Continuous reactive power support for wind turbine generators

ES 2 593 005 T3

DESCRIPTION

Continuous Reactive Power Support for Wind Turbine Generators

The invention relates to the generation and distribution of electrical power. More particularly, the invention relates to the compensation of electrical power sources for reactive loads.

Power transmission and distribution networks transmit electrical energy from generation facilities to end users. Managing the stress in the transmission and distribution system is an important consideration for the operation and design of the system. In a typical system the reactive power flow has a strong influence on the voltage. Reactive power flow can be influenced by the generator source, changes in the transmission and distribution system, the addition of reactive shunt elements, and loads. On the other hand, excessive flow of reactive power can raise voltage and put undue stress on transmission lines, transformers, and other electrical components.

With reference to Figures 1, 2 and 3, electrical power has at least two characteristics relevant to power distribution: voltage and current. In a large-scale power distribution network both voltage and current vary with time. When the instantaneous voltage is multiplied by the instantaneous current, the result is the instantaneous power. In most power distribution networks the voltage and current signals are in the form of a sine wave.

If the reactive power flow (i.e., VAR) is zero, the voltage and current waves are in phase as illustrated in Figure 1, where ν (ωί) is the time-varying voltage waveform e ί (ωί) is the time-varying current waveform. However, if the reactive (i.e. inductive or capacitive) power is not zero, the voltage waveform, v (ut), will not be in phase with the current waveform, ί (ωί). The amount by which the current lags or leads the voltage can be quantified by a power factor angle,, which is representative of the fraction of a cycle by which the current lags or leads the voltage. One cycle is 2π or 360 °, and the power factor angle, Φ, is the difference between the cycles of current and voltage.

With respect to a constant voltage waveform, ν (ωί), a lagging current is illustrated as ί (ωί - Φ) in figure 2 and a conducted current is illustrated as ί (ωί + Φ) in figure 3 The amount by which the current lags or leads the voltage can be quantified by a power factor angle Φ, which is representative of the fraction of a cycle by which the current lags or the voltage is conducted. One cycle is 2π or 360 °, and the power factor angle, Φ, is the difference between the cycles of current and voltage.

The reactive power factor is important from a power supply point of view. Since most transmission systems are inductive, increasing the reactive current component (i.e. capacitive VARs) will cause the voltage to rise. Conversely, decreasing the reactive power component (i.e. inductive VARs) will cause the voltage to decrease.

Reactive power flow control of the wind farm can be achieved by the individual wind turbine generator, shunt elements (for example, switching capacitors or switching reactors), transformer tap changers, or some combination of these.

US 5,225,712 discloses a variable speed wind turbine with reduced power fluctuation and a static VAR mode of operation. Multiple wind turbines can be connected to a power grid through an energy collection center. The wind turbine comprises an inverter that can control the reactive power output as a power factor angle or directly as a number of VARs independent of the actual power. Reactive power can be controlled in an operating mode when the wind turbine is generating power or in a static VAR mode when the wind turbine is not operating to produce actual power.

Sáenz JR et al, Reactive power control of a windfarm through different control algorithms, Conference Proceedings, vol. 1, October 22, 2001, pages 203-207, XP010571522 discloses the control algorithms that take into account the operating limits of the machine. A calculation is made of the reactive power necessary to maintain the electrical parameters of the network. Based on this calculation, the reactive power that each machine must generate is sent to the machine as the reactive power setpoint to follow.

The invention is defined by the features of the independent claims. Preferred embodiments are defined in the dependent claims.

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

Figure 1 illustrates a waveform of voltage and current in phase.

Figure 2 illustrates a main current voltage waveform.

Figure 3 illustrates a lag current voltage waveform.

ES 2 593 005 T3

Figure 4 is a block diagram of a closed loop embodiment of a wind turbine system in which the wind turbine generators can be individually controlled to provide reactive power support.

