes & Groenendijk's model and Goda's model. Furthermore, approximation formula for the incipient depth for breaking wave is presented.
In IEC61400-3 (2009), extreme wave height (EWH) is specified to calculate based on the long term metocean data. However, when the sufficiently long data are not available, it can be calculated using Eq. 1 by assuming the Rayleigh distribution on the water surface motion,
where (equation) and (equation) are the TR-year expected value of the EWH and the significant wave height for the 3-hour averaging, respectively. (equation) is a wave height with the exceedance probability of 1/1000, which is approximately equal to the maximum wave height among 1000 individual waves (see, Table A.1).
In the case of deep water where wave shoaling deformation is negligible, it is reasonable to estimate EWH by assuming the Rayleigh distribution, thus Eq. 1 can be used. However, as the wave propagates to shorelines, wave height is amplified by the shoaling effect and at the same time the higher portion of wave height distribution is removed by wave breaking and is transformed into the lower portion, consequently wave height distribution deviates from the Rayleigh distribution. Therefore the Battjes & Groenendijk model (Groenendijk and van Gent, 1998; Battjes and Groenendijk, 2000) and the Goda model (Goda, 1975; Goda, 2010) are provided in the IEC standard and JSCE guideline (2010), respectively.
On the other hand, as DNVGL-ST-0437 (2016) and JIS C 1400–3 (2014) have pointed out, BG model has a drawback that it was optimized with a few flume experiment data and was not validated against field measurement data. In DNV, for the usage of the BG model, validation by site-specific wave data is required. Since the BG model is inherently a mathematical fitting by experimental data, wave shoaling and wave breaking which are essential for wave deformation in shallow water zone are considered implicitly through data which were used for curve fitting, however, the Goda model considers these theoretically.