CN107810323B

CN107810323B - Method and system for generating a wind turbine control arrangement

Info

Publication number: CN107810323B
Application number: CN201680038982.0A
Authority: CN
Inventors: C·斯普鲁斯; C·比雷迪
Original assignee: Vestas Wind Systems AS
Current assignee: Vestas Wind Systems AS
Priority date: 2015-06-30
Filing date: 2016-06-21
Publication date: 2020-04-07
Anticipated expiration: 2036-06-21
Also published as: CN107810323A; WO2017000950A1

Abstract

There is provided a method of generating a control schedule for a wind turbine, the control schedule indicating how the turbine maximum power level varies over time, the method comprising: determining a value indicative of a current remaining fatigue life of the turbine or one or more turbine components based on measured wind turbine site and/or operational data; applying an optimization function that changes an initial control schedule to determine an optimized control schedule by changing a tradeoff between energy capture and fatigue life consumed by the turbine or the one or more turbine components until an optimized control schedule is determined, the optimization comprising: estimating a future fatigue life consumed by the turbine or turbine component for a duration of the altered control schedule based on the current remaining fatigue life and the altered control schedule; and constraining optimization of the control schedule in accordance with one or more input constraints; wherein the optimizing further comprises changing an initial value of wind turbine life and changing an initial value of a number of component replacements to be performed on the one or more components over the course of the schedule to determine a combination of the number of component replacements for the one or more turbine components and the target minimum wind turbine life.

Description

Method and system for generating a wind turbine control arrangement

Technical Field

Embodiments of the invention relate to methods and systems for determining a control schedule (controlschedule) for a wind turbine power output.

Background

Fig. 1A shows a large conventional wind turbine 1 as known in the art, comprising a tower 10 and a wind turbine nacelle 20 positioned on top of the tower 10. The wind turbine rotor 30 comprises three wind turbine blades 32, each having a length L. The wind turbine rotor 30 may include other numbers of blades 32, such as one, two, four, five, or more. The blades 32 are mounted on a hub 34 at a height H above the bottom of the tower. The hub 34 is connected to the nacelle 20 by a low speed shaft (not shown) extending from the front of the nacelle 20. The low speed shaft drives a gearbox (not shown) that speeds up the rotational speed and, in turn, drives a generator within the nacelle 20 to convert the energy extracted from the wind by the rotating blades 32 into an electrical power output. The wind turbine blades 32 define a swept area a, which is a circular area delineated by the rotating blades 32. The swept area indicates how much of a given air mass is intercepted by the wind turbine 1 and therefore affects the power output of the wind turbine 1 and the forces and bending moments experienced by the components of the turbine 1 during operation. As shown, the turbine may be located onshore or offshore. In the latter case, the tower will be connected to a monopile tripod lattice or other foundation structure, and the foundation may be fixed or floating.

For example, each wind turbine has a wind turbine controller, which may be located at the tower base or at the tower top. The wind turbine controller processes inputs from the sensors and other control systems and generates output signals for actuators such as pitch actuators, generator torque controllers, generator contactors, switches for activating shaft brakes, yaw motors, and the like.

FIG. 1B schematically shows an example of a conventional wind power plant 100 comprising a plurality of wind turbines 110, the controller of each of the plurality of wind turbines 110 being in communication with a plant controller PPC 130. The PPC 130 may be in bidirectional communication with each turbine. The turbine outputs power to the grid connection point 140 as indicated by the thick line 150. In operation, and assuming wind conditions permit, each of the wind turbines 110 will output maximum active power up to its rated power as specified by the manufacturer.

Fig. 2 shows a conventional power curve 55 of a wind turbine, plotting wind speed on the x-axis and power output on the y-axis. Curve 55 is the normal power curve for the wind turbine and defines the power output by the wind turbine generator as a function of wind speed. As is known in the art, a wind turbine is cut into wind speed V_minSuccess rate begins to occur. The turbine is then operated under partial load (also called partial load) conditions until point V_RThe rated wind speed is reached. At rated wind speed, the rated (or nominal) generator power is reached and the turbine is operated at full load. For example, the cut-in wind speed in a typical wind turbine may be 3m/s and the rated wind speed may be 12 m/s. Point V_maxIs the cut-out wind speed, which is the highest wind speed at which the wind turbine can operate while delivering power. At wind speeds equal to or higher than the cut-out wind speed, the wind turbine is shut down for safety reasons, in particular to reduce the loads acting on the wind turbine. Alternatively, the power output may gradually drop to zero power as a function of wind speed.

The power rating of a wind turbine is defined in IEC 61400 as the maximum continuous electrical power output that the wind turbine is designed to achieve under normal operating and external conditions. Large commercial wind turbines are typically designed for a 20 to 25 year life and are designed to operate at rated power so that the design loads and fatigue life of the components are not exceeded.

The rate of fatigue damage accumulation for individual components in a wind turbine varies widely under different operating conditions. As the power generated increases, the wear rate, or rate of damage accumulation, tends to increase. Wind conditions also affect the rate of damage accumulation. For some mechanical components, operating in very high turbulence results in a fatigue damage accumulation rate many times higher than operating in normal turbulence. For some electrical components, operation at very high temperatures, which may be caused by high ambient temperatures, results in a fatigue damage accumulation rate (such as the insulation breakdown rate) that is many times higher than that of operation at normal temperatures. As an example, the rule of thumb for the generator winding is that a 10 ℃ drop in winding temperature will increase the lifetime by 100%.

The Annual Energy Production (AEP) of a wind power plant is related to the production rate of the wind turbines forming the wind power plant and typically depends on the annual wind speed at the location of the wind power plant. For a given wind power plant, the larger the AEP, the greater the profit for the operator of the wind power plant and the greater the amount of electrical energy supplied to the grid.

Thus, wind turbine manufacturers and wind power plant operators are always trying to increase the AEP of a given wind power plant.

One such method may be to over-rate the wind turbine under certain conditions, in other words, to allow the wind turbine to operate at power levels up to the rated or nameplate power level of the wind turbine for a period of time (as shown by the shaded area 58 of FIG. 2) in order to generate more electrical energy when the wind is high and thus increase the AEP of the wind power plant. In particular, the term "over-rated" should be understood to mean that power in excess of rated active power is produced during full load operation by controlling turbine parameters such as rotor speed, torque or generator current. An increase in speed demand, torque demand, and/or generator current demand will increase the additional power produced by over-rating, while a decrease in speed demand, torque demand, and/or generator current demand will decrease the additional power produced by over-rating. It should be understood that over-rating applies to active power, not reactive power. When the turbine is over-rated, the turbine operates more aggressively than normal, and the generator has a higher power output than the rated power for a given wind speed. For example, the over-rated power level may be up to 30% above the rated power output. This allows greater power extraction to be achieved when it is beneficial to the operator, especially when external conditions such as wind speed, turbulence and electricity prices allow for more profitable power generation.

Over-rating causes higher wear or fatigue on the components of the wind turbine, which may lead to early failure of one or more components and the need to shut down the turbine for maintenance. Thus, over-rating is characterized by transient behavior. When the turbine is over-rated, it may last as little as a few seconds, or it may last for an extended period of time if wind conditions and fatigue life of the components are favorable for over-rating.

While over-rating allows turbine operators to increase the AEP and otherwise modify the power generation to meet their requirements, there are several problems and drawbacks associated with over-rating wind turbines. Wind turbines are typically designed to operate at a given nominal rated or nameplate power level and for a certified number of years, for example, 20 years or 25 years. Thus, if the wind turbine is over-rated, the life of the wind turbine may be shortened.

The present invention seeks to provide the turbine operator with the flexibility to have their turbine operate in a manner that meets their requirements, for example by returning an optimized AEP.

Disclosure of Invention

The invention is defined in the independent claims, to which reference will now be made. Preferred features are set forth in the dependent claims

Embodiments of the present invention seek to improve the flexibility available to the turbine operator when employing a control approach that compromises energy capture and fatigue loading. An example of such a control method is the use of over-rating.

According to a first aspect of the invention, there is provided a method of generating a control schedule for a wind turbine, the control schedule indicating how the turbine maximum power level varies over time, the method comprising:

receiving an input indicative of a target minimum wind turbine life;

determining a value indicative of a current remaining fatigue life of the wind turbine or one or more turbine components based on the measured wind turbine site and/or operational data;

changing a parameter of an initial predefined control schedule specifying how the turbine maximum power level varies over time by:

i) adjusting parameters of an initial predefined control schedule;

ii) estimating a future fatigue life consumed by the turbine or one or more turbine components for the duration of the altered control schedule based on the altered control schedule;

and

iii) repeating steps (i) and (ii) until the estimated future fatigue life consumed by each of the one or more turbine components or the wind turbine is sufficient to allow the target minimum wind turbine life to be reached.

The parameters may be changed until the estimated future fatigue life consumed by the heaviest loaded components is sufficient to allow just reaching the target minimum wind turbine life, or in other words until the total fatigue life consumed is made substantially the same as the target minimum wind turbine life. This may be achieved based on a predetermined margin of target minimum wind turbine life (e.g., within 0 to 1 month, within 0 to 3 months, within 0 to 6 months, or within 0 to 12 months from the target).

Optionally, step (iii) also requires that energy capture over the life of the turbine is maximised.

Optionally, the control schedule indicates an amount by which the wind turbine may be over-rated to power in excess of its rated power.

Optionally, the method further comprises receiving, for each of one or more of the turbine components, an input indicative of a maximum number of permitted replacements for that turbine component. Step (i) may then further comprise adjusting for one or more of the turbine components the number of times that component may be replaced within the remaining life of the turbine. Step (i) may also include making adjustments to one or more of the turbine components as to when the components may be replaced during the remaining life of the turbine. The one or more turbine components may include one or more of the following: a blade, a pitch bearing, a pitch actuation system, a hub, a main shaft, a main bearing, a gearbox, a generator, a converter, a yaw drive mechanism, a yaw bearing or a transformer.

Optionally, the initial predefined control schedule specifies a relative change in the turbine maximum power level over time.

Optionally, determining a value indicative of a current remaining fatigue life of the turbine or one or more turbine components comprises applying sensor data from one or more turbine sensors to one or more life usage estimation algorithms.

Optionally, determining a value indicative of the current remaining fatigue life of the turbine or one or more turbine components comprises using data from a condition monitoring system.

Optionally, determining a value indicative of the current remaining fatigue life of the turbine or one or more turbine components comprises using data obtained from wind plant sensors in conjunction with a site check program that determines loads acting on the turbine components based on the data obtained from the wind plant sensors and parameters related to the wind plant and wind turbine design. The sensor data may comprise sensor data collected prior to commissioning and/or construction of the wind turbine or wind power plant.

