Wakes in and between very large offshore arrays

Projected power output and wake extents are presented from new simulations with the Weather Research and Forecasting (WRF) model v4.2.2 for the large offshore wind energy lease areas along the U.S. east coast. These simulations assume nearly 2000 IEA 15 MW reference turbines are deployed with a spacing equal to the mean of smaller European offshore wind farms (7.7 rotor diameters). Results show marked differences across two wind farm parameterizations. Generally, the modified Fitch parameterization (wherein TKE generation by the rotor has been decreased) generates lower power production estimates, and more spatially extensive and deeper wind farm wakes than are manifest in output from the Explicit Wake Parameterization (EWP). For example, under conditions of moderate freestream wind speeds (∼ 4-10 ms −1 at hub-height) and turbulent kinetic energy (TKE ∼ 0.2 to 1 m 2 s −2 ), cumulative power output (summed over all 15 lease areas) is substantially greater (∼ 25% higher) in output from EWP than Fitch. These differences have real implications for power production and thus both expected revenues and grid integration. The cumulative power production and mean normalized wake extent also exhibit sensitivity to the order in which the overlapping inner domains are computed and the number of inner domains. This effect is smaller than differences from two wind farm parameterizations. Analyses focusing on the seven adjoining lease areas south of Massachusetts indicate differences in the two schemes are magnified over the largest offshore wind clusters (with expected installed capacity of > 10 GW and spatial extent of 3675 km 2 ).