Extreme Wind Speeds at US Offshore Windfarms in the Future Climate

Title: Extreme Wind Speeds at US Offshore Windfarms in the Future Climate

Authors: Fred Letson, Jacob Coburn, Xin Zhou, Kelsey Thompson, Melissa Bukovsky, Rebecca J. Barthelmie, Sara C. Pryor

Offshore wind energy is an important and quickly growing part of the US energy sector. More than 50 GW of offshore wind energy capacity is currently in the construction pipeline. With an excellent wind resource and proximity to coastal load centers, the US Bureau of Ocean Energy Management offshore lease areas off the Atlantic coast and in the Gulf of Mexico have the potential to accommodate thousands of wind turbines and to significantly contribute to national electricity supply (Pryor et al. (2021): Wind power production from very large offshore wind farms. Joule 5 2663-2886) with both fixed bottom and floating wind turbines.

Extreme sustained and gust wind speeds are key drivers of wind turbine structural loading, and thus influence the design and cost of wind turbines and support structures. Current annual maximum hub-height wind speeds range from 22-37 ms^-1 at the East-Coast lease areas, and are caused by cyclones with diverse origins and storm tracks (Barthelmie et al. (2021): Extreme wind and waves in US east coast offshore wind energy lease areas. Energies 14 1053). Future projections of extreme wind speeds are highly uncertain, and are dependent on the types of cyclones which occur in the future climate and how those cyclones may be affected by both external and internal climate forcing (Coburn et al., Northeastern Windstorms and Midlatitude Cyclones in the MPI Large Ensemble, this conference). Estimates of extreme winds offshore are affected by air-sea interactions (mass, heat and momentum exchange) and can be improved by models with explicit and reciprocal ocean-atmosphere coupling (Thompson et al., A comparison of how Hurricane Sandy could have impacted offshore wind turbines based on WRF-only and COAWST model simulations, this conference).

In the current work, output from seven Weather Research and Forecasting (WRF) simulations covering the CORDEX North American domain and nested in three CMIP5-generation Earth System Models (ESM) as well as ERA-Interim, are used to characterize the historical, contemporary and future extreme wind speed and cyclone climate of seven offshore lease areas. Each lease area is identified by the nearest US state: Massachusetts (MA), New York (NY), New Jersey (NJ), Virginia (VA), North Carolina (NC) (East Coast areas), Louisiana (LA) and Texas (TX) (Gulf of Mexico areas). The ESMs used are from the Geophysical Fluid Dynamics Laboratory (GFDL), the Hadley Centre (Global Environment Model version 2; HadGEM2) and the Max Planck Institute (MPI). They utilize the Representative Climate Pathway (RCP) 8.5, a ‘business as usual’ scenario for future emissions. The WRF simulations, are performed using a grid spacing 0.22° (~25 km), and are subject to nudging at the top of the simulation domain (Mearns, L.O., et al., 2017: The NA-CORDEX dataset, version 1.0. NCAR Climate Data Gateway, Boulder CO, https://doi.org/10.5065/D6SJ1JCH). The NA-CORDEX archive makes 10-m wind speed and sea level pressure output available every 3 hours.

WRF output at the seven lease-area centroids are used to identify and characterize the annual maximum wind speed for 30 year periods in each simulation; 1981 – 2010 for ERA-Interim, 1976 – 2005 for the three historical ESM simulations, and 2070-2099 for the three future ESM simulations. A logarithmic shear profile with a surface roughness of 0.5 mm (corresponding to ‘blown sea’) is assumed when estimating hub height wind speeds. These annual maximum wind speeds are fitted to Generalized Extreme Value distributions and used to infer 50-year return period wind speeds for each WRF simulation and each lease area. Cyclones associated with each annual maximum wind speed in each lease area are also identified and are tracked using spatially-smoothed sea-level pressure fields and classified by their location of origin.

The goals of the current work are to: (1) Assess the similarity of annual maximum wind speeds and the inferred 50-year return period (extreme) wind speeds and cyclone types responsible for the annual maxima in the three WRF simulations nested within the ESM historical simulations to those from the WRF simulation within ERA-Interim. (2) Quantify possible changes in 50-year return period wind speeds, the distribution of annual maximum wind speeds and their associated cyclones in the WRF simulations nested in the ESM projections. (3) Characterize the magnitude and uncertainty of future extreme wind speeds relevant to wind turbine loading based on (1) and (2).

Initial analyses show that annual maximum wind events at all seven locations across all seven simulations occur overwhelmingly during September to April. Co-occurrence of annual maximum wind speeds among the lease areas located in the Gulf or along the US east coast is common. The 150 events identified in each simulation for the East-Coast lease areas (30 years × 5 lease areas) occur on 104 to 122 unique days indicating that approximately 20% of cyclones identified as responsible for generating the annual maximum wind speeds influence more than one lease area. Cyclone tracks associated with annual maximum wind speeds at the lease areas nearly always originate south of 49° latitude, meaning Alberta Clippers are uncommon as a source of extreme wind speeds in the lease areas off the east coast. However, somewhat surprisingly, a few southward-tracking cyclones affect the TX lease area which originate in western Canada in the lee of the Rocky Mountains in the WRF simulation performed in the future HadGEM2 lateral boundary conditions. 50-year return period winds (U_RP50) at hub height vary between 28 and 41 ms^-1, with the lowest U_RP50 occurring at the Gulf of Mexico lease areas across all simulations. These extreme wind speeds are similar between historical and future runs. The highest U_RP50 (41 ms^-1) occurs at the NC area in the ERA-Interim-nested simulation. All other U_RP50 values are smaller than the estimate derived using the International Electrotechnical Commission (IEC) standard approximation of five times the long-term mean wind speed (IEC, 2005: IEC 61400-1 third edition 2005-08 Wind turbines – Part 1: Design requirements. International Electrotechnical Commission, 177. Geneva, Switzerland).