Optionally, adjusting the parameter comprises applying an offset, an amplification, an attenuation (de-amplification) or a gain factor to the control arrangement. The parameters may be adjusted until all or substantially all of the fatigue life of the heaviest loaded component is consumed for the scheduled duration. The offset may be adjusted by equalizing the area of the curve above and below the line shown below: the line indicates fatigue damage caused by individual turbines operating at a maximum power level set at site-specific capability over a desired lifetime. The offset may be adjusted until fatigue damage caused over time as a result of operating the turbine according to the control schedule equals fatigue damage caused over time as a result of operating the turbine according to a constant maximum power level set at an individual turbine maximum power level for a target minimum lifetime.

Optionally, the initial predefined control schedule specifies a gradient of variation of the maximum power level over time. Adjusting the parameter may then comprise adjusting the gradient.

Optionally, the control schedule indicates an amount of fatigue damage that should be caused over time, the method further comprising operating the wind turbine based on the one or more LUEs to cause fatigue damage at a rate indicated by the control schedule.

Optionally, the method further comprises providing the determined control schedule to a wind turbine controller to control the power output of the wind turbine.

The method may be performed only once, or may be performed aperiodically as needed. Alternatively the method may be repeated periodically. In particular, the method may be repeated daily, monthly or yearly.

A corresponding controller for a wind turbine or wind power plant may be provided, configured to perform the method described herein.

According to a first aspect, there is provided a method for generating a control schedule for a wind power plant comprising two or more wind turbines, the control schedule indicating for each wind turbine how the maximum power level varies over time, the method comprising:

receiving an input indicative of a target minimum expected life for each turbine;

determining a value indicative of a current remaining fatigue life of each of the wind turbines or one or more turbine components of each of the wind turbines based on the measured wind turbine sites and/or operational data;

changing a parameter of an initial predefined control schedule that specifies how the maximum power level of the power plant varies over time by:

i) adjusting parameters of an initial predefined control schedule;

ii) estimating future fatigue life consumed by the turbine or one or more turbine components for the duration of the changed control schedule based on the changed control schedule using a site check program that determines loads acting on the turbine components and including interactions between the turbines of the wind power plant based on data obtained from wind power plant sensors and parameters related to the wind power plant and wind turbine design; and

Optionally, the sensor data comprises sensor data collected prior to commissioning and/or construction of the wind turbine or wind power plant.

Optionally, step (iii) is further constrained such that for any given period of time within the schedule, when the power of all turbines is added together, the sum of the powers does not exceed the amount of power that can be carried in the connection from the power plant to the power grid.

According to a second aspect of the invention, there is provided a method of generating a control schedule for a wind turbine, the control schedule indicating how the turbine maximum power level varies over time, the method comprising:

receiving input indicative of a maximum number of times each of the one or more turbine components is to be replaced within a remaining life of the turbine;

determining a value indicative of one or more of the turbine components or a current remaining fatigue life of the turbine based on the measured wind turbine site and/or operating data;

iv) adjusting parameters of the initial predefined control schedule;

v) based on the changed control schedule and taking into account the replacement of one or more turbine components,

estimating a future fatigue life consumed by the turbine or one or more turbine components for the duration of the changed control schedule; and

vi) repeating steps (i) and (ii) until the estimated future fatigue life consumed by each of the one or more turbine components or the wind turbine is sufficient to allow a target minimum wind turbine life to be reached.

Optionally, step (iii) also requires that energy capture over the lifetime of the turbine is maximised.

Optionally, step (i) may further comprise adjusting for one or more of the turbine components the number of times that component may be replaced within the remaining life of the turbine. Step (i) may also include making adjustments to one or more of the turbine components as to when the components may be replaced during the remaining life of the turbine.

Optionally, the target minimum wind turbine life is a predetermined target value corresponding to a turbine design life.

Optionally, the method further comprises receiving an input indicative of a user-defined target minimum wind turbine life.

Optionally, adjusting the parameter comprises applying an offset, amplification, attenuation or gain factor to the control arrangement. The parameters may be adjusted until all or substantially all of the fatigue life of the heaviest loaded component is consumed for the scheduled duration. The offset may be adjusted by equalizing the area of the curve above and below the line shown below: the line indicates fatigue damage caused by individual turbines operating at a maximum power level set at site-specific capability over a desired lifetime. The offset may be adjusted until fatigue damage caused over time as a result of operating the turbine according to the control schedule equals fatigue damage caused over time as a result of operating the turbine according to a constant maximum power level set at an individual turbine maximum power level for a target minimum lifetime.

Optionally, the initial predefined control schedule specifies a gradient of variation of the maximum power level over time. Adjusting the parameter may comprise adjusting the gradient.

Optionally, the one or more turbine components comprise one or more of the following: a blade, a pitch bearing, a pitch actuation system, a hub, a main shaft, a main bearing, a gearbox, a generator, a converter, a yaw drive mechanism, a yaw bearing or a transformer.

A corresponding controller for a wind turbine or wind power plant configured to perform the method described herein may be provided

Still according to the second aspect, there is provided a method of generating a control schedule for a wind power plant comprising two or more wind turbines, the control schedule indicating for each wind turbine how the maximum power level varies over time, the method comprising:

receiving input indicative of a maximum number of times each of one or more turbine components of each turbine is to be replaced within a remaining life of the turbine;

iv) adjusting parameters of the initial predefined control schedule;

v) estimating future fatigue life consumed by the turbine or one or more turbine components for the duration of the changed control schedule using a site check program based on the changed control schedule and taking into account the replacement of one or more turbine components, the site check program determining loads acting on the turbine components and including interactions between the turbines of the wind power plant based on data obtained from wind power plant sensors and parameters related to the wind power plant and wind turbine design;

and

vi) repeating steps (i) and (ii) until the estimated future fatigue life consumed by each of the one or more turbine components or the wind turbine is sufficient to allow the target minimum wind turbine life to be reached.

According to a third aspect of the invention, there is provided a method of generating a control schedule for a wind turbine, the control schedule indicating how the turbine maximum power level varies over time, the method comprising:

determining a value indicative of a current remaining fatigue life of the turbine or one or more turbine components based on the measured wind turbine site and/or operational data;

applying an optimization function that changes an initial control schedule to determine an optimized control schedule by changing a tradeoff between energy capture and fatigue life consumed by the turbine or one or more turbine components until an optimized control schedule is determined, the optimization comprising:

estimating a future fatigue life consumed by the turbine or turbine component for the duration of the altered control schedule based on the current remaining fatigue life and the altered control schedule; and constraining optimization of the control schedule in accordance with one or more input constraints;

wherein the input constraints comprise a maximum number of allowable component replacements for one or more turbine components, and the optimizing further comprises changing an initial value of wind turbine life to determine a target wind turbine life.

According to a fourth aspect of the present invention there is provided a method of generating a control schedule for a wind turbine, the control schedule indicating how the turbine maximum power level varies over time, the method comprising:

wherein the input constraints comprise a target minimum wind turbine life, and the optimization further comprises changing an initial value of a number of component replacements to be performed on one or more components over the course of the schedule to determine a maximum number of component replacements.

According to a fifth aspect of the present invention there is provided a method of generating a control schedule for a wind turbine, the control schedule indicating how the turbine maximum power level varies over time, the method comprising:

wherein the optimizing further comprises changing an initial value for wind turbine life and changing an initial value for a number of component replacements to be performed on the one or more components over the course of the schedule to determine a combination of the number of component replacements for the one or more turbine components and the target minimum wind turbine life.

The following optional features may be applied to the third, fourth or fifth aspect.

The control arrangement may be applied throughout the life of the turbine.

Optionally, the method further comprises optimizing the control schedule by changing the timing of the component replacements and changing the number of component replacements until a maximum number is reached.

Optionally, the one or more turbine components that can be replaced include one or more of: a blade, a pitch bearing, a pitch actuation system, a hub, a main shaft, a main bearing, a gearbox, a generator, a converter, a yaw drive mechanism, a yaw bearing or a transformer.

Optionally, the initial control schedule specifies a relative change over time in the achievable turbine maximum power level at which the turbine may operate.

Optionally, the input constraints further comprise an upper turbine maximum power output and/or a minimum turbine power output allowed by the turbine design.

Optionally, determining a value indicative of the current remaining fatigue life of the turbine or one or more turbine components comprises using data obtained from wind plant sensors in conjunction with a site inspection program that determines loads acting on the turbine components based on the wind plant sensors and parameters related to wind plant and wind turbine design.

Optionally, the optimization of the control schedule comprises changing the control schedule to minimize a leveled energy cost (LCoE). LCoE may be determined using an LCoE model that includes parameters for one or more of: a capacity factor indicative of the energy generated over a period of time divided by the energy that would be generated if the turbine were continuously operating at rated power over the period of time; availability, which indicates the time at which the turbine will be available to generate electricity; and a field efficiency indicating the energy generated over a period of time divided by the energy that would be generated if the turbine were operating in a wind completely undisturbed by the upstream turbine. The model may also include parameters for one or more of: costs associated with replacing one or more components, including turbine downtime, labor and equipment for component replacement, manufacturing or refurbishment costs for replacement components, and transportation costs for transporting refurbished or replaced components to a power plant; and the service costs associated with replacement of worn parts.

Optionally, the optimized control schedule is a schedule of maximum power levels that can be reached at which the turbine can operate, and which may specify a maximum power level that is higher than the rated power of the wind turbine. Alternatively, the control schedule may specify an amount of fatigue damage that should be caused over time, the method further comprising operating the wind turbine based on the one or more LUEs to cause fatigue damage at a rate indicated by the control schedule.

The control schedule may indicate how the turbine maximum power level varies over the lifetime of the turbine.

Optionally, the method may further comprise providing the optimized control schedule to a wind turbine controller or a wind power plant controller to control the power output of the wind turbine.

Optionally, the method is repeated periodically. The method may be repeated daily, monthly or yearly.

A corresponding controller for a wind turbine or wind power plant may be provided configured to perform a method according to the third, fourth or fifth aspect described herein.

According to a third aspect, there is provided an optimizer for generating a control schedule for a wind turbine, the control schedule indicating how the turbine maximum power level varies over time, the optimizer comprising:

an optimization module configured to receive: initial values of a set of variables, the set of variables being operational variables of the wind turbine and comprising an initial control schedule; one or more constraints; and data indicative of the current remaining fatigue life of the turbine or one or more turbine components;

wherein the optimization module is configured to:

optimizing a control schedule by minimizing or maximizing the operating parameters received at the optimization module that are dependent on the variables by changing one or more of the set of variables from their initial values as a function of the remaining fatigue life of the turbine or one or more turbine components and the one or more constraints; and

outputting the optimized control schedule;

wherein the constraints include a maximum number of allowable component replacements for one or more turbine components, and the optimization module is further configured to change an initial value of wind turbine life to determine a target wind turbine life.

According to a fourth aspect, there is provided an optimizer for generating a control schedule for a wind turbine, the control schedule indicating how the turbine maximum power level varies over time, the optimizer comprising:

wherein the optimization module is configured to:

the optimized control schedule is output and,

wherein the constraints include a target minimum wind turbine life, and the optimization module is further configured to change an initial value of a number of component replacements to be performed on one or more components over the schedule to determine a maximum number of component replacements.

