GB2410386A

GB2410386A - Controlling reactive power output

Info

Publication number: GB2410386A
Application number: GB0401375A
Authority: GB
Inventors: Phillip Cartwright
Original assignee: Areva T&D UK Ltd
Current assignee: UK Grid Solutions Ltd
Priority date: 2004-01-22
Filing date: 2004-01-22
Publication date: 2005-07-27
Also published as: GB0401375D0

Abstract

A wind farm comprises at least one wind turbine and the combination of a voltage controller and a tap-changing transformer arranged to control a reactive power output of the wind farm. The tap-changing transformer is arranged to transform an output voltage of the wind farm at a predetermined position in the network. The controller has a first input arranged to provide a measure of the output voltage of the wind farm at the predetermined position and a second input representative of a desired reference voltage set by a network operator. The controller is arranged to monitor the input and generate a first output to the tap-changing transformer to change its tap setting and keep the voltage at the predetermined position within predetermined limits, and to generate a second output to control the reactive power output of the wind farm.

Description

A WIND FARM AND METHOD OF CONTROLLING A WIND FARM

This invention relates to a method of controlling at least one windturbine and particularly, but not exclusively, to a method of controlling at least one doubly fed induction generator (DFIG) based wind-turbine. The invention also relates to apparatus for providing such a method.

Due, at least in part, to concerns about emissions from burning fossil fuels and the depletion of fuel resources, renewable energy systems are a topic of great interest and investment. In particular, wind farms are now considered a viable energy source and their utility and efficiency has been the subject of much recent research and development.

Most early wind farms used asynchronous generators with fixed speed and stall control of the blades. Such wind farms are directly coupled to a power grid, with the connection being present when the wind speed (and therefore the electrical power produced) is sufficient to supply energy to the grid.

When the power falls below a set amount (which may be zero as, if left connected, the wind-turbine will operate as a motor) they are de-coupled from the grid. The performance of fixed speed wind-turbines depends on the characteristics of the mechanical system and often provide variable output power which can lead to instability in the grid voltage.

In order to overcome the problems associated with fixed speed systems, new wind farms tend to use variable speed turbines that utilise an active voltage source. There are two types of generators used in variable speed generation. Fully Fed Synchronous Generators (FFSGs) and Doubly Fed Induction Generators (DFIGs).

# a; A method of controlling a DFIG based wind farm is desirable to allow it to comply with wind farm connection requirements, or grid connection codes, that are set by the power supply grid operators. Such a method may also have commercial benefits in that the commercial viability of winddriven embedded renewable generation is sensitive to connection costs. These costs are related to the voltage level at which the wind generator is to be connected. The wind farm developer will prefer to connect at the lower voltage as this is associated with lower connection costs. However, the grid operator may wish the connection to be implemented at the higher voltage so as to more closely match the existing supply; any impact on the distribution system's voltage profile is then minimal. Employing a voltage control strategy may allow generators to be connected while generating a lower voltage and such a new connection may make a positive contribution to the operation of the grid.

Active voltage control of networks containing distributed generation (i.e. a number of small wind farms contribute) offers several advantages when compared with a passive network or a network with a so called 'connect and forget policy' where the distributed generation is considered a negative load. Introducing such a strategy to a network considerably reduces the generation that is 'wasted' due to voltage constraints and thus increases the net generation.

DFIGs have been modelled, for example, in Ekanayake J B. Holdsworth L, Wu X G. Jenkins N. "Dynamic Modelling of Doubly Fed Induction Generator Windturbines", IEEE Transactions on Power Systems, Vol l, p 803-809, May 2003. In that article, the converters and their control system were assumed to be 'ideal'. The d.c. side voltage was assumed to be constant and dynamic linear time invariant models of the DFIG including its associated voltage source converters and controllers are derived in the synchronous d-q reference frame. A local on-load tap changer is modelled as a finite state machine.

e . 68. e' ' ^ e A e e, Another voltage control strategy was proposed in Ledesma P. Usaola J: O"Contribution of Variable-Speed Wind-turbines to Voltage Control", Wind Engineering, Volume 26, No. 6, p 347-358, 2002 which used a 'grid' side converter of a DFIG with active and reactive power control.

According to a first aspect of the invention there is provided a voltage controller arranged to control a wind-turbine in combination with a tap changing transformer arranged to transform an output voltage of a wind lO turbine, the controller having at least one input arranged to provide a measure of the output of a wind-turbine controlled by the controller the controller further having at least one output, the output being arranged to control which tap the tap changing transformer is set to, the controller being arranged, in use, to monitor the input and being capable of generating an output to cause the tap changing transformer connected thereto to change its tap setting in order to keep the output of the transformer within predetermined limits.

