Sandia is developing accurate methods to simulate altered wave propagation due to the operation of wave farms by enhancing the open-source spectral wave model, SWAN. The new model, SNL-SWAN, remains open source and will help the WEC industry assess the potential environmental effects created by changes in wave climates associated with the deployment of various sizes and configurations of wave farms in the ocean.

Energy spectra at gage in the lee of the WEC-Array for OR2 (top) and OR3 (bottom) and directional spreading on (LHS) and off (RHS).

Energy spectra at gage in the lee of the WEC-Array for OR2 (top) and OR3 (bottom) and directional spreading on (LHS) and off (RHS).

Recently, Sandia modified the SNL-SWAN Alpha code to include a frequency-dependent WEC module. This latest version has been named SNL-SWAN Beta. The SWAN test cases were run to verify the baseline code’s functionality, and the frequency-dependent functionality of SNL-SWAN Beta’s WEC module was verified. This was accomplished by comparing the shape of the incident energy spectrum (before the WEC) to the lee energy spectra (after the WEC), and noting varying energy absorption in different frequency bins.

SNL-SWAN Beta was then used to model the Columbia Power array tests performed in the Oregon State Tsunami Wave Basin. The results of these simulations were then compared to the experimental data from these wave tank tests, and also to simulations using

  • the baseline SWAN code,
  • SNL-SWAN Alpha, and
  • the OSU Module for SWAN.

Baseline SWAN models the WEC as a constant transmission coefficient; SNL-SWAN Alpha calculates a constant effective power transmission coefficient based on the WEC’s power performance data; and the OSU module is a function external to SWAN that is used to modify the wave spectra at the line of WECs, and then repropagate that spectra to the next line of WECS.

Theoretically, the OSU Module and SNL-SWAN Beta version should be very similar because they are both frequency dependent. However, because the OSU module is external to SWAN, and requires multiple runs of SWAN, there may be numerical artifacts from the additional boundary conditions.

This work was published in the GMREC/METS 2014 conference proceedings, and presented in Seattle on April 17, 2014.