Taming hurricanes with arrays of o shore wind turbines

Hurricanes are causing increasing damage to many coastal regions worldwide1,2. O shore wind turbines can provide substantial clean electricity year-round, but can they alsomitigate hurricane damage while avoiding damage to themselves? This study uses an advanced climate–weather computer model that correctly treats the energy extraction of wind turbines3,4 to examine this question. It finds that large turbine arrays (300+ GW installed capacity) may diminish peak near-surface hurricane wind speeds by 25–41ms−1 (56–92 mph) and storm surge by 6–79%. Benefits occur whether turbine arrays are placed immediately upstream of a city or along an expanse of coastline. The reduction in wind speed due to large arrays increases the probability of survival of even present turbine designs. The net cost of turbine arrays (capital plus operation cost less cost reduction from electricity generation and from health, climate, and hurricane damage avoidance) is estimated to be less than today’s fossil fuel electricity generation net cost in these regions and less than the net cost of sea walls used solely to avoid storm surge damage. Hurricane damage is increasing with expanding coastal development1 and rising sea levels2. Increasing temperatures may also increase hurricane intensity, but it is uncertain whether hurricane intensity changes so far have exceeded natural variability5. Continuing a long-term problem of hurricane damage, Hurricane Sandy in 2012 caused ∼$82 billion in damage to three US states6 and 253 fatalities in seven countries. Hurricane Katrina destroyed much of New Orleans, Louisiana. Following Hurricane Sandy, sea walls were proposed to protect cities from hurricane storm surge. Such walls might cost $10–$29 billion for one city7, protect the areas only right behind the walls, and limit the access of populations to coastal zones. Large arrays of wind-wave pumps, which bring deep, cool water to the surface have also been proposed to reduce hurricane intensity8. This technology also serves one purpose. This study quantitatively tests whether large arrays of wind turbines installed offshore in front of major cities and along key coastal areas can extract sufficient kinetic energy from hurricane winds to reduce wind speed and storm surge, thus preventing damage to coastal structures as well as to the offshore turbines themselves. Unlike sea walls, offshore wind turbines would reduce both wind speed and storm surge and would generate electricity year-round. The hypothesis is tested here through numerical simulationswith GATOR–GCMOM, a global-through-local climate–weather–airpollution–ocean forecast model3,4 (Supplementary Information). The model extracts the correct amount of energy from the wind at different model heights intersecting the turbine rotor3 given the instantaneous model wind speed, which is affected by turbulence and shear due to the hurricane and turbine itself (Supplementary Section 1.H). Several three-dimensional computer simulations without and with wind turbines were run for hurricanes Katrina and Isaac (US Gulf Coast) and Sandy (US East Coast; Methods and Table 1). Figure 1 compares modelled with observed storm tracks and peak near-surface wind speeds for hurricanes Katrina, Sandy and Isaac. Model results include those from GATOR–GCMOM and two operational hurricane models (Geophysical Fluid Dynamic Laboratory (GFDL) and Hurricane Weather Research and Forecasting (HWRF)). TheGATOR–GCMOMmodelled tracks followed observed tracks, particularly for Katrina and Isaac. For Katrina, GATOR–GCMOM-modelled peak wind speeds and their rates of change with time were similar to observed peak winds and slightly more accurate than those from the GFDL model. For Sandy, GATOR–GCMOM-modelled peak winds slightly exceeded observed values, but ‘its results are comparable with those of the other operational and semi-operational models (T. Marchok, NOAA/GFDL, personal communication). Supplementary Fig. 6 shows results from a case where turbine arrayswere added offshore ofCuba and fromFlorida toTexas during Hurricane Katrina (Simulation A). Such arrays, in comparison with the base-case simulation of Katrina without turbines, reduced wind speeds by up to 41ms (92 mph) at 15 m height and by up to 80ms (179 mph) at the 100 m hub height typical of an offshore wind turbine while producing 1.1 TW of power. For Simulation E, where turbine arrays were placed along most of the East Coast during Hurricane Sandy, turbines reduced 15m wind speeds up to 39ms while extracting up to 2.65 TW (Table 1, Supplementary Fig. 7). The greater modelled power extraction during Sandy was due to the larger radius of hurricane-force winds, a factor also observed9. Figure 2 and Supplementary Fig. 8 show the two-dimensional time evolution of several parameters with and without turbines just to the southeast of New Orleans during Hurricane Katrina (Simulation D). Without turbines, the strongest winds when the hurricane was close to landfall were on the eastern side of the core (Supplementary Fig. 8), consistent with observations.With turbines, the hurricane dissipated faster. Comparisons of Simulation D with A for Katrina and of Simulation H with E for Sandy indicate that wind speed reductions and power extraction were similar in the overlapping regions of turbines whether the turbines were just upstream of the region of interest or along much of the coast (Supplementary Fig. 9). Thus,