According to a fifth aspect, there is provided an optimizer for generating a control schedule for a wind turbine, the control schedule indicating how the turbine maximum power level varies over time, the optimizer comprising:

wherein the optimization module is configured to:

the optimized control schedule is output and,

wherein the optimization module is further configured to change an initial value of wind turbine life and change an initial value of a number of component replacements to be performed on the one or more components over the course of the schedule to determine a combination of the number of component replacements for the one or more turbine components and the target minimum wind turbine life.

The optional features described below may be applied to the optimiser of the third, fourth or fifth aspect.

Optionally, the optimizer further comprises an initialization module configured to receive initial values of the set of variables and sensor data, the initialization module configured to calculate initial values of the operating parameters.

Optionally, the one or more turbine components are one or more of the following: a blade, a pitch bearing, a pitch actuation system, a hub, a main shaft, a main bearing, a gearbox, a generator, a converter, a yaw drive mechanism, a yaw bearing or a transformer.

Optionally, the operating parameter is a flattened energy cost (LCoE) of the turbine, and optimizing the control schedule includes minimizing the flattened energy cost (LCoE). LCoE may be determined using an LCoE model that includes parameters for one or more of: a capacity factor indicative of the energy generated over a period of time divided by the energy that would be generated if the turbine were continuously operating at rated power over the period of time; availability, which indicates the time at which the turbine will be available to generate electricity; and a field efficiency indicating the energy generated over a period of time divided by the energy that would be generated if the turbine were operating in a wind completely undisturbed by the upstream turbine. The model may also include parameters for one or more of: costs associated with replacing one or more components, including turbine downtime, labor and equipment for component replacement, manufacturing or refurbishment costs for replacement components, and transportation costs for transporting refurbished or replaced components to a power plant; and the service costs associated with replacement of worn parts.

There may be provided a controller comprising an optimizer according to any of the third, fourth or fifth aspects.

According to a third aspect, there is provided a method of generating a control schedule for a wind power plant comprising a plurality of wind turbines, the control schedule indicating for each wind turbine how the maximum power level varies over time, the method comprising:

determining a value indicative of a current remaining fatigue life of each of the turbines or one or more turbine components of each of the turbines based on the measured wind turbine site and/or operating data;

applying an optimization function that changes an initial control schedule for each of the turbines to determine an optimized control schedule by changing a tradeoff between energy capture and fatigue life consumed by each of the turbines or one or more turbine components of each of the turbines until an optimized control schedule is determined, the optimization comprising:

estimating future fatigue life consumed by the turbine or turbine component for the duration of the altered control schedule based on the current remaining fatigue life and the altered control schedule using a site check program that determines loads acting on the turbine components and including interactions between the turbines of the wind power plant based on data obtained from wind power plant sensors and parameters related to the wind power plant and wind turbine design; and

constraining optimization of the control schedule in accordance with one or more input constraints;

wherein the constraints include a maximum number of allowable component replacements for each of the one or more turbine components of each of the wind turbines, and the optimization module is further configured to change the initial value of wind turbine life to determine a target wind turbine life.

According to a fourth aspect, there is provided a method of generating a control schedule for a wind power plant comprising a plurality of wind turbines, the control schedule indicating for each wind turbine how the maximum power level varies over time, the method comprising:

wherein the constraints include a target minimum wind turbine life for each of the wind turbines, and the optimization module is further configured to change an initial value of a number of component replacements to be performed on one or more components of each of the wind turbines over the course of the schedule to determine a maximum number of component replacements.

According to a fifth aspect, there is provided a method of generating a control schedule for a wind power plant comprising a plurality of wind turbines, the control schedule indicating for each wind turbine how the maximum power level varies over time, the method comprising:

wherein the optimizing further comprises changing the initial value for each of the wind turbine lifetimes, and changing the initial value for the number of component replacements to be performed on the one or more components of each of the wind turbines over the course of the schedule to determine a combination of the number of component replacements for the one or more turbine components of each of the wind turbines and the target minimum wind turbine lifetime for each of the wind turbines.

The optional features described below may be applied to the power plant level method of the third, fourth or fifth aspect.

Optionally, the initial control schedule specifies for each turbine a relative change over time in the turbine maximum power level to which the turbine may operate.

Optionally, the optimisation function varies the number of times a component can be replaced over the remaining life of the turbine for one or more of the turbine components. The optimization function may change for one or more of the turbine components when a component may be replaced during the remaining life of the turbine.

Optionally, the method may be further constrained such that for any given period of time within the schedule, when the power of all turbines is added together, the sum of the powers does not exceed the amount of power that can be carried in the connection from the power plant to the power grid.

There may be provided a corresponding wind power plant controller configured to perform the method of the third, fourth or fifth aspect described above.

Any of the methods described herein may be embodied in software that, when executed on a processor of a controller, will cause the controller to perform the associated method.

The site inspection software referred to herein includes site inspection tools known to those skilled in the art for simulating the operation of wind turbines based on pre-construction and/or pre-commissioning sensor data and other site information (e.g., terrain, etc.) to determine operational characteristics of the wind turbines and wind power plants. The site inspection tool may also use operational data from the turbine or power plant or from similar turbines or power plants, if available. Examples include the vestas (tm) site review tool. DNV GL provides an alternative site check software package. It consists of three related programs, "WindFarmer", "WindFarmer blanked Link" and "blanked", which allow the user to perform all-around performance and load calculations.

Drawings

The invention will now be described, by way of example only, with reference to the accompanying drawings, in which:

FIG. 1A is a schematic front view of a conventional wind turbine;

FIG. 1B is a schematic representation of a conventional wind power plant including a plurality of wind turbines;

FIG. 2 is a graph illustrating a conventional power curve of a wind turbine;

FIG. 3 is a graph showing how the power produced by a wind turbine over time may vary over the target life of the turbine;

FIG. 4 is a chart showing different power schedules for wind turbines in which individual maximum wind turbine power levels vary over the life of the turbine to control power output;

FIG. 5 is a graph illustrating exemplary variations in total life fatigue accumulated between different turbine components;

FIG. 6 is an example of a simplified flattened energy cost model for a wind power plant;

FIG. 7 is a block diagram of an exemplary optimizer for optimizing a wind turbine control strategy;

FIG. 8 is an example of a method for determining a maximum power level for a wind turbine type; and

FIG. 9 is a schematic view of a wind turbine controller arrangement.

Detailed Description

Embodiments of the present invention seek to improve the flexibility available to the turbine operator when employing a control approach that compromises energy capture and fatigue loading. Specifically, embodiments provide an optimization method that allows turbine operators to optimize turbine performance (e.g., AEP) according to their requirements.

To optimize performance, three parameters are available for change in the overall wind turbine control strategy. The parameters are (i) the power schedule of the wind turbine; (ii) remaining life of the wind turbine; and (iii) the number of component replacements allowed during the remaining life of the wind turbine. One or more of these parameters may be varied relative to one or more of the other parameters to achieve an optimized control strategy. The parameters may also be subject to constraints.

For example, optimization may be performed to increase its AEP and increase profitability over the life of the turbine. The turbine operator may specify one or more constraints, after which the optimization may be performed. The operator may require one or more of a minimum wind turbine life (e.g., 19 years), a maximum number of individual component replacements (e.g., replacement of a gearbox), and/or a particular power schedule, schedule curves or forms, or schedule gradients.

Power scheduling is a variable used by wind turbine controllers to trade off energy capture and fatigue loading over the remaining turbine life, for example, when over-rating the turbine. The additional power generated by over-rating a given turbine may be controlled by specifying the value of a variable such as the individual wind turbine maximum power level. This maximum power level specifies a power above the rated power until which the turbine can be operated over-rated. The power schedule may specify a constant maximum power level over the lifetime of the turbine. Alternatively, the power schedule may specify a maximum power level that varies over the life of the wind turbine such that the amount of additional power that can be generated by over-rating varies over time. For example, a power plant operator may wish to generate more power during the early years of a wind turbine life at the expense of increased fatigue life consumption of the turbine components, because the financial value of generating power in the early years of the project is disproportionately high.

The individual wind turbine maximum power level for a given turbine type is constrained by the ultimate load limits of the wind turbine mechanical components and the design limits of the electrical components, as the maximum power cannot be safely increased beyond a level that would subject the turbine to a higher mechanical or electrical load value than its ultimate design load limit. The upper maximum power level that an individual wind turbine maximum power level cannot exceed may be referred to as a "wind turbine type maximum power level" and specifies a maximum power level at which the determined load does not exceed the design load for that type of wind turbine. An example of the way in which the maximum power level of a wind turbine type can be calculated is given in the "maximum power level calculation" section below.

The individual wind turbine maximum power level is the power level specified in the arrangement according to an embodiment of the invention and may be referred to simply as the maximum power level. The individual wind turbine maximum power level may be refined for each individual turbine, i.e. calculated based on one or more of the fatigue load values of each turbine, based on the conditions to which each of the wind turbines is exposed at its specific location or location in the wind power plant, wherein the individual wind turbine maximum power level is determined for each turbine in a given site. The individual wind turbine maximum power level may be set such that the rate at which fatigue life is consumed by the turbine or by individual turbine components gives a fatigue life that corresponds to or exceeds a particular target life.

The remaining life of the wind turbine specifies the amount of operating life that the operator is willing to accept in order to optimize the AEP. The remaining life will depend on the point in time since the first start-up of the AEP optimization method was implemented, as the available remaining life decreases with turbine operation.

The number of component replacements allowed during the remaining life of the wind turbine may also be used to optimize the AEP. As turbine components fatigue at different rates under different conditions, the actual life of some components may well exceed the 20 year expected life of the wind turbine, or equivalently the components can be over-rated by a greater amount over a given life. Components with longer life do not contribute to the overall turbine life and have idle production capacity. However, those components with shorter life may have a limiting effect on over-rating, and the AEP may be increased by replacing one or more of these components during the life of the turbine. In particular, the over-rating achieved by increasing the torque has a particularly large effect on the fatigue life of the gearbox, generator and power take off components. In contrast, in the event of an over-rating achieved by increasing the rotor speed, the fatigue life of the blades and structural components is more heavily influenced.

Replaceable components in the context of embodiments of the invention are considered to be essential components, for example components that each account for 5% or more of the total wind turbine cost and that can be replaced in the field. Typical wear parts, which only account for a small fraction of the total cost of the wind turbine, do not have to be considered. In particular, the components considered for replacement may include one or more of a blade, a pitch bearing, a pitch actuation system, a hub, a main shaft, a main bearing, a gearbox, a generator, a converter, a yaw drive mechanism, a yaw bearing, or a transformer.

