floris.optimization.load_optimization.load_optimization

Module for the load optimization class and functions.

Functions

compute_farm_revenue(fmodel)	Compute the farm revenue of the FlorisModel object using the values from fmodel.wind_data.
compute_farm_voc(fmodel, A, ambient_lti[, ...])	Compute the farm-total Variable Operating Cost (VOC) for each findex.
compute_lti(fmodel, ambient_lti[, ...])	Compute the turbine 'load turbulence intensity' (lti) for the current layout.
compute_net_revenue(fmodel, A, ambient_lti)	Compute the net revenue for the current layout as the difference between the farm revenue the farm VOC for each index.
compute_turbine_voc(fmodel, A, ambient_lti)	Compute the turbine Variable Operating Cost (VOC) for each findex and turbine.
find_A_to_satisfy_rev_voc_ratio(fmodel, ...)	Find the value of A that satisfies the target ratio of total farm revenue over all findices to total farm VOC over all findices.
find_A_to_satisfy_target_VOC_per_MW(fmodel, ...)	Find the value of A that satisfies the target average cost per total farm power per findex over all findices.
optimize_power_setpoints(fmodel, A, ambient_lti)	Optimize the derating of each turbine to maximize net revenue sequentially from upstream to downstream.

floris.optimization.load_optimization.load_optimization.compute_lti(fmodel, ambient_lti, wake_slope=0.3, max_dist_D=10.0)

Compute the turbine 'load turbulence intensity' (lti) for the current layout.

LTI represents the turbulence intensity used in load calculations and follows the method of computing wake added turbulence described in Annex E of the IEC 61400-1 Ed. 4 standard. In principle this can be the same as the turbulence models used in the wake velocity and deflection models within FLORIS, but for consistency with the IEC standard is computed separately here.

Args:

fmodel (FlorisModel): FlorisModel object ambient_lti (list or np.array): Ambient 'load' turbulence intensity (lti) for each findex wake_slope (float, optional): Wake slope, defined as the lateral expansion of the wake on

each side per unit downstream distance along the axial direction. Defaults to 0.3.

max_dist_D (flat, optional): Maximum distance downstream of a turbine beyond which wake

added turbulence is assumed to be zero, in rotor diameters. Defaults to 10.0 (see IEC 61400-1 Ed. 4 Annex E).

Returns:

np.array: Array of load turbulence intensity for each findex and turbine

Parameters:

• fmodel (FlorisModel)

• ambient_lti (array)

• wake_slope (float)

• max_dist_D (float)

floris.optimization.load_optimization.load_optimization.compute_turbine_voc(fmodel, A, ambient_lti, wake_slope=0.3, max_dist_D=10.0, exp_ws_std=1.0, exp_thrust=1.0)

Compute the turbine Variable Operating Cost (VOC) for each findex and turbine.

Variable Operating Cost (VOC) is meant to represent the cost of operating a turbine at a particular rating in particular conditions. We envision in the future there can be several possible functions to determine VOC for a turbine, but for now we use a simple model that is proportional to the wind speed standard deviation and the absolute thrust of the turbine.

Args:

fmodel (FlorisModel): FlorisModel object A (float): Coefficient for the VOC calculation ambient_lti (list or np.array): Ambient 'load' turbulence intensity for each findex. wake_slope (float, optional): Wake slope, defined as the lateral expansion of the wake on

each side per unit downstream distance along the axial direction. Defaults to 0.3.

max_dist_D (flat, optional): Maximum distance downstream of a turbine beyond which wake

added turbulence is assumed to be zero, in rotor diameters. Defaults to 10.0 (see IEC 61400-1 Ed. 4 Annex E).

exp_ws_std (float, optional): Exponent for the wind speed standard deviation.

Defaults to 1.0.

exp_thrust (float, optional): Exponent for the thrust. Defaults to 1.0.

Returns:

np.array: Array of VOC for each findex and turbine

Parameters:

• fmodel (FlorisModel)

• A (float)

• ambient_lti (array)

• wake_slope (float)

• max_dist_D (float)

• exp_ws_std (float)

• exp_thrust (float)

floris.optimization.load_optimization.load_optimization.compute_farm_voc(fmodel, A, ambient_lti, wake_slope=0.3, max_dist_D=10.0, exp_ws_std=1.0, exp_thrust=1.0)

Compute the farm-total Variable Operating Cost (VOC) for each findex.

