Wind Farm Components

Wind Farm Components#

Hercules provides four wind farm simulation components that differ in their approach to wake modeling and data sources. The first three components support both simple filter-based turbine models and 1-degree-of-freedom (1-DOF) turbine dynamics, while the fourth component uses SCADA power data directly.

Overview#

The WindFarm component applies wind speed time signals to turbine models to simulate wind farm behavior over extended periods. This is available with different methods for how wakes are applies, as described below. The WindFarmSCADAPower component uses a fundamentally different approach by using actual SCADA power measurements as input.

WindFarm (with Dynamic wake method)#

WindFarm is a comprehensive wind farm simulator. When wake_method="dynamic" (the default), WindFarm computes wake effects dynamically at each time step (or at intervals specified by floris_update_time_s). It focuses on meso-scale phenomena by applying a separate wind speed time signal to each turbine model derived from data. This model combines FLORIS wake modeling with detailed turbine dynamics for long-term wind farm performance analysis.

Use this model when:

  • Turbines have individual power setpoints or non-uniform operation

  • Precise wake modeling is required for each control action

  • Turbines may be partially derated or individually controlled

WindFarm (with Precomputed wake method)#

WindFarm with wake_method="precomputed" is an optimized variant that pre-computes all FLORIS wake deficits at initialization for improved simulation speed. This approach provides significant speed improvements while conservatively assuming wakes are always based on nominal operation.

Use this model when:

  • Not investigating wakes of derated turbines or wake losses can be conservatively estimated.

WindFarm (with No Added Wakes method)#

Using WindFarm with wake_method="no_added_wakes" assumes that wake effects are already included in the input wind data and performs no wake modeling during simulation. This model is appropriate for using SCADA data of operational farm since wake losses already included in data.

WindFarmSCADAPower (SCADA Power Data)#

WindFarmSCADAPower uses SCADA power measurements directly rather than computing power from wind speeds and turbine models. This component applies a filter to the SCADA power data to simulate turbine response dynamics and respects power setpoint constraints.

This model is a beta feature and is not yet fully tested.

Configuration#

Common Required Parameters#

Required parameters for both components in h_dict (see timing for time-related parameters):

  • floris_input_file: FLORIS farm configuration

  • wind_input_filename: Wind resource data file

WindFarm Specific Parameters#

Required parameters for WindFarm:

  • wake_method: One of "dynamic", "precomputed", or "no_added_wakes" (defaults to "dynamic")

  • floris_update_time_s: How often to update FLORIS (the last floris_update_time_s seconds are averaged as input). Required for "dynamic" and "precomputed" wake methods; for "no_added_wakes", this parameter is not required and ignored if provided.

  • turbine_file_name: Turbine model configuration

  • log_channels: List of output channels to log. See Logging Configuration section below for details.

WindFarmSCADAPower Specific Parameters#

Required parameters for WindFarmSCADAPower:

  • scada_filename: Path to SCADA data file (CSV, pickle, or feather format)

  • turbine_file_name: Turbine model configuration (for filter parameters)

  • log_channels: List of output channels to log. See Logging Configuration section below for details.

SCADA File Format:

The SCADA file must contain the following columns:

  • time_utc: Timestamps in UTC (ISO 8601 format or parseable datetime strings)

  • wd_mean: Mean wind direction in degrees

  • pow_###: Power output for each turbine (e.g., pow_000, pow_001, pow_002)

Optional columns:

  • ws_###: Wind speed for each turbine (e.g., ws_000, ws_001, ws_002)

  • ws_mean: Mean wind speed (used if individual turbine speeds not provided)

  • ti_###: Turbulence intensity for each turbine (defaults to 0.08 if not provided)

The number of turbines and rated power are automatically inferred from the SCADA data.

Turbine Models#

Note: WindFarmSCADAPower does not use a filter model as power values come directly from SCADA data rather than being computed from wind speedes.

Filter Model#

Simple first-order filter for power output smoothing with configurable time constants.

1-DOF Model#

Advanced model with rotor dynamics, pitch control, and generator torque control. Not applicable to WindFarmSCADAPower.

Outputs#

Common Outputs#

All four components provide these outputs in the h_dict at each simulation step:

  • power: Total wind farm power (kW)

  • turbine_powers: Individual turbine power outputs (array, kW)

  • turbine_power_setpoints: Current power setpoint values (array, kW)

  • wind_speed_mean_background: Farm-average background wind speed (m/s)

  • wind_speed_mean_withwakes: Farm-average with-wakes wind speed (m/s)

  • wind_direction_mean: Farm-average wind direction (degrees)

  • wind_speeds_background: Per-turbine background wind speeds (array, m/s)

  • wind_speeds_withwakes: Per-turbine with-wakes wind speeds (array, m/s)

Note for WindFarm with no_added_wakes and WindFarmSCADAPower: In these models (no wake modeling), wind_speeds_withwakes equals wind_speeds_background and wind_speed_mean_withwakes equals wind_speed_mean_background.

Logging Configuration#

The log_channels parameter controls which outputs are written to the HDF5 output file. This is a list of channel names. The power channel is always logged, even if not explicitly specified.

Available Channels#

Scalar Channels:

  • power: Total wind farm power output (kW)

  • wind_speed_mean_background: Farm-average background wind speed (m/s)

  • wind_speed_mean_withwakes: Farm-average with-wakes wind speed (m/s)

  • wind_direction_mean: Farm-average wind direction (degrees)

Array Channels:

  • turbine_powers: Power output for all turbines (creates datasets like wind_farm.turbine_powers.000, wind_farm.turbine_powers.001, etc.)

  • turbine_power_setpoints: Power setpoints for all turbines

  • wind_speeds_background: Background wind speeds for all turbines

  • wind_speeds_withwakes: With-wakes wind speeds for all turbines

Selective Array Element Logging#

For large wind farms, logging all turbine data can significantly increase file size and slow down the simulation. You can log specific turbine indices by appending a 3-digit turbine index to the channel name:

# Log only turbines 0, 5, and 10
log_channels:
  - power
  - wind_speed_mean_background
  - wind_speed_mean_withwakes
  - wind_direction_mean
  - turbine_powers.000
  - turbine_powers.005
  - turbine_powers.010

Example Configurations#

Minimal Logging:

log_channels:
  - power
  - wind_speed_mean_background
  - wind_speed_mean_withwakes
  - wind_direction_mean

Detailed Logging (all turbines):

log_channels:
  - power
  - wind_speed_mean_background
  - wind_speed_mean_withwakes
  - wind_direction_mean
  - turbine_powers
  - wind_speeds_withwakes

Selected Turbine Logging:

# Log first 3 turbines only
log_channels:
  - power
  - wind_speed_mean_background
  - wind_speed_mean_withwakes
  - wind_direction_mean
  - turbine_powers.000
  - turbine_powers.001
  - turbine_powers.002
Contents