FIG. 3 shows a first example of optimization in which the power schedule is varied with respect to the target life of the turbine. In this example, the design life of the turbine is 20 years, and the power level is fixed over the life of the turbine. It can be seen that the amount of power generated in a given year increases as the life of the wind turbine decreases. As turbine life decreases, the rate of consumption of fatigue life of the turbine or turbine components may increase, allowing additional power to be generated by over-rating. Optimization may be applied depending on the preferences of the turbine operator. For example, the life, net present price (NPV), or net present value (NPW) of the turbine that maximizes AEP may be determined and selected. NPV/NPW can be calculated using known methods.

FIG. 4 illustrates another optimization example, where the power schedule is again changed relative to the target life of the turbine. In this example, the maximum power level specified by the schedule is variable over the lifetime of the turbine. An initial schedule may be specified, for example, a turbine operator may have a desired schedule modality to be used. The schedule defines how the individual wind turbine maximum power levels change over time, but this may be done in a relative rather than absolute manner. In this example, the desired schedule 401 is from the wind turbine type maximum power level P over a turbine life of 20 years_maxTo a nominal or rated power level P of the turbine type_nomIs arranged linearly. For a typical exemplary site where the annual average wind speed is lower than the design wind speed of the turbine, the site-specific capabilities of the individual turbine over a 20 year lifetime are shown by dashed line a. For a particular turbine, it may not be possible to meet the desired schedule 401 without exceeding the fatigue life of the turbine or certain turbine components over the life of the turbine. Thus, the schedule is adjusted until the total fatigue caused according to the power schedule does not exceed the design fatigue life of the heaviest loaded component.

This may be achieved by estimating fatigue damage due to following the schedule over its duration (e.g. up to the turbine design life or a user specified turbine life). The induced fatigue damage may be estimated using a site review function and may be supplemented with LUE data, both taking into account the fatigue damage induced by the load in view of a given micro-site condition. The arrangement may be adjusted until the resulting fatigue life of the heaviest loaded component is equal to the design fatigue life of that component. In other words, the schedule may be adjusted until all or substantially all of the fatigue life in the heaviest loaded component is exhausted for the duration of the schedule.

The schedule may be adjusted by adjusting one or more parameters of the schedule. This may include:

-applying an offset to the schedule by adding or subtracting a value across the entire schedule;

-applying a gain greater than or less than 1 to the arrangement;

-any other means for non-linearly raising or lowering the control schedule via adjustment of the relevant parameters

It is suitably a function to expand/compress or otherwise expand/compress the arrangement as required

Causing it to increase/decrease to change the scheduled power level value.

In one example, the scheduling adjustment may be based on an equivalent graph of the resulting fatigue damage versus time, or the remaining fatigue life versus time, of the most fatigued component, which may be determined by a power schedule graph and using site review software to determine the fatigue damage to the component that will be caused at a given power level at a particular turbine location (otherwise referred to as a turbine micro-site) within the power plant. The graph is adjusted until the areas defined by each schedule above and below the corresponding force line on the equivalent fatigue curve for the desired turbine life are equal. This is achieved, for example, by making the area of the curve above and below the line shown below equal: the line indicates fatigue damage caused by individual turbines operating at a constant maximum power level set at site-specific capability over a desired lifetime. For example, it may be a line equivalent to dashed line a of fig. 3, but it indicates fatigue damage caused by the maximum power of the individual wind turbines over time. Area equalization may be achieved by shifting the power schedule curve up or down by adding or subtracting an offset until the areas are equalized, or by scaling the curve up or down by adjusting one or more parameters of the curve. The total fatigue life consumed by the turbine or turbine components will then reach 20 years of operation. An exemplary arrangement is shown by line 402, which line 402 terminates in a black square i.

Capabilities specific to the site of the turbine over a 19 year lifetime for the same exemplary site are shown by dashed line B. It can be seen that the capacity over the 19 year lifetime is higher than the capacity over the 20 year lifetime. Thus, the resulting 19-year schedule, exemplified by line 403, may have an initial maximum power level value P that is greater than the 20-year schedule 402^I _20yrsHigher initial maximum power level value P^I _19yrs. The schedule 403 ends over 19 years, which is indicated in black square ii.

In the example of FIG. 4, the schedule adjustment is further subject to an additional constraint that the slope or gradient of the schedule should be equal to the slope or gradient of the initial schedule 401 over a 20 year lifetime. Another constraint as used in the example of fig. 4 may also be applied, according to which the slope of the schedule just before the nominal power level (which may be the rated power of the turbine) is reached is equal to the slope of the initial schedule 401, from which point onwards the maximum power level is maintained at the nominal power level. Alternatively, embodiments may employ a derating of the turbine such that the maximum power level specified by the schedule is set to a level below the rated power of the turbine.

The schedule may be adjusted in a step-wise manner, which may be from P_maxDecrease, or from P_nomIncrease, or increase from the power value of line a until a suitable schedule is reached for which the fatigue life of the heaviest loaded turbine component is sufficient to reach the target turbine life. For example, the initial maximum power level P may be made^IFrom P_nomIncrease or decrease in 1% steps until the proper schedule is reached.

There are other possibilities for optimizing the power schedule according to the number of years of the turbine life. For example, the schedules may all start at the same initial value (e.g., P)_max) And the gradient is varied until the areas defined by each schedule above and below the corresponding force line on the equivalent fatigue curve for the desired turbine life are equal.

Another line 404 illustrates an example of a schedule that may be achieved over a 20 year life for a turbine taking into account one or more component replacements. The schedule 404 terminates with a black box i. One or more components may be particularly susceptible to fatigue damage caused by over-rating. For example, as shown in FIG. 5, after 20 years of operation, one component may reach a 20 year life fatigue limit while other components still have some life. In this case, replacing one or more components that cause a higher rate of fatigue damage will allow the AEP to be increased. This may still increase the profitability of the turbine when calculating the NPV, taking into account the total cost of replacement, and taking into account the lifetime of the turbine.

As an alternative to a schedule specifying a maximum power level value, it is also possible to specify a schedule of fatigue damage or remaining fatigue life, as the rate of fatigue damage caused is related to the maximum power level setting of the turbine. The turbine power output is then controlled to maintain the remaining fatigue life at that specified by the schedule, for example by using LUE tracking fatigue life in the turbine controller. As another alternative, an energy schedule may also be used, as it still indicates how the turbine maximum power level changes over time. The energy schedule may be yearly, calendar months, or the like.

For the avoidance of doubt, the arrangement may also have a non-linear morphology, for example a morphology following a polynomial curve.

Although the schedules are shown as continuously varying over their duration, they may also vary in a step-wise manner, specifying a given maximum power level over a particular time period, such as a month, quarter, or year. The schedule may, for example, be a series of age values over the life of the turbine.

The schedule may be calculated once, or the calculation may be repeated at certain intervals. For example, the schedule may be calculated monthly or yearly. For schedules that specify maximum power levels on an annual basis, it may be advantageous to calculate the schedule (e.g., monthly or weekly) because changes made to the schedule may alert the user to parameters that change faster than expected.

If the schedule is calculated once, the calculation may occur before the wind power plant is put into production, or may occur at any time after production. For calculations that are repeated at intervals, the first calculation may occur before the wind power plant is put on production, or may occur at any time after production.

First example

According to a first example, a control arrangement is generated which can be used for controlling the wind turbine. A relative arrangement may be defined and one or more of a minimum wind turbine life or a maximum number of major component replacements may be defined. Thereafter, the schedule is adjusted to ensure that the fatigue life of the turbine meets the target life while maximizing AEP.

The wind turbine is operated according to one of the over-rating control techniques described herein using an over-rating controller, which may be implemented by the wind turbine controller.

A Life Usage Estimator (LUE) may be used to determine and monitor the life usage of the component. The life usage estimator may be used to ensure that the fatigue load limit of all turbine components remains within their design life. The load (e.g., its bending moment, temperature, force or motion) experienced by a given component may be measured and the amount of component fatigue life consumed calculated, for example, using techniques such as rain flow counting and Miner's law or chemical decay equations. Thereafter, based on the life usage estimator, the individual turbines may be operated in a manner that does not exceed their design limits. The device, module, software component or logic for measuring the fatigue life consumed by a given turbine component may also be referred to as its life usage estimator, and the same abbreviation (LUE) will be used to refer to the algorithm for determining the life usage estimate and the corresponding device, module or software or logic. The LUE will be described in more detail below.

According to the default operating mode, the over-rating controller will control the amount of over-rating applied on a function or schedule basis over the expected or certified life of the wind turbine. Typically, this is 20 or 25 years.

The controller is configured to receive input parameters from, for example, a site operator, which input defines a new target life for the wind turbine or one or more specific turbine components. The LUE is used to determine the up to now life usage of the turbine or related turbine components. This imposes a constraint on the amount of component life remaining for the wind turbine and, therefore, a constraint on the control schedule. Furthermore, the revised target life imposes constraints on the amount of time that the remaining component life must be extended.

The future available fatigue life may be calculated offline or online using site inspection software, and may be used to specify a revised control schedule. The site review function may include calculations or one or more simulations to determine an expected fatigue damage rate using site-based historical data including site climate data measured prior to and/or after construction, and/or data from LUEs. Site climate data typically includes data from a MET mask or ground-based LIDAR, and may include wind speed, turbulence intensity, wind direction, air density, vertical wind shear, and temperature. Site check calculations may be performed remotely or by the turbine/power plant level controller as needed.

Information or parameters relating to the topography, terrain, wind conditions, etc. of a given WPP site may be input to the site review software. Topographical and topographical information may be provided by site survey and/or by knowledge of the WPP site, which may include details of the slope, cliff, inflow angle of each turbine within the WPP, and the like. The wind conditions, such as wind speed (seasonal, annual, etc.), turbulence intensity (seasonal, annual, etc.), air density (seasonal, annual, etc.), temperature (seasonal, annual, etc.), and the like, may be provided by Met mask data and/or by wind conditions experienced and recorded by the wind turbine and/or the WPPC at the location of the WPP.

The site review tool may include one or more memories, databases, or other data structures to store and maintain fatigue load values for each type of wind turbine, a wind turbine type maximum power level for each type of wind turbine, and information and/or parameters related to WPP site conditions.

A revised control schedule is thus generated whereby the additional power produced by over-rating is adjusted to expose the turbine to a higher or lower rate of fatigue damage accumulation depending on whether the new target date of end-of-life is earlier or later than the previous target date (which may be a certified life).

The ability to make revisions to turbine control schedules allows operators to change their priorities over time. For example, a primary generator on a local grid may be taken out of service for mid-life service, or may be completely decommissioned, and the grid may require additional support. This can be reflected in a significantly higher long-term tariff, and it would therefore be advantageous for the operator to increase energy production in the short term. Thus, the operator may decide to reduce the turbine life or the life of the affected components (e.g., gearbox and generator) and generate additional power by over-rating while accepting a shorter wind turbine or turbine component life.