Variable Operating Cost (VOC) is meant to represent the cost of operating a turbine at a particular rating in particular conditions. We envision in the future there can be several possible functions to determine VOC for a turbine, but for now we use a simple model that is proportional to the wind speed standard deviation and the absolute thrust of the turbine. The farm-total VOC is the sum of the VOC for each turbine in the farm.

Args:

fmodel (FlorisModel): FlorisModel object A (float): Coefficient for the VOC calculation ambient_lti (list or np.array): Ambient 'load' turbulence intensity for each findex,

expressed as fractions of mean wind speed

wake_slope (float, optional): Wake slope, defined as the lateral expansion of the wake on

each side per unit downstream distance along the axial direction. Defaults to 0.3.

max_dist_D (flat, optional): Maximum distance downstream of a turbine beyond which wake

added turbulence is assumed to be zero, in rotor diameters. Defaults to 10.0 (see IEC 61400-1 Ed. 4 Annex E).

exp_ws_std (float, optional): Exponent for the wind speed standard deviation.

Defaults to 1.0.

exp_thrust (float, optional): Exponent for the thrust. Defaults to 1.0.

Returns:

np.array: Array of farm VOC for each findex

Parameters:

• fmodel (FlorisModel)

• A (float)

• ambient_lti (array)

• wake_slope (float)

• max_dist_D (float)

• exp_ws_std (float)

• exp_thrust (float)

floris.optimization.load_optimization.load_optimization.compute_farm_revenue(fmodel)

Compute the farm revenue of the FlorisModel object using the values from fmodel.wind_data.

Args:

fmodel (FlorisModel): FlorisModel object

Returns:

np.array: Array of farm revenue for each findex

Parameters:

fmodel (FlorisModel)

floris.optimization.load_optimization.load_optimization.compute_net_revenue(fmodel, A, ambient_lti, wake_slope=0.3, max_dist_D=10.0, exp_ws_std=1.0, exp_thrust=1.0)

Compute the net revenue for the current layout as the difference between the farm revenue the farm VOC for each index.

Args:

fmodel (FlorisModel): FlorisModel object A (float): Coefficient for the VOC calculation ambient_lti (list or np.array): Ambient 'load' turbulence intensity for each findex,

expressed as fractions of mean wind speed

wake_slope (float, optional): Wake slope, defined as the lateral expansion of the wake on

each side per unit downstream distance along the axial direction. Defaults to 0.3.

max_dist_D (flat, optional): Maximum distance downstream of a turbine beyond which wake

added turbulence is assumed to be zero, in rotor diameters. Defaults to 10.0 (see IEC 61400-1 Ed. 4 Annex E).

exp_ws_std (float, optional): Exponent for the wind speed standard deviation.

Defaults to 1.0.

exp_thrust (float, optional): Exponent for the thrust. Defaults to 1.0.

Returns:

np.array: Array of net revenue for each findex

Parameters:

• fmodel (FlorisModel)

• A (float)

• ambient_lti (array)

• wake_slope (float)

• max_dist_D (float)

• exp_ws_std (float)

• exp_thrust (float)

floris.optimization.load_optimization.load_optimization.find_A_to_satisfy_rev_voc_ratio(fmodel, target_rev_voc_ratio, ambient_lti, wake_slope=0.3, max_dist_D=10.0, exp_ws_std=1.0, exp_thrust=1.0)

Find the value of A that satisfies the target ratio of total farm revenue over all findices to total farm VOC over all findices.

Args:

fmodel (FlorisModel): FlorisModel object target_rev_voc_ratio (float): Target revenue to VOC ratio ambient_lti (list or np.array): Ambient 'load' turbulence intensity for each findex,

expressed as fractions of mean wind speed

wake_slope (float, optional): Wake slope, defined as the lateral expansion of the wake on

each side per unit downstream distance along the axial direction. Defaults to 0.3.

max_dist_D (flat, optional): Maximum distance downstream of a turbine beyond which wake

added turbulence is assumed to be zero, in rotor diameters. Defaults to 10.0 (see IEC 61400-1 Ed. 4 Annex E).

exp_ws_std (float, optional): Exponent for the wind speed standard deviation.

