By Gordon Stewart

Loads Analysis of Floating Offshore Wind Turbines

2

Offshore Wind Resources

3

Floating Offshore Wind Turbines (FOWTs)

Can be placed over deep (>50m) water, allows for placement far from shore out of sight

Possibility of being assembled in port and towed out

Increased costs associated with support platform, operation and maintenance, and power transmission

4

Research Questions

What are the metrological/ocean conditions that floating wind turbines will experience in the U.S.?

How long do simulations of floating wind turbines need to be to be accurate?

What are the effects of wind/wave misalignment on floating wind turbines?

Can the shapes of offshore wind farms be optimized to reduce avian impacts?

5

Metocean Conditions

5 6 7 8 9 100.5

1

1.5

2

2.5

3

Mean Wind Speed at Hub Height (m/s)

Mean S

ignifi

cant W

ave

Heig

ht (m

)

SE NANTUCKET

LONG ISLAND

SOUTH SANTA ROSA

CAPE SAN MARTIN

BERMUDA

CAPE ELIZABETH

ONSLOW BAY

TILLAMOOK

MONTEREY

FRYING PAN SHOALSW GULFBAY OF CAMPECHE

MID GULF

TANNER BANK

SANTA MARIA

NE BAHAMAS

GALVESTON

E GULF

CORPUS CHRISTI

LUKE OFFSHOREPENSACOLA

W TAMPA

SAN CLEMENTE BASIN

East Coast

West Coast

Gulf of Mexico

• Data downloaded from the National Data Buoy Center from NOAA.

• Sites must have wind speed, wind direction, significant wave height, wave peak spectral period, and wave direction

• Over 200 years total of hourly data; created a representative site for East Coast, Gulf of Mexico, and West Coast.

This data will be published in a open source format for future use

6

Wind/Wave Misalignment

While wind and wave directions are correlated, all ocean sites experience some degree of misalignment

Greater compliance of a floating platform compared to fixed-bottom substructures can lead to more wave-induced motion in FOWT

This, coupled with the lack of side-to-side aerodynamic damping, may cause increased loading

7

Fatigue Loads – Misalignment

Edgewise Flapwise Tower SS Tower FA Anchor Tension0

0.2

0.4

0.6

0.8

1

1.2

1.4DELs for various loads normalized by all misalignment bins (All simulations)

24 bins

12 bins

8 bins

4 bins

2 bins

1 bin

Increasing the bin size of the misalignment bins (equivalent to reducing the number of bins) has very little effect on loads, except for 1 direction

Using only the 0˚ bin drastically under-predicts side-to-side tower loads and over-predicts tower fore-aft loads

8

Avian Micro-siting Optimization

Masden et al. (2012) used radar data from the Nysted offshore wind farm to create a model that simulates trajectories of flocks of common eider

9

Preliminary Results from Genetic Algorithm

For a proof-of-concept, a genetic algorithm is used to optimize the placement of 100 wind turbines in the farm.

30 random wind farms are used for an initial population, and the GA is allowed to run for 50 iteration.

This movie shows the best solution at each iteration.