It is possible to use methods other than LUE to determine the lifetime usage of a wind turbine or turbine component. Instead, the operation of the turbine up to now can be checked and the fatigue damage that has occurred up to now can be calculated. This may be particularly advantageous when retrofitting over-rated control for a wind turbine, and the future available fatigue life is calculated off-line, again using site review software, and used to specify the maximum power level. The site review function may also include offline or online calculations or one or more simulations to determine an expected fatigue damage rate using site-based historical data or site data measured for the installation site, although in this case the calculations may be made without available LUE data.

The operation of the wind turbine up to the date of assembly of the over-rating controller employing the functionality described herein may be checked using site review software to calculate fatigue loads on turbine components using measurements relating to the exact location of the wind turbine within the wind power plant site, for example, one or more of energy output, wind speed, wind direction, turbulence intensity, wind shear, air density, turbine mechanical load measurements (e.g., from blade load sensors), turbine electrical component temperatures and loads, icing events, component temperatures and condition monitoring system outputs, based on input parameters specifying site geography, site meteorological conditions, etc. These values may be used to calculate an estimate of the fatigue damage that has occurred to date to the turbine component. The future available life of the turbine or turbine component may be calculated by applying the measured values to a site check function wind turbine model or simulation that provides as output an estimated fatigue damage and/or remaining fatigue life based on one or more of these measured values and a value for the wind turbine type maximum power level of the turbine. The simulation or model may provide fatigue damage and/or remaining fatigue life of the turbine at the component level or as a whole. The fatigue load calculation may be performed according to various calculation processes. Various examples of such site check procedures will be known to those skilled in the art and will not be described in detail.

The resulting estimate of fatigue life consumed by the turbine or turbine component may be used to determine an over-rating strategy to be applied by the controller. The estimation may be used once when over-rating control is initialized, which may be performed in the middle of the life of the turbine if the turbine is being retrofitted. Alternatively, the estimation may be performed periodically during the life of the turbine, such that the over-rating strategy is periodically updated depending on how the life fatigue consumption changes throughout the life of the turbine.

The over-rating strategy is determined based on the remaining fatigue life of the wind turbine or wind turbine components, which is itself based on the operating life of the wind turbine. The amount of over-rating applied is controlled to cause fatigue damage to the turbine or turbine component at a rate sufficiently low to ensure that the fatigue life of the turbine is only at the end of the predetermined turbine life, and preferably is exhausted just at the end of the predetermined turbine life.

The determination of the component fatigue life estimate may be further extended or replaced by using data from one or more condition monitoring systems. Condition Monitoring Systems (CMS) include several sensors in the turbine gearbox, generator or other critical components at the control critical points of the drive train. The condition monitoring system provides an early warning of component failure before the component actually fails. Thus, the output of the condition monitoring system may be provided to the controller and may be used as an indication of the fatigue life consumed by the monitored component, and in particular can provide an indication of when the fatigue life of the component has reached its end. This provides an additional way of estimating the lifetime used.

Second example

A second example is provided to implement a more general optimization process, which can be used to implement a similar kind of optimization as described above, as well as other more general optimizations. The optimization process of the second example may be implemented by a controller applying an optimization scheme.

A complete financial cost model of the turbine, or a normalized energy cost (LCoE) model, is included and used in offline calculations performed prior to installation of the over-rating control system, or used online as part of the wind turbine controller or wind power plant controller. The use of the LCoE model allows for optimization of the over-rating strategy and may also take into account replacement of the primary components based on their cost. As used herein, the term "leveled energy cost" refers to a measure of the cost of energy from a turbine that is calculated by dividing the life cost of the turbine by the life energy output of the turbine.

FIG. 6 illustrates an example of a simplified LCoE model in which various costs associated with building and operating a wind turbine and wind turbine plant are considered.

Wind Turbine Generator (WTG) costs take into account the total cost of manufacturing a wind turbine. The transportation costs will take into account the costs of transporting the turbine components to the installation site. The operating and maintenance (O & M) costs take into account the operating costs of the turbine and may be updated as operations and maintenance occur. The service technician may provide this information to the local turbine controller, to the wind farm controller, or elsewhere. The capacity factor indicates the energy generated over a given time period (e.g., over a year) divided by the energy that would be generated if the turbine were continuously operating at rated power over that time period. Availability indicates the time at which the turbine will be available to generate power. The field efficiency indicates the efficiency of extracting energy from the wind, which is affected by the spacing of the turbines within the farm.

Only those LCoE elements that are affected by the control and component replacement strategies must be included in the LCoE model, as many parameters that may be included in the LCoE model are fixed at the time of the turbine or wind farm construction. The affected elements are:

● operation and maintenance (O & M) costs

If more parts are replaced, the cost increases

● capacity factor

If more aggressive over-rating is used, and therefore more MWh is generated, the capacity factor increases

● availability

If more major components are replaced, the availability is slightly reduced due to the downtime required for the replacement process

If a more aggressive over-rating causes increased preventive replacement of wearing parts or failure not included in the schedule, the availability is slightly reduced

● life span

Depending on the constraint selection, decrease or increase.

Including a financial cost of the turbine (LCoE) model in the turbine or WPP controller allows for a more flexible and efficient control strategy to be determined. For example, if the conditions at a particular site are found to be particularly adverse to the gearbox, then the condition will be identified and the operator may choose whether to over-rate the turbine and take into account a certain number of replacements of the gearbox. The turbine controller can then determine when the gearbox should be replaced and cause the turbine to operate accordingly, and also optionally provide an indication when the gearbox is to be replaced.

FIG. 7 illustrates a block diagram of an exemplary optimizer for optimizing a wind turbine control strategy, which may be incorporated into a controller and used to implement various embodiments of the present invention.

At algorithm start-up, a block labeled "initialize" is run once. This provides initial conditions for the optimization loop. The loop labeled "optimized" is performed periodically, such as once per day, once per month, or once per year. When the loop is executed, it runs as much as necessary to achieve a good enough convergence for the optimization process. Following convergence, the new set of outputs is sent to the wind turbine controller (x1) and the operator (other outputs) to implement the determined control strategy. The two blocks of "calculating the estimate of LCoE" contain the same calculation method. They include all the not yet fixed elements of fig. 6, i.e., O & M cost, capacity factor, availability and lifetime. For example, the tower CAPEX is already fixed and thus does not have to be included. But the operating and maintenance (O & M) costs are not fixed, since the gearbox may be more work intensive and replaced once during the life of the turbine, thus including the cost.

Not all connections are shown in fig. 7, where there are many similar connections, for example between the optimization algorithm block and the "calculate estimate of LCoE" block. The following notation is used in fig. 7 or with reference to fig. 7:

● N: the number of cycles (e.g., years) of remaining life. The user can change it, if necessary, to meet his desired operating policy.

● x 1: a one-dimensional array of individual wind turbine maximum power levels in year 1.. N, e.g., [3.5MW,3.49MW,3.48MW,3.47MW.. ] for a 3WM turbine

● x 2: a one-dimensional array of gearbox replacement times in N years 1.. 0,0,0,0,0,1,0,0,0,0,0]

● x 3: one-dimensional array of generator replacement times in the 1 st

● x 4: one-dimensional array of main bearing replacement times in year 1

● x 5: one-dimensional array of blade set replacement times in year 1

And optionally:

● x 6: one-dimensional array of converter replacement times in year 1

● x 7: one-dimensional array of pitch bearing replacement times in year 1.. N

● x 8: one-dimensional array of pitch actuator (hydraulic or electrical) replacement times in N years 1

● x 9: one-dimensional array of yaw drive mechanism change times in year 1

● x 10: one-dimensional array of yaw bearing replacement times in year 1

● x 11: one-dimensional array of transformer replacement times in the 1 st

● "_ 0" indicates an initial condition, e.g., x1_0 is an initial condition of x 1.

Referring to FIG. 7, the optimization process entails determining several constants for a given turbine and calculating the initial conditions for optimization using the values of several physical and control parameters. Once the initial conditions have been calculated, the optimization process applies a function defining the relationship between the flattened energy costs and the input values of the physical and control parameters to determine a combination of input values that minimizes the flattened energy costs without exceeding certain optimization constraints.

To calculate the initial conditions for optimization, several parameter values for a given turbine are determined and input into an "initialization" block. These values are constant for any given periodic optimization (e.g., monthly). They are operator entered parameters and can change at any time, but if changed, will be applied the next time the optimization is run. These parameters may include one or more of the following: the life of the turbine/individual turbine components; the cost of gear box replacement; cost of bearing replacement; generator replacement cost; blade replacement cost; the pitch system replacement cost; as well as the replacement cost of any other necessary components.

For example, the site check function and/or one or more LUEs may be used to determine the life of the turbine and/or the life of one or more components, or the life of the turbine and/or the life of one or more components may be provided as constraints to be met. The exchangeable part comprises a blade, a pitch bearing, a pitch actuation system, a hub, a main shaft, a main bearing, a gearbox, a generator, a converter, a yaw drive mechanism, a yaw bearing or a transformer.

Determining a total cost of replacing each of the components. For example, for replacing a gearbox, the cost will consider whether a new or retrofit gearbox is being assembled, transportation costs, and crane and labor costs. According to the availability section in fig. 6, turbine down time costs are also included.

Other costs are also included, such as financial costs including Weighted Average Capital Cost (WACC) and the like, as well as any other elements needed to calculate the impact of future turbine operating strategies on LCoE.

The lifetime parameter may be set by the operator depending on its operating strategy for the site, or may be determined as part of the optimization. The other constants are based on best knowledge so they can be updated occasionally, but such updating will be quite rare. In particular, the O & M cost can only be estimated in advance, and these estimates can be replaced with actual data over time, thereby obtaining a more accurate estimate of the future O & M cost.

The "initialization" block and optimization algorithm uses the following variables:

● x 1: one-dimensional array of maximum power levels in year 1.. N, e.g., [3.5MW,3.49MW,3.49MW,3.48MW,3.47MW.. ] for a 3WM turbine

● x 3: one-dimensional array of generator replacement times in the 1 st

● x 4: one-dimensional array of main bearing replacement times in year 1

● x 5: one-dimensional array of blade set replacement times in year 1

And optionally:

● x 6: one-dimensional array of converter replacement times in year 1

● x 7: one-dimensional array of pitch bearing replacement times in year 1.. N

● x 9: one-dimensional array of yaw drive mechanism change times in year 1

● x 10: one-dimensional array of yaw bearing replacement times in year 1

● x 11: one-dimensional array of transformer replacement times in the 1 st

The initial calculation of the estimate of LCoE uses initial estimates from the operator for initial conditions, i.e., x1_0, x2_0, x _0, etc.