Defaults to 1.0.

exp_thrust (float, optional): Exponent for the thrust. Defaults to 1.

Returns:

float: Value of A that satisfies the target revenue to VOC ratio

Parameters:

• fmodel (FlorisModel)

• target_rev_voc_ratio (float)

• ambient_lti (array)

• wake_slope (float)

• max_dist_D (float)

• exp_ws_std (float)

• exp_thrust (float)

floris.optimization.load_optimization.load_optimization.find_A_to_satisfy_target_VOC_per_MW(fmodel, target_VOC_per_MW_findex, ambient_lti, wake_slope=0.3, max_dist_D=10.0, exp_ws_std=1.0, exp_thrust=1.0)

Find the value of A that satisfies the target average cost per total farm power per findex over all findices. Note that if each findex represents 1 hour of operation, this is equivalent to the target average cost/MWh.

Args:

fmodel (FlorisModel): FlorisModel object target_VOC_per_MW_findex (float): Target average cost per MW per findex ambient_lti (list or np.array): Ambient 'load' turbulence intensity for each findex,

expressed as fractions of mean wind speed

wake_slope (float, optional): Wake slope, defined as the lateral expansion of the wake on

each side per unit downstream distance along the axial direction. Defaults to 0.3.

max_dist_D (flat, optional): Maximum distance downstream of a turbine beyond which wake

added turbulence is assumed to be zero, in rotor diameters. Defaults to 10.0 (see IEC 61400-1 Ed. 4 Annex E).

exp_ws_std (float, optional): Exponent for the wind speed standard deviation.

Defaults to 1.0.

exp_thrust (float, optional): Exponent for the thrust. Defaults to 1.

Returns:

float: Value of A that satisfies the target cost/MW/findex

Parameters:

• fmodel (FlorisModel)

• target_VOC_per_MW_findex (float)

• ambient_lti (array)

• wake_slope (float)

• max_dist_D (float)

• exp_ws_std (float)

• exp_thrust (float)

floris.optimization.load_optimization.load_optimization.optimize_power_setpoints(fmodel, A, ambient_lti, wake_slope=0.3, max_dist_D=10.0, exp_ws_std=1.0, exp_thrust=1.0, power_setpoint_initial=None, power_setpoint_levels=array([1.0e+12, 7.5e+11, 5.0e+11, 2.5e+11, 1.0e-03]))

Optimize the derating of each turbine to maximize net revenue sequentially from upstream to downstream.

Args:

fmodel (FlorisModel): FlorisModel object A (float): Coefficient for the VOC calculation ambient_lti (list or np.array): Ambient 'load' turbulence intensity for each findex,

expressed as fractions of mean wind speed

wake_slope (float, optional): Wake slope, defined as the lateral expansion of the wake on

each side per unit downstream distance along the axial direction. Defaults to 0.3.

max_dist_D (flat, optional): Maximum distance downstream of a turbine beyond which wake

added turbulence is assumed to be zero, in rotor diameters. Defaults to 10.0 (see IEC 61400-1 Ed. 4 Annex E).

exp_ws_std (float, optional): Exponent for the wind speed standard deviation.

Defaults to 1.0.

exp_thrust (float, optional): Exponent for the thrust. Defaults to 1. power_setpoint_initial (np.array, optional): Initial power setpoint for each turbine.

If None, each turbine's rated power will be used. Defaults to None.

power_setpoint_levels (np.array, optional): Array of power setpoint levels to consider

in optimization in W. Defaults to np.linspace(POWER_SETPOINT_DEFAULT, POWER_SETPOINT_DISABLED, 5).

Parameters:

• fmodel (FlorisModel)

• A (float)

• ambient_lti (array)

• wake_slope (float)

• max_dist_D (float)

• exp_ws_std (float)

• exp_thrust (float)

• power_setpoint_initial (array)

• power_setpoint_levels (array)