Wind Farm Layout and Operational Optimization

Wind Farm Layout Optimization

Positioning and layout of wind turbines inside a wind farm require some pretty cutting edge science like Computational Dynamics Modeling and Algorithms. While it sounds like a mouth full, these modeling and algorithms are critical for energy generation for a wind farm. In a wind farm only the front-row turbines get most of the blowing power of the wind, therefore generate the maximum energy, whereas the rest of the wind turbines inside the wind farm produce much less than the front-row (on the order of up to 70% less). This is because every wind turbines generate turbulent wakes downwind with slower winds , basically high turbulence results in lower power generation; Take another good look at the background, you can see the wake clouds from the front-row turbines are covering the back-row turbines, now that is the turbulent wake. Placing the turbines in a optimized "smart" layout, to minimize wake losses and maximize energy production, is therefore very important in wind farms. For a given area and a given number of turbines, there are hundreds of thousand possible layouts, depending on the spacing along and across the prevailing wind direction, the staggering of rows, and the angle between rows and columns; oh did we mention the all crazy complex terrain variation?? Identifying the optimal layout is really tough, because it requires today's most sophisticated computer simulations, called LES (Large-Eddy Simulation), which can take up to 45 days to determine the power generation of a single layout under a single wind direction. To make these analysis and simulation commercially relevant, one needs to analyze 100,000 possible layouts, use LES to do this analysis would take multiple lifetimes. Now, this is where HKF GoodWind comes in, we have developed a super duper computer software, which can identify multiple optimal wind farm layouts with accuracy comparable to that of today's most sophisticated LES models but much faster, instead of multiple lifetimes, we can do it in just a few hours, yessss!

Atmosphere Simulation With Heavy Computer Model

Atmosphere simulation without wind turbines

  • Large-Eddy Simulations with WiTTS (Wind Turbine and Turbulence Simulator);
  • Finite-difference (4th order in space, 2nd in time);
  • SGS model: Lagrangian scale-dependent;
  • Incompressible, Boussinesq, all stabilities.

Atmosphere simulation with one wind turbines

  • Actuator lines;
  • Deficit maximum for x<5D, reduced by 10D but not gone

Atmosphere simulation with a wind farm

  • Large-Eddy Simulations with SOWFA (Software for Offshore/onshore Wind Farm Applications);
  • SOWFA developed at NREL;
  • Actuator lines;
  • Finite-volume, C++;
  • SGS model: Lagrangian scale-invariant;
  • Incompressible, Boussinesq, all stabilities.

Atmosphere simulation with a wind farm

  • Lillgrund wind farm (Denmark) simulation; Neutral stability;
  • 48 Siemens 2.3 MW wind turbines;
  • Spacing 3D x 4D;
  • 80 million grid cells;
  • Resolution 3.5-7m;
  • Complex initialization method to provide non-periodic boundary conditions;
  • Complex sensitivity to:
  • - Layout;
  • - Stability.

Layout Optimization For Increased Power Output

Layout effect: Original vs. staggered

Staggered layout better for wind from 225°

WIND FROM SOUTH-WEST

ORIGINAL
  • Power: 695 kW
  • Losses: 35.5%
  • CF: 0.3
STAGGERED
  • Power: 787 kW
  • Losses: 19.9%
  • CF: 0.34

Energy Generation Trade-offs, variable wind directions

WIND FROM SOUTH-WEST
  • Power: 695 kW
  • Losses: 35.5%
  • CF: 0.3
  • Power: 787 kW
  • Losses: 19.9%
  • CF: 0.34
WIND FROM NORTH-WEST
  • Power: 964 kW
  • Losses: 10.1%
  • CF: 0.42
  • Power: 939 kW
  • Losses: 12.4%
  • CF: 0.41

Atmospheric stability effects

  • Initialized with same prescribed wind speed at 90 m (9 m/s);
  • Neutral and stable case have reached equilibrium;
  • Unstable case shows patterns of convergence and divergence;
  • Wakes shorter in unstable, longer in stable, than neutral case.

Vertical Profiles

Our Technology Is Much Faster

Effect of stability still unclear...therefore the opportunity

Our results are awesome!

Results at Lillgrund

Results at Horns Rev

LES data at Horns Rev from Porte-Agel et al. (Energies, 2014)

Did we say awesome? Yesss! Because we can assess all layouts combinations!
Oh and complex terrain analysis capability is coming right up!

Hey Sport! Hello to you too!

HKF Good Wind LLC
2711 CENTERVILLE RD SUITE 400, WILMINGTON, DE, 19808

contact@hkftechnology.com