The signal labeled "measurement data" in fig. 7 contains data from the sensors as well as data determined by the O & M process. The measurement data from the sensor may be from a turbine or a wind power plant, and may include one or more of the following options:

LUE values for one or more of turbine components such as gearbox, generator, main bearing, blades, converter, pitch bearing, pitch actuator (hydraulic or electrical), yaw drive mechanism, yaw bearing, transformer, etc.;

-wind speed and environmental data, or other data obtained by a site inspection program;

CMS data of one or more of the turbine components.

The measurement data from the operation and maintenance (O & M) activities contains O & M costs, which may include estimates based on costs up to now, if any. It can be used together with future scheduled service patterns, experience of other turbines of the same design from the same or other wind farms, and experience of certain components for other turbines of different designs using the same components to give an estimate of future O & M costs in LCoE calculations.

The optimization process uses the inputs and constraints to minimize the normalized energy costs (LCoE) by calculating LCoE directly or by calculating certain LCoE variables, depending on the initial conditions. Only the portion of the LCoE that changes after the turbine is built, i.e., the portion affected by O & M cost, capacity factor, availability and life, needs to be calculated. The optimization is run until the LCoE is minimized, e.g., until the step-wise variation in the calculated LCoE is within a given tolerance.

The constraint for optimization is the area that the optimization algorithm cannot enter when searching for the minimum of LCoE. The constraints may include one or more of the following: wind turbine type maximum power level; minimum power output for the turbine type; the maximum active power capacity of the connection of the wind power plant to the grid, i.e. the maximum sum of the active power outputs of all turbines; and any other suitable constraints.

The constraints may also include one or more of the following, which may be user defined:

-a minimum or target expected wind turbine life;

-a maximum number of component replacements for all components or one or more given components;

-a predefined maximum power level schedule or a predefined relative maximum power schedule defining a shape of the maximum power schedule.

The number of inputs per one-dimensional array may be selected to make the runtime of the optimization algorithm easier to manage. The one-dimensional arrays x1, x2, etc. are described above as being provided for yearly operations. However, it is possible to provide inputs for monthly or quarterly operations, which would provide 12 or 4 times the number of inputs. Therefore, the year value can be used. Of course, different time periods may be used as desired, depending on the desired computation time or optimization granularity.

Again, to make the runtime easier to manage, the wind turbine components may be selected such that only the most relevant components are used in the optimization. The selection of the components to be included may be based on whether the component lifetime is significantly affected by the active power output above the rated wind speed, which may be, inter alia, the gearbox, the generator, the main bearings and the blades.

Additionally or alternatively, the components used in the optimization may be selected based on their values. For example, only components having a value of 5% or more of the total cost of the turbine are included.

The optimizer algorithm generates several outputs each time it runs to convergence. A scheduled one-dimensional array x1 representing the maximum power level of the turbine over the 1 st.. N years can be used in closed-loop control by automatically communicating that data to the wind turbine controller for use as the turbine power demand until the next optimization loop operation (e.g., 1 month later). Alternatively, the maximum power level may be used in the advisor status without an automated control loop, for example, by sending the maximum power level data to a computer system for output on a display for viewing by a maintenance department.

The other one-dimensional arrays x2, x3, x4 represent arrangements for component replacement. The scheduling data may be output to another computer system, allowing action to be taken. The data may be provided directly to the component replacement scheduling software. Alternatively, the component replacement data including the proposed replacement date may be used as a advisory output that is sent to a display for viewing by a maintenance department to determine a manual implementation of the component replacement plan.

It should be noted that the above-described one-dimensional array of maximum power levels (x1) may be provided as an over-rated level only, as an over-rated level or a under-rated level, or as a under-rated level only, so that the maximum power level variable need only specify an amount that exceeds (or is below) the rated power. Alternatively, the power demand may be a speed demand and/or a torque demand per cycle, or may be a fatigue life consumption, wherein the power is controlled by a life usage control function as described below. A drawback of using both the speed and torque requirements is that the calculation time for calculating the optimal configuration will be longer.

Although the optimizer is described above as being performed periodically, it may be used occasionally, or even once. For example, the optimization may be performed offline at the point where the over-rating controller is installed. Alternatively, the optimizer may be embodied in a controller at the wind turbine, at the wind power plant or elsewhere, in which case it will be executed in specific time steps.

As described above, optimization may be performed with or without LUEs, as the site data may be used to determine component fatigue and thus give an indication of the remaining life available for the turbine or turbine component.

Although the optimization algorithm is described primarily in connection with use with an over-rating controller, this is not required. Optimization may be applied in combination with any control action that compromises energy capture and turbine fatigue loading. This may include one or more of the following: changing power requirements by, for example, derating; thrust limits, which limit power output to avoid high thrust loads by reducing rotor thrust at the "corners" of the power curve at the expense of power output; or any other control feature that compromises energy capture and fatigue loading.

Although the required calculations may be performed anywhere, in practice strategic actions such as these are best performed in the wind power plant controller, e.g. in the SCADA server. This allows the service data to be entered directly at the site, avoiding communication problems from the site to the control center. However, the calculation may also be performed at the control center. The same applies to the other methods described herein, including the method in the first example.

Maximum power level calculation

Exemplary techniques for determining a maximum power level that may be applied to a turbine will be described next.

A method for determining a wind turbine type maximum power level for a type of wind turbine may comprise: simulating a load spectrum for the two or more test power levels to determine a load on the type of wind turbine for each of the two or more power levels; comparing the determined load for each test power level with a design load for the type of wind turbine; and setting a wind turbine type maximum power level for the type of wind turbine to a maximum test power level at which the determined load does not exceed the design load for the type of wind turbine.

Accordingly, a wind turbine type maximum power level can be determined for one or more types of wind turbines.

FIG. 8 shows a flow chart detailing an example of setting a turbine maximum power level that may be used in connection with any embodiment. In step 301, a check is performed to determine wind turbine mechanical component design limits for one or more types of wind turbines. In this example, the design limits are determined using an offline computer system. However, it should be appreciated that the described functionality may be implemented by an online computer system or any other software and/or hardware associated with the wind turbine and/or the WPP.

The wind turbine type maximum power level is: in the case that a wind turbine of a given type is to be operated at the limits of the design loads of the components of the wind turbine, the maximum power level that the wind turbine of the given type is allowed to produce when the wind is suitably high. Wind turbine type maximum power levels are effectively applied for the design life of the turbine. Thus, the wind turbine type maximum power level will typically be higher than the nominal nameplate rating of the wind turbine type, as the nominal nameplate rating is typically a more conservative value.

"one type of wind turbine" as used in the following examples and embodiments may be understood as wind turbines having the same electrical system, mechanical system, generator, gearbox, turbine blades, turbine blade length, hub height, etc. Accordingly, for the purposes of embodiments of the present invention, any differences in the main structure or components of a wind turbine will in fact result in a new type of wind turbine. For example, a wind turbine that is identical except for the difference in hub height (e.g., tower height) would be two different types of wind turbines. Similarly, wind turbines that are identical except for the different lengths of the turbine blades will also be considered as different types of wind turbines. 50Hz and 60Hz wind turbines are also considered to be different types of wind turbines, which are designed for cold and hot climates.

Thus, the type of wind turbine does not necessarily correspond to the International Electrotechnical Commission (IEC) class of wind turbines, as different types of wind turbines may be in the same IEC class of wind turbines, where each type of wind turbine may have a different wind turbine type maximum power level based on the design of the components in the wind turbine.

In the following example, the wind turbine is rated for a nominal nameplate rated power level of 1.65MW (1650KW), has a hub height of 78 meters, and is designed to service under specific IEC wind class conditions.

Design limits for wind turbine type mechanical components of a wind turbine of the type may then be determined by simulating a load spectrum for a first test over-rated power level to identify loads acting on the type of wind turbine for the first power level. The load may be a mechanical load, a fatigue load, any other load that the wind turbine may experience, or any combination of different loads. In this example, mechanical loads are considered, however it should be appreciated that other loads, such as fatigue loads, may also be considered. The process of simulating the load spectrum may also include other forms of analysis that may be used to determine the load on this type of wind turbine or may be an extrapolation of the analysis.

The load spectrum typically includes a range of different test cases that may be run in a computer simulation of the wind turbine. For example, the load spectrum may include test cases for: a wind power of 8m/s for 10 minutes, a wind power of 10m/s for 10 minutes, different wind directions, different wind turbulence levels, start-up of the wind turbine, shut-down of the wind turbine, etc. It should be appreciated that there may be many different wind speeds, wind conditions, wind turbine operating conditions, and/or fault conditions for which test cases may be run in a wind turbine simulation of the load spectrum. The test case may include real actual data or artificial data (e.g., a 50 year gust of wind as defined in standards related to wind turbines). The simulation of the load spectrum may determine forces and loads affecting the wind turbine for all test cases in the load spectrum. The simulation may also estimate or determine the number of times a test case event may occur, for example, a test case of 10m/s wind lasting 10 minutes may be expected to occur 2000 times over the 20 year life of the wind turbine, thus enabling the fatigue on the wind turbine over the life of the wind turbine to be calculated. The simulation may also calculate or determine fatigue damage or loads that may be caused by various components in the wind turbine based on the determined loads affecting the wind turbine.

In this example, the first test power level may be 1700KW, as it is higher than the nominal nameplate rated power level for the type of wind turbine considered in this example. Thereafter, a load spectrum may be simulated for a given type of wind turbine to determine whether the type of wind turbine is capable of operating at the first test power level without exceeding a final design load of mechanical components of the type of wind turbine. If the simulation shows that a wind turbine of this type is capable of operating at a first test power level, the same procedure may be repeated for a second test power level. For example, in this example, the second test power level may be 1725 KW. Thereafter, a load spectrum may be simulated for a given type of wind turbine to identify whether the type of wind turbine is capable of operating at the second test power level without exceeding a final design load of mechanical components of the type of wind turbine.

The process of simulating the load spectrum for other test power levels may be performed iteratively if the final design load of the mechanical component is not exceeded. In this example, the test power level is incremented in steps of 25KW, but it will be appreciated that the incremental steps may be any steps suitable for the purpose of identifying the maximum power level of the wind turbine type, e.g. 5KW, 10KW, 15KW, 20KW, 30KW, 50KW etc., or the incremental steps may be increased by a percentage of the test power level, e.g. 1% increments, 2% increments, 5% increments etc. Alternatively, the process starts at a high first test power level and for each iteration the test power level is decremented by an appropriate amount until a wind turbine type maximum power level is identified, i.e. a first test power level at which the type of wind turbine can operate without exceeding the final design limits.

In this example, a wind turbine of a given type is identified as being capable of operating at other test power levels of 1750kW, 1775kW and 1800kW before exceeding the design limits of one or more mechanical components on 1825 kW. Thus, the process identifies a wind turbine type maximum power level of 1800KW for this type of turbine.

In other embodiments, since wind turbines of the type described do not exceed the final design load of the mechanical components over 1800KW, but exceed the final design load of the mechanical components over 1825KW, the process may further iteratively increment the test power level by small increments, such as 5KW, to identify whether the wind turbine can operate at a power level between 1800KW and 1825KW without exceeding the mechanical final design load. However, in the present example, a power level of 1800KW is taken as a wind turbine type mechanical component design limit for this type of wind turbine.