Figure 5 is a block diagram of an open loop embodiment of a wind turbine system in which the wind turbine generators can be individually controlled to provide reactive power support.

Figure 6 is a conceptual illustration of the active and reactive power capacity curve for a wind turbine.

Figure 7 is a dynamic reactive power control flow diagram of individual wind turbines within a wind turbine system.

Figure 8 is a block diagram of an optimization controller.

Figure 9 is an example transfer function for providing a desired voltage profile at a fixed location in the power system.

Figure 10 is an optimization comparison example for a simple network.

Figure 11 is the simple network corresponding to the optimization comparison of Figure 10.

Figure 12 is a block diagram of a doubly fed induction generator system.

In the following description, for purposes of explanation, numerous specific details are set forth in order to provide a complete understanding of the invention. It will be apparent, however, to one skilled in the art that the invention can be practiced without these specific details. In other cases, structures and devices are shown in block diagram form in order to avoid obscuring the invention.

The technique described in this document provides the possibility to use the full capacity of a wind turbine generator system (ie, a wind farm) to provide dynamic VAR (reactive power support). The VAR support provided by individual wind turbine generators in a system can be dynamically varied to suit application parameters.

Wind turbine generators can provide VAR support based on actual power generation and power factor. This type of VAR support can be described, for example, by the equation:

VAR = watt * tan (θ) where θ is the power factor angle. Power factor control has some shortcomings. Since the VAR support is proportional to the square of the power output this technique does not use the full capacity of the respective wind turbine generators, as shown in figure 6. Additionally, power factor control can sometimes result in the wrong measure being taken to maintain a desired voltage at the common point of coupling (PCC) itself.

A voltage controller controls the common point of coupling (PCC) between a wind turbine generator system (eg, a wind farm) and a connection to a power grid. The voltage controller measures the mains voltage and compares the mains voltage to a desired voltage. The voltage controller calculates the amount of reactive power required from the wind farm system so that the grid voltage is within a desired range.

In one embodiment, to provide the desired power (including reactive power) at the PCC, a dynamic voltage controller transmits reactive power commands to individual wind turbine generators through a distributed control network. Wind turbines interpret received commands and excite their generator to produce the requested reactive power. When reactive power changes, the measured grid voltage moves towards the desired voltage level. Therefore, the system provides a closed-loop tension control system.

A wind turbine generator consists of a rotating electrical machine mechanically coupled to the turbine blades. The mechanical energy of the turbine is converted into electrical power supplied to the electrical network through the collector system. An electronic power converter is used to control the flow of real and reactive power.

In Figure 12, the generator is a doubly fed induction generator with a wound rotor and slip rings. A variable frequency power converter drive system connected to the generator rotor allows the generator to run (for example) at speeds ranging from 800 rpm to 1,600 rpm. The variable frequency drive power converter system is also used to adjust the reactive power output of the generator.

ES 2 593 005 T3

For reactive power generation, the response time of the frequency converter generating system is equivalent to a static VAR regulator. The power converter can also be controlled, independently of the generator, to contribute reactive power to the collector system as shown in figure 12.

Figure 4 is a block diagram of a closed loop embodiment of a wind turbine system in which wind turbine generators can be individually controlled to provide reactive power support. A wind turbine system 400 is illustrated with two wind turbines (410 and 412), each including a generator. However, any number of wind turbines can be included in a system using the techniques described herein.

Individual wind turbines are electrically coupled to common point of coupling (PCC) 420. Since many wind turbine systems include a large number of wind turbines spread over a large area, the distance between each wind turbine and pCc 420 can vary.

A 430 measurement system is coupled with the PCC 420. One of the functions of the 430 measurement system is to monitor voltage, current and power at the PCC 420. The 430 measurement system provides signals to the filters 440 and 441, the compensator bypass line 445 and to the voltage controller 450 based on the voltage in the PCC 420. Other factors can also be included in the signals provided to filter 440, line bypass compensator 445, and voltage controller 450.

