The wave energy industry is entering a new phase of pre-commercial and commercial deployments of full-scale devices, so better understanding of seaway variability is critical to the successful operation of devices. The response of Wave Energy Converters (WEC) to incident waves govern their operational performance and for many devices, this is highly dependant on spectral shape due to their resonant properties. Resource assessments, device performance predictions and monitoring of operational devices will often be based on summary statistics and assume a standard spectral shape such as Pierson-Moskowitz or JONSWAP. Furthermore, these are typically derived from the closest available wave data, frequently separated from the site on scales in the order of 1km. Therefore, variability of seaways from standard spectral shapes and spatial inconsistency between the measurement point and the device site will cause inaccuracies in the performance assessment. Potential differences in estimated and actual incident wave power are investigated using measured data. Concurrent wave data from two wavebuoys have been collected from the EMEC full scale test facility in Orkney, Scotland and the Galway Bay benign test site, Ireland along with single buoy data from the Wave Hub site off the north Cornwall coast in the UK.
Spatial variability is investigated through the examination of limited data from wave measurements taken at separations of 200m, 500m and 1500m. Differences between concurrent measurements are identified and compared to theoretically predicted variability, in terms of their magnitude and statistical properties. Comparisons of the data over the available separations indicate a possible dependence of the measured differences in the spatial domain, although site-specific properties must also be considered.
The deviation of the measured data from the empirically derived spectral shapes is investigated, quantified and compared to other sites from published data. This investigation involves the identification of bi-modal seaways and it is found that this deviation reduces with increasing wave height, although the degree of deviation, which is being investigated here, is site specific to some degree.
Finally, this analysis is implemented for chosen elements of the wave height-period scatter diagram that are populated at the selected sites. The analysis techniques derived are employed to quantify not only spatial and spectral variation and variability, but also variations from benign site to full- scale site. The analysis highlights potential variations in power incident at a site from those estimated using a standard spectral shape at a separate measurement point.
The results of the work presented here are only indicative as the spectral variation work is based on just four months of data, while the spatial variation exercise is applied to just 1 month’ s concurrent measurements. Robust conclusions can only be drawn when the methods are applied to a more extensive database.