The process of determining the maximum power level of a wind turbine type may then be performed for any other type of wind turbine to be analyzed. In step 302 of FIG. 8, design constraints for electrical components in wind turbines of the type described may be considered or evaluated against previously determined wind turbine mechanical component design limits.

In step 302, the primary electrical components may be considered to ensure that the wind turbine type power level determined for the mechanical component design limits does not exceed the design limits of the primary electrical components of the type of wind turbine being analyzed. For example, the main electrical components may include a generator, a transformer, an internal cable, a contactor, or any other electrical component in a wind turbine of the type described.

Based on the simulation and/or calculation, it is then determined whether the primary electrical component is capable of operating at the wind turbine type maximum power level previously determined for the mechanical component design limits. For example, operation at a mechanical component design limit power level may cause the temperature of one or more cables inside the wind turbine to increase and thus reduce the current carrying capacity of the cables, which is determined by the size of the cable conductors and the heat dissipation conditions. Thus, the current carrying capacity will be calculated for the new temperature conditions in order to determine whether the cable can be operated at a power level up to the maximum power level of the wind turbine type. Similar considerations may be considered for other electrical components, such as temperature of the component, capacity of the component, and so forth, to identify whether the electrical component is capable of operating at power levels up to the mechanical component design limit.

If it is determined or identified that the primary electrical component is capable of operating at the previously determined mechanical component design limit, then in step 303 of FIG. 8, for a given type of wind turbine, a wind turbine type maximum power level is set or recorded as the maximum power level of the given type of wind turbine as a function of the mechanical component design limit. However, if it is determined or identified that the primary electrical component is not capable of operating at the previously determined mechanical component design limits, further investigation or action may be taken to arrive at a turbine type maximum power level that coordinates both the mechanical and electrical components.

Once the wind turbine type maximum power level for each type of wind turbine is determined, this parameter may be used as a constraint in the method described above to arrive at a schedule of individual maximum power levels (e.g., maximum over-rated power levels) for each wind turbine in the WPP.

The different individual maximum power levels for each wind turbine in the WPP is advantageous because the conditions in the WPP may be different throughout the WPP site. Thus, it may be the case that a wind turbine at a certain location in a WPP may face a different situation than another wind turbine of the same type at a different location in the WPP. Thus, two wind turbines of the same type may require different individual maximum power levels, or a minimum individual maximum power level may be applied to all wind turbines of that type in a WPP depending on the preferred embodiment. As described herein, individual maximum power levels specific to individual wind turbines are determined as part of a determination schedule.

Over-rating control

Embodiments of the invention may be applied to a wind turbine or a wind power plant, which is operated by applying an over-rating control to determine an amount of over-rating to be applied.

The over-rating control signal is generated by an over-rating controller and is used by the wind turbine controller to over-rate the turbine. The control arrangement described above may be used within or in conjunction with such an over-rating controller to set an upper limit on the amount of power that can be generated by over-rating. The particular manner in which the over-rating control signal is generated is not critical to embodiments of the present invention, but examples will be given to facilitate understanding.

Each wind turbine may include an over-rating controller as part of the wind turbine controller. The over-rating controller calculates an over-rating request signal that instructs the turbine to over-rate the power output by an amount that is greater than the rated output. The controller receives data from the turbine sensors, e.g., pitch angle, rotor speed, power output, etc., and is capable of sending commands, e.g., set points for pitch angle, rotor speed, power output, etc. The controller may also receive commands from the grid, for example from a grid operator, to boost or reduce the active or reactive power output in response to a demand or fault on the grid.

FIG. 9 shows a schematic example of a turbine controller arrangement in which an over-rating controller 901 generates an over-rating control signal that the wind turbine controller can use to apply over-rating to the turbine. The over-rating control signal may be generated in dependence on the output of one or more sensors 902/904 detecting operating parameters and/or local conditions (e.g. wind speed and wind direction) of the turbine. The over-rating controller 901 includes one or more functional control modules that may be used in various aspects of over-rating control. Additional functional modules may be provided, the functions of the modules may be combined, or some of the modules may be omitted.

The values for the individual turbine maximum power levels are provided by an optimizer 907 according to an arrangement determined as described herein. Which provides the maximum power level to which the turbine can operate according to the schedule.

The additional functional module generates a power demand and will generally act to reduce the final power demand to be complied with by the turbine controller. A specific example of an additional functional module is an operational constraint module 906. Over-rating takes advantage of the difference that typically exists between the component design loads and the loads experienced by each turbine in operation, which are typically better than the IEC standard simulation conditions under which the design loads are calculated. Over-rating causes the power demand on the turbine to be increased at high winds until the operating limits specified by the operating constraints (temperature, etc.) are reached, or until the upper power limit set to avoid exceeding component design loads is reached. The operating constraints enforced by the operating constraint control module 906 limit the possible over-rated power requirements as a function of various operating parameters. For example, in the case where the protection function is ready to initiate a shutdown when the gearbox oil temperature exceeds 65 ℃, the operating constraints may dictate that for temperatures exceeding 60 ℃, the maximum possible over-rating set point signal falls linearly as a function of the gearbox oil temperature, thereby achieving "no over-rating" (i.e., a power set point signal equal to the rated power) at 65 ℃.

The maximum power level and power requirement from the functional module are provided to a minimization function, block 908, and a minimum value is selected. Another minimization block 909 may be provided that selects the minimum power requirements from the over-rating controller 901 as well as any other turbine power requirements, such as those specified by the grid operator, to produce the final power requirements to be applied by the wind turbine controller.

Alternatively, for example, the over-rating controller may be part of the PPC controller 130 of fig. 1B. The PPC controller is in communication with each of the turbines and is capable of receiving data from the turbines, such as pitch angle, rotor speed, power output, etc., and sending commands, such as set points for pitch angle, rotor speed, power output, etc., to the individual turbines. The PPC 130 also receives commands from the grid, e.g., from a grid operator, to boost or reduce active or reactive power output in response to a demand or fault on the grid. The controller of each wind turbine is in communication with the PPC 130.

The PPC controller 130 receives power output data from each of the turbines, and thus knows the power output of each turbine and the power output of the power plant as a whole at the grid connection point 140. If desired, the PPC controller 130 may receive an operating set point for the power output of the power plant as a whole and divide it among each turbine so that the output does not exceed the operator assigned set point. The power plant set point may be anywhere from 0 up to the rated power output of the power plant. The "rated power" output of the power plant is the sum of the rated power outputs of the individual turbines in the power plant. The power plant set point may be higher than the rated power output of the power plant, i.e., the entire power plant is over-rated.

The PPC may receive an input directly from the grid connection, or it may receive a signal that is a measure of the difference between the total plant output and the nominal or rated plant output. The difference may be used to provide a basis for over-rating of individual turbines. In theory, it is possible to over-rate only a single turbine, but it is preferred to over-rate a plurality of turbines, and most preferred to send over-rated signals to all turbines. The over-rating signal sent to each turbine may not be a fixed control, but may instead be an indication of the maximum amount of over-rating that each turbine may perform. Each turbine may have an associated controller which may be implemented within the turbine controller or may be implemented centrally, for example at the PPC, which will implement one or more of the functions shown in figure 9 to determine whether the turbine is able to respond to the over-rating signal and, if so, by the amount of over-rating. For example, a controller within the turbine controller may respond positively and over-rate a given turbine if it determines that the conditions at the turbine are favorable and above rated wind speed. As the controller implements the over-rating signal, the output of the power plant will increase.

Thus, an over-rating signal is generated centrally or at each individual turbine, the signal being indicative of the amount of over-rating that one or more turbines or turbines of the power plant as a whole may perform.

Life use estimator

As described above, embodiments of the present invention utilize a Lifetime Usage Estimator (LUE). The lifetime usage estimator will now be described in more detail. The algorithms required to estimate life usage will vary between components, and LUEs may comprise a library of LUE algorithms including some or all of the following: load duration, load speed profile, rain flow count, stress cycle damage, temperature cycle damage, generator thermal response rate, transformer thermal response rate, and bearing wear. Other algorithms may alternatively be used. As described above, the lifetime usage estimation may be used only for selected critical components, and the use of a library of algorithms enables selection of a new component for the LUE and setting of the appropriate algorithm and specific parameters selected from the library for that component.

In one embodiment, LUEs are implemented for all major components of a turbine, including: a blade; a pitch bearing; a pitch actuator or drive mechanism; a hub; a main shaft; a main bearing housing; a main bearing; a gearbox bearing; gear teeth; a generator; a generator bearing; a converter; a generator terminal box cable; a yaw drive mechanism; a yaw bearing; a tower; an offshore support structure (if present); a foundation; and a transformer winding. Alternatively, one or more of these LUEs may be selected.

As an example of a suitable algorithm, rain flow counting may be used in blade structures, blade bolts, pitch systems, main shaft systems, converters, yaw systems, towers and ground estimators. In the blade structure algorithm, rain flow counts are applied to the blade root bending shimmy and flapping moments to identify stress cycle ranges and averages, and the output is sent to a stress cycle damage algorithm. For the blade bolts, the rain flow counts are applied to the bolt bending moments to identify the stress cycle range and average, and the output is sent to the stress cycle damage algorithm. In the pitch system, the main shaft system, the tower and the foundation estimator, a rain flow counting algorithm is also applied to identify the stress cycle range and average and the output is sent to the stress cycle damage algorithm. Parameters for applying the rain flow algorithm may include:

-pitch system-pitch force;

-main shaft system-main shaft torque;

-tower stress;

foundation-foundation stress.

In a yaw system, a rain flow algorithm is applied to the tower top torque to identify load duration and this output is sent to a stress cycle damage algorithm. In the converter, the generator power and RPM are used to infer the temperature, and rain flow counting is used over that temperature to identify the temperature cycle and average.

Life usage in the blade bearings may be monitored by inputting blade lag loads and pitch speed as inputs to a load duration algorithm or a bearing wear algorithm. For the gearbox, the load speed duration is applied to the main shaft torque to calculate the lifetime used. For the generator, the generator RPM is used to infer the generator temperature, which is used as an input to the thermal reaction rate generator algorithm. For transformers, the transformer temperature is inferred from the power and ambient temperature to provide input to the transformer thermal reaction rate algorithm.

Where possible, it is preferred to use existing sensors to provide the input on which the algorithm operates. Thus, for example, wind turbines typically directly measure blade root edgewise and flapping moments required for blade structure, blade bearings, and blade bolt estimators. For a pitch system, the pressure in the first chamber of the cylinder may be measured and the pressure in the second chamber may be inferred, thereby enabling the pitch force to be calculated. These are examples only and other parameters needed as inputs may be measured directly or may be inferred from other available sensor outputs. For some parameters, it may be advantageous to use additional sensors if one value cannot be inferred with sufficient accuracy.