Line bypass compensator 445 is an optional component that can be used to compensate for voltage drops caused by transmission from PCC 420 to utility grid 430. In one embodiment, the compensation includes the effects of the load on the power supply. line. In one embodiment, the relationship between the voltage (Es) and current (Is) measurements at the PCC and the calculated voltage (Er) in the power supply network is: E_r = A E_s + B · Is where A and B are complex coefficients derived from transmission line parameters (eg line impedance and shunt reactance). Figure 9 is an example transfer function for providing a desired voltage profile at a fixed location in the power system. In general, the shape of the transfer function may be different for each application and is determined based on detailed studies of the power system.

Measurement system 430 provides a signal to filter 440 that corresponds to the voltage measured at PCC 420; and a signal to filter 441 that corresponds to the current measured at PCC 420. Filter 440 filters out harmonics and noise, and provides a filtered signal corresponding to the voltage at PCC 420 to voltage controller 450 and shunt compensator. line 445. The filter 441 filters out harmonics and noise, and provides a filtered signal corresponding to the current in PCC 420 to the line bypass compensator 445. The line bypass compensator 445 is an optional component that can be used to compensate for voltage drops. voltage caused by transmission of pCc 420 to utility 430 or within system 400. In one embodiment, the output signals from filter 440 and line bypass compensator 445 are combined to provide an input signal to voltage controller 450.

The PI controller 450 determines the desired reactive power to be provided to PCC 420. The reactive power to be provided can be in terms of the system 400 as a whole, individual wind turbines or groups of turbines.

The optimization controller 460 is a component that receives a power setting signal from the power modulator 470, a VAR signal from the wind farm from the voltage controller 450, and a VAR setting signal from the condensation manager / power reactor. derivation 480. Optimization controller 460 calculates an individual reactive power command for each wind turbine that minimizes wind farm system losses or optimizes collector system voltage distribution. An optimization controller is described in greater detail below with respect to Figure 8.

Mathematically an optimization problem is generally described by minimizing an objective function J (u, x) subject to some constraint conditions. In matrix notation:

Minimize J (u, x)

Subject to: d (u, x) = 0 e (u, x) <0 x: system variables (for example, bus voltages, power factor) u: control variables (for example, generator reactive power)

J (u, x): objective function d (u, x): equality constraints

ES 2 593 005 T3 e (u, x): inequality constraints

A sample objective function is shown in equation (1). This function is designed to minimize losses in the distribution line (PL) subject to maintaining the power factor at the PCC of the wind farm. It may also be desirable to establish a performance hierarchy, applying a tighter stress tolerance band on a subset of nodes.

n

Minimize J = PL Equation (1) k = 1

Subject to: PF system = 0.95 (over excited)

As an illustration, a simple optimization comparison is shown in Figure 10 for the simple network of Figure 11.

Some wind farm applications may require the addition of switched capacitors 404 and switched reactors 406 within the wind farm. The condensation manager / bypass reactor 480 is an optional component that coordinates and optimizes the operation of these switching elements with the reactive power output of the wind turbines 410 and 412. A transform bypass switch 402 can also be coordinated with switching reactors 406, capacitors 404, and wind turbine VAR signals.

System 400 wind turbines receive power commands from voltage regulator 450 and optional optimization controller 460, and react to the commands individually. Power setting and VAR commands can be distributed to the wind turbines through a shared data bus with each wind turbine that has an address or other identification. Alternatively, the VAR and power setting commands can be distributed to the wind turbines through individual connections, for example through a hub device.

When wind turbines react to commands, the individual wind turbine control system produces the necessary changes (for example, blade pitch angle changes, generator torque changes) to provide the active and reactive power indicated by the power commands. As a result of the control of individual wind turbines, the real and reactive power in PCC 420 can be dynamically adjusted to provide the desired characteristics, increasing the performance and side benefits of the 400 wind turbine system. Benefits include, but are not limited to limit to; flicker reduction, voltage management and power limiting, and electrical system stabilization.

