Abstract
Designing wave energy converters (WECs) relies on numerical simulations extensively in the early design phase due to the expenses associated with physical testing and prototyping. However, model assumptions and simplifications are often made to decrease computational expenses with numerical models which can decrease the realistic representation and accuracy of the simulation and increase the overall uncertainty of WEC performance. As wave energy technology advances toward commercialization, realistic representation of wave resources and WEC performance will be essential for at-sea testing and applications such as grid integration. Misrepresenting projected power variability could lead to system failures. The conventional time series generation method used in WEC modeling, the random-phase method, does not capture the true variability of real waves. Performance outputs obtained from simulations featuring more realistic, statistically accurate wave scenarios will yield a more comprehensive analysis of WECs, enabling more informed design decisions for physical prototyping and testing. Prior work using the random-amplitude method has demonstrated improved representation of realistic waves and shown the impact this representation has on increasing the power variability of WECs. In our study, a statistical analysis is conducted to determine the number of time series required to represent key wave height percentiles from artificial time series. Here, we focus on larger wave percentiles to better understand how time series generation methods can impact the assessment of design load cases, key to the survival and operation of WECs. Additionally, we investigate whether the time series duration impacts the overall time required to represent desired wave percentiles most effectively.