Algorithms for various types of fatigue estimation are known and can be found in the following standards and texts:

load speed distribution and load duration:

Guidelines for the Certification of Wind Turbines,GermainischerLloyd,Section 7.4.3.2Fatigue Loads

rain flow:

IEC 61400-1 'Wind turbines-Part 1: Design requirements', Annex G Miners, sum:

IEC 61400-1 'Wind turbines-Part 1: Design requirements', Annex G power law (chemical denaturation):

IEC 60076-12‘Power Transformers–Part 12:Loading guide for dry-typepower transformers’,Section 5。

plant level control of power generation

Any of the methods described herein may be performed at the wind farm level, thereby generating a farm control schedule comprising individual control schedules for each wind turbine. Doing so has the benefit of allowing for the interaction between turbines in a given power plant to be taken into account.

Changes to the power demand/power level of the upstream turbine or turbines affect the power output and the rate of fatigue damage accumulation for any turbine that immediately follows the upstream turbine or turbines. The site check software includes information about the location of the turbines within the wind farm and takes into account the relative positions of the turbines with respect to each other within the wind farm. Thus, wake effects from upstream turbines are taken into account in the site inspection software implemented calculations.

In the case of some wind power plants, the power carrying capacity of the connection from the plant to the utility grid is less than the sum of the power generated by each turbine in the case when all turbines generate power at the wind turbine type maximum power level. In this case, the control schedule of the wind turbines or wind power plant is further constrained such that for any given period of time within the schedule, when the power of all turbines is added together, the sum of the powers does not exceed the amount of power that the connection from the power plant to the power grid is able to carry.

The embodiments described herein rely on analysis of turbine attributes and turbine site attributes in determining a control schedule for a turbine. Various calculations, including those performed by the site check software, may be performed offline in one or more different computing systems, and the resulting control arrangement provided to the wind turbine or power plant controller. Alternatively, the calculation may be performed online at the wind turbine controller or the power plant controller.

The above-described embodiments are non-exclusive and one or more of the features may be combined or cooperate to obtain improved over-rating control by setting a maximum power level for each wind turbine in a wind power plant, the improved over-rating control taking into account environmental and site conditions to which the wind turbine is exposed or which have an impact on the wind turbine.

It should be noted that embodiments of the present invention may be applied to both constant speed and variable speed turbines. Turbines may employ active pitch control whereby power limits above rated wind speed are achieved by feathering, which involves rotating part or all of each blade to reduce the angle of attack. Alternatively, the turbine may employ active stall control which achieves a power limit above rated wind speed by pitching the blades to stall in a direction opposite to that used in active pitch control.

While embodiments of the present invention have been shown and described, it will be understood that such embodiments are described by way of example only. Numerous variations, modifications, and substitutions will occur to those skilled in the art without departing from the scope of the invention as defined by the appended claims.

Claims

1. A method of generating a control schedule for a wind turbine in a power plant, the control schedule indicating how the turbine maximum power level varies over time, the method comprising:

determining a value indicative of a current remaining fatigue life of the turbine or one or more turbine components based on measured wind turbine site and/or operational data;

applying an optimization function that changes an initial control schedule to determine an optimized control schedule by changing a tradeoff between energy capture and fatigue life consumed by the turbine or the one or more turbine components until an optimized control schedule is determined, the optimization comprising:

estimating a future fatigue life consumed by the turbine or turbine component for a duration of the altered control schedule based on the current remaining fatigue life and the altered control schedule; and

wherein the optimizing further comprises changing an initial value of wind turbine life and changing an initial value of a number of component replacements to be performed on the one or more components over the course of the schedule to determine a combination of the number of component replacements for the one or more turbine components and the target minimum wind turbine life.

2. A method according to claim 1, wherein the initial control schedule specifies a relative change over time in the turbine maximum power level to which the turbine can operate.

3. The method of claim 1, further comprising:

the control arrangement is optimized by changing the timing of the component replacements and changing the number of component replacements until a maximum number is reached.

4. The method of claim 1, wherein the one or more turbine components that are replaceable include one or more of: a blade, a pitch bearing, a pitch actuation system, a hub, a main shaft, a main bearing, a gearbox, a generator, a converter, a yaw drive mechanism, a yaw bearing or a transformer.

5. A method according to claim 1, wherein the input constraints further comprise an upper maximum power output of the turbine and/or a minimum power output of the turbine allowed by the turbine design.

6. The method of claim 1, wherein determining a value indicative of the current remaining fatigue life of the turbine or one or more turbine components comprises: sensor data from one or more turbine sensors is applied to one or more life usage estimation algorithms.

7. The method of claim 1, wherein determining a value indicative of the current remaining fatigue life of the turbine or one or more turbine components comprises: data from the condition monitoring system is used.

8. The method of claim 1, wherein determining a value indicative of the current remaining fatigue life of the turbine or one or more turbine components comprises: data obtained from wind farm sensors is used in conjunction with a site check program that determines loads acting on turbine components based on the wind farm sensors and parameters related to the wind farm and the wind turbine design.

9. The method of claim 1, wherein the optimization of the control schedule comprises:

the control arrangement is changed to minimize the levelled energy costs LCoE.

10. The method of claim 9, wherein the flattened energy costs are determined using a flattened energy costs model comprising parameters for one or more of:

a capacity factor indicative of energy generated over a period of time divided by energy that could be generated if the turbine were to operate continuously at rated power over the period of time;

availability indicating a time at which the turbine will be available to generate power; and

a field efficiency indicating the energy generated over a period of time divided by the energy that could be generated if the turbine were operating in wind completely undisturbed by the upstream turbine.

11. The method of claim 10, wherein the model further comprises parameters for one or more of:

costs associated with replacing one or more components, including turbine downtime, labor and equipment for component replacement, manufacturing or refurbishment costs for replacement components, and transportation costs for transporting refurbished or replaced components to the power plant; and

service costs associated with replacement of worn parts.

12. A method according to any of claims 1 to 11, wherein the optimised control schedule is a schedule of maximum power levels to which the turbine can operate.

13. A method according to any of claims 1 to 11, wherein the control schedule indicates an amount of fatigue damage that should be caused over time, the method further comprising operating the wind turbine based on one or more LUEs to cause fatigue damage at a rate indicated by the control schedule.

14. A method according to any of claims 1 to 11, wherein the control schedule specifies a maximum power level above the rated power of the wind turbine.

15. A method according to any one of claims 1 to 11 wherein the control schedule indicates how the turbine maximum power level varies over the lifetime of the turbine.

16. The method according to any of claims 1 to 11, further comprising providing the optimized control schedule to a wind turbine controller or a wind power plant controller to control the power output of a wind turbine.

17. The method of any one of claims 1 to 11, wherein the method is repeated periodically.

18. The method of claim 17, wherein the method is repeated daily, monthly, or yearly.

19. A controller for a wind turbine or wind power plant configured to perform the method of any of claims 1 to 18.

20. An optimizer for generating a control schedule for a wind turbine in a power plant, the control schedule indicating how the turbine maximum power level varies over time, the optimizer comprising:

wherein the optimization module is configured to:

optimizing the control schedule by minimizing or maximizing the operational parameters received at the optimization module that are dependent on the set of variables by changing one or more of the variables from their initial values as a function of the remaining fatigue life of the turbine or the one or more turbine components and the one or more constraints; and

the optimized control schedule is output and,

wherein the optimization module is further configured to change an initial value of wind turbine life and an initial value of a number of component replacements to be performed on the one or more components over the course of the schedule to determine a combination of the number of component replacements for the one or more turbine components and a target minimum wind turbine life.

21. An optimiser according to claim 20 wherein the initial control schedule specifies the relative variation over time of the turbine maximum power level to which the turbine can operate.

22. The optimizer of claim 20, further comprising an initialization module configured to receive initial values of the set of variables and sensor data from turbine sensors, the initialization module configured to calculate initial values of the operating parameters.

23. The optimizer of claim 20, 21 or 22 wherein the one or more turbine components that are replaceable are one or more of: a blade, a pitch bearing, a pitch actuation system, a hub, a main shaft, a main bearing, a gearbox, a generator, a converter, a yaw drive mechanism, a yaw bearing or a transformer.

24. The optimizer of any of claims 20 to 22, wherein the operating parameter is a flattened energy cost of the turbine and optimizing the control schedule comprises minimizing the flattened energy cost.

25. The optimizer of claim 24, wherein the flattened energy costs are determined using a flattened energy costs model comprising parameters for one or more of:

26. The optimizer of claim 25, wherein the model further comprises parameters for one or more of:

service costs associated with replacement of worn parts.

27. A controller comprising an optimizer as claimed in any of claims 20 to 26.

28. A wind turbine comprising a controller according to claim 27.

29. A wind power plant comprising a controller according to claim 27.

30. A method of generating a control schedule for a wind power plant comprising a plurality of wind turbines, the control schedule indicating for each wind turbine how the maximum power level varies over time, the method comprising:

determining a value indicative of a current remaining fatigue life of each of the turbines or one or more components of each of the turbines based on measured wind turbine site and/or operating data;

applying an optimization function that changes an initial control schedule of each of the turbines to determine an optimized control schedule by changing a tradeoff between energy capture and fatigue life consumed by each of the turbines or the one or more components of each of the turbines until an optimized control schedule is determined, the optimization comprising:

estimating future fatigue life consumed by the turbine or turbine component for the duration of the altered control schedule based on the current remaining fatigue life and the altered control schedule using a site check program that determines loads acting on turbine components and including interactions between turbines of the wind power plant based on data obtained from wind power plant sensors and parameters related to the design of the wind power plant and the wind turbines; and

wherein the optimizing further comprises changing an initial value for each of the wind turbine lifetimes and changing an initial value for a number of component replacements to be performed on one or more components of each of the wind turbines over the course of the schedule to determine a combination of the number of component replacements for one or more turbine components of each of the wind turbines and the target minimum wind turbine lifetime for each of the wind turbines.

31. A method according to claim 30 wherein the initial control schedule specifies for each turbine the relative change over time of the turbine maximum power level to which the turbine can operate.

32. A method according to claim 30, wherein the sensor data comprises sensor data collected prior to commissioning and/or construction of the wind turbine or wind power plant.

33. A method according to any one of claims 30 to 32 wherein the optimisation function varies the number of times components can be replaced over the remaining life of the turbine for one or more of the turbine components.

34. A method according to claim 33 wherein the optimisation function makes changes to one or more of the turbine components to when the components can be replaced during the remaining life of the turbine.

35. A method according to any of claims 30 to 32, wherein the method is further constrained such that for any given period of time within the schedule, when the power of all of the turbines is added together, the sum of the powers does not exceed the amount of power that can be carried in the connection from the power plant to the power grid.

36. A wind power plant controller configured to perform the method of any of claims 30 to 35.