Such a controller and transformer combination is advantageous at it may allow the voltage output from the transformer to be maintained within the predetermined limits. The transformer will be able to respond to disturbances applied to a wind-turbine connected thereto such as a change in load as well as to changes in the generated power so as to keep its output voltage within the predetermined limits. Further, use of an on line tap changer can increase the dynamic range of a wind-turbine connected to the controller and transformer.

Preferably, the controller has a voltage and a current input arranged to measure the voltage and current output of the wind-turbine respectively.

Such an arrangement is convenient because is allows calculations to be ::; #' A:: :: it. . made that provide quantities such as reactive and/or real power, and the like.

The controller may have a second output arranged to control a converter.

Such an output is advantageous because it can be used to provided finer control of the output from the transformer. The skilled person will appreciate that changing the tap of the OLTC will provide a step change to the output voltage whereas a converter may be used to provide smooth control of the transformer output voltage within predetermined limits.

The controller may be arranged to determine the reactive power of a windturbine connected thereto and control the transformer and/or converter such that the reactive power output tends towards a predetermining value. This is advantageous as the output of the turbine will generally be for use to supply electricity to a power supply grid and such supply will have to be of a reactive power set by the grid operator and it is therefore advantageous to meet this requirement.

The controller may be arranged, in use, to use a dynamic linear time invariant model of the wind-turbine in order to.generate the outputs. This has the advantage that the effect of the interaction between the phases may be neglected.

Preferably, the controller is arranged to compare the input to its input to a predetermined limit and, if the input exceeds the predetermined limit, the controller is arranged, as a first step, to generate an output on its second output in an attempt to cause the input to be within the predetermined limit and, if the input to the input still exceeds the predetermined limit, the controller Is arranged, as a second step, to generate an output on its first output in order to change the tap setting of the transformer. Such an arrangement may be thought of as trying to control the output from the 8 C 8 a a transformer, within an envelope provided by the current tap setting, using a converter and if the output cannot be controlled within this envelope changing the tap setting to provide a different envelope.

According to a second aspect of the invention there is provided a controller for use in the combination of the first aspect of the invention.

According to a third aspect of the present invention there is provided a system comprising a wind-turbine capable of producing an output, at least to one converter, at least one tap changing transformer and a controller having an input, the controller further having a tap changing output arranged to control which tap the tap changing transformer is set to, and a second output arranged to control the converter wherein in use of the turbine the controller is arranged to control the tap changer of the transformer in coordination with the at least one converter in order that the output voltage and/or current tend towards predetermined limits.

In a preferred embodiment the controller has a voltage input and a current input, respectively arranged to have the output voltage and current of the wind-turbine input thereto. Having both current and voltage inputs is convenient because it allows parameters such as power (real and/or imaginary) to be calculated.

Preferably, the converter is a rotor side converter. This is advantageous as controlling the active and/or reactive power via the rotor side allows for a lower converter rating than the machine rating.

The system may further comprise a low pass filter arranged to filter the output current before it is input to the controller. This is advantageous as, when employing a low-pass filter, the controller should not respond to transients.

. . e.

Preferably, the turbine comprises a Doubly Fed Induction Generator (DFIG). When used with a wind-turbine, a DFIG offers several advantages when compared with fixed speed generators. These advantages, include speed control and reduced flicker, which are primarily achieved via control of the voltage source converter within the DFIG. A voltage source converter has inherent four-quadrant active and reactive power capabilities and may be used to control the active and reactive power on the rotor side.

Taking into account the amplification factor of the inverse slip ratio of a DFIG, the converter may control the active and reactive power on the stator side. Controlling the active and reactive power via the rotor side allows for a converter rating typically 25% of the machine rating rather than a potential 100% rating if control is realised using the grid side converter.

The DFIG in combination with the converter may provide a smooth voltage control. The tap changer may provide coarse control with step changes.

The controller may be arranged to provide dynamic reactive power control.

The converter may be a voltage source grid side converter.

The system may also comprise a pulse width modulator (PWM). The PWM may be driven by the controller and arranged to control the converter.

The system further comprise a second converter. The second converter may be provided as a grid side converter arranged to control the direct voltage across a d.c. link. The d.c. link may be provided by a capacitor.

The second converter may be controlled by a second controller. In one embodiment the second controller may be arranged to drive a PWM. The PWM may be arranged to drive the second converter.

c . c . . . be. e According to a fourth aspect of the invention, there is provided a method of controlling the output of a generator comprising measuring a parameter indicative of the output, comparing the measured parameter with at least one predetermined value and if the measured parameter falls outside a predetermined limit of the predetermined value controlling a tap changing transformer to try and bring the output back within the predetermined limit of the predetermined value.