Figure 5 is a block diagram of an open loop embodiment of a wind turbine system in which wind turbine generators can be individually controlled to provide reactive power support. Wind turbine system 500 includes wind turbines (410 and 412) and PCC 420 along with an electrical supply network 430 as described above. As with the system 400 of Figure 4, any number of wind turbines can be included.

The measurement system 550 monitors the power supplied to the PCC 420 by the wind turbines. The measurement system 550 provides a signal to the filter 560 that corresponds to the power measured in the PCC 420. The filter 560 filters out fast current fluctuations, and provides a filtered signal corresponding to the power in the PCC 420 of the transfer function. VAR / watts 590. In one embodiment, the VAR / watt transfer function 590 is a constant power factor characteristic. The VAR / watt transfer function approximates a desired voltage profile at a point in the power system. An example of a VAR / watt transfer function is illustrated in Figure 9.

The VAR / watt transfer function 590 compares the power signal (P) from filter 560 to the VAR / watt curve to dynamically determine the reactive power to be provided by the system 500.

In one embodiment two signals (Q y) are transmitted by the VAR / Watt transfer function

590 to wind turbines (410 and 412) for local control. Local control of wind turbines (410 and 412) is of the form

<img file="ES2593005T3_D0001.tif" />

where Q is a common reactive power term that is provided to all wind turbine generators in a system and

ES 2 593 005 T3

<img file="ES2593005T3_D0002.tif" />

is a slope term that can be used for fast dynamic control by local control of wind turbines (410 and 412).

Figure 6 is a conceptual illustration of the active and reactive power capacity of a wind turbine generator. Reactive power can be dynamically adjusted within the limits shown in figure 6.

The system configurations of Figures 4 and 5 provide several important features. The capacity of each wind turbine generator in a system to operate as a static VAR compensator can be utilized within the capacity curve shown in Figure 6. In one embodiment, the reactive power compensation capacity of a VAR converter Wind turbine power can be used when the turbines are not running. In one embodiment, the damping power system (for example, power swing angle, frequency, and rate of change of power swing angle) can be controlled by modulating real and reactive power generation at the individual generator level. or at the system level.

Figure 7 is a dynamic power control flow diagram of individual wind turbines within a wind turbine system. The desired utility control signal (eg, voltage, power factor, or VARs) to be provided at a predetermined location is determined, 710. The default location is usually the Common Coupling Point (PCC); however, if, for example, the wind turbine system is located remotely, the predetermined location may be a projected point based on a fall line compensation algorithm.

The wind farm power or voltage output supplied to the predetermined location is measured, 720. The measured output is compared to the desired control signal, 730. In response to the comparison, the control system for the wind farm determines the magnitude any corrections that may be necessary to provide the desired power to the predetermined location.

The control system determines the active and reactive power to be provided by each wind turbine generator in order to provide the desired control at the predetermined location. Each wind turbine generator can be commanded to provide a different combination of real and reactive power. The physical configuration of the wind farm including, for example, the types of generators, the placement of the wind turbines, the design of the collector system, and the distances between the wind turbines and the PCC can be used to determine the power of the commands to be provided to individual wind turbines in order to improve system performance for loss reduction and stress profile.

Power commands are transmitted to individual wind turbines, 740. Power commands can be transmitted using any means, whether wired or wireless, known in the art. Any protocol known in the art capable of transmitting commands to individual destinations within a group of possible destinations can also be used. Individual wind turbine generators modify their respective outputs, if necessary, in response to power commands, 750.

Figure 8 is a block diagram of an optimization controller. Block 810 implements a deadband characteristic in the error between the wind farm VAR signal and the VARs generated by the switched capacitors and reactors. The time integral of block 820 is used to establish an inverse time characteristic that initiates the switching of the capacitor and reactor banks. Block 830 is used to determine which capacitor or reactor bank to change.