The method may also use a converter to try and maintain the output within the predetermined limit of the predetermined value. An advantage of using the converter in this manner is that smoother control may be achieved when compared to only using a tap changing transformer.

Generally the predetermined value will be the voltage of a grid to which the generator is connected. The predetermined limits may be a maximum (plus and/or minus) variation from the voltage of the grid, generally set by the operator of the grid.

In a preferred embodiment the method tries to maintain the output within the predetermined limit of the predetermined value using the converter and if this cannot be achieved the method may then alter the tap at which the tap changing transformer is set. Thus, the converter may be used to affect change within an envelope of change and changing the tap setting of the transformer may alter the envelope.

The method may allow a plurality of generators to be controlled generally by a single tap changing transformer.

Conveniently, the method monitors the outputs of each of the generators being controlled. In the preferred embodiment each generator controlled by cc. e e t a e e a r ee. a e * ce.

e e e .. e e the method is provided with a converter having an associated controller that is capable of controlling the output of that generator. Such a method provides some control of each of the generators.

The method may further comprise providing a supervising controller to monitor the controllers of each of the generators. The skilled person will appreciate that it would be possible to arrange a single controller, perhaps the supervising controller, to control each of the converters associated with generators monitored by the method.

The method may implement a control strategy in which the tap of the tap changing controller is changed only when the output can not be maintained within the predetermined limit of the predetermined value by altering any one of the converters. This may be thought of as altering the tap of the tap changing transformer only when the converter of all (or substantially all) the generators has saturated. Such a method is convenient because it provides maximum control within the envelope provided by a tap of the tap changing transformer and as such should reduce the number of tap changes made by the transformer.

The method may be arranged to alter the tap on the tap changing transformer only once substantially all of the generators have saturated.

The method may model the tap changer transformer as a finite state machine.

According to a fifth aspect of the invention, there is provided a wind farm, the wind farm comprising a controller, a plurality of wind-turbines each having a converter associated therewith and a tap changing transformer wherein said controller is arranged to calculate a total output for the wind farm and compare the total output to a desired total output, the controller e.e e e e e ret ece . e A- e being arranged such that, if the total output does not match the desired total output then it is arranged to control the converters in order to attempt to maintain the total output at the desired total output and further arranged such that if control of the converters fails to maintain the total output at the desired total output then it is arranged to cause the tap of the tap changing transformer to be changed.

Embodiments of the invention are now described by way of example only and with reference to the accompanying figures of which: Figure 1 shows a simplified schematic network; Figure 2 shows a typical configuration of a DFIG wind-turbine arranged to carry out one embodiment of the present invention; Figure 3 shows a schematic representation of a portion of the configuration shown in Figure 2; Figure 4 shows a tap changer transformer; Figure 5 shows a representation of a steady state On Line Tap Changer; Figure 6 shows a State chart for a tap changing transformer; Figure 7 shows an overview of the control system; Figure 8 shows a Speed control reference of the DFIG; Figure 9 shows an exemplary Transmission System; . * . #. .:. he:: :: e e. Figure 10 shows an exemplary distribution System; Figure 11 shows the results seen in use of a Transmission System; Figure 12 shows the results seen in use of a distribution System; Figure 13 shows the output of an example system according to the present invention; and Figure 14 shows a series of possible control envelopes.

Throughout the following description and in the figures, the following notation is employed: Vvsc = Rotor side converter voltage Vvsc2 = Grid side converter voltage La = Grid side converter inductance is = Grid side converter a.c. current Vdc = DC voltage C = DC capacitance ic = Current through capacitor C Vs. Vr Van = Stator, rotor and double-cage voltage is, ir = Stator and rotor current Rs, Rr = Stator and rotor resistance R,, Rn = Double-cage machine and grid side converter resistance 6's, rib = Synchronous and rotor angular frequency = Flux linkage Lm = Magnetising inductance Lrm = Mutual inductance between two rotor coils a a e C a a . L5 Lr Ld = Stator, rotor and double-cage leakage inductance Lss Lrr Ldd = Stator, rotor and double-cage self-inductance s = Rotor slip J = Moment of inertia of entire wind-turbine Tm' Te' Tip = Mechanical, electromagnetic, set point torque Topt = Optimal torque Kopt = Optimal torque/speed constant of the wind-turbine PAC' Power on a. c. side of the rotor side of Converter 4 Pact = Power on a.c. side of the grid side of Converter 2 M = Modulation index H = Lumped inertia constant * = Reference value for controller (as a superscript) d, q = Synchronous d-q axis (as a subscript) Figure I shows a simplified network with a local load 102, a distributed generator (DG) and potential power factor compensation connected at a second busbar 2 via a complex impedance 104 to a first busbar 1. The first busbar 1 is connected via an On Load Tap Changer (OLTC) transformer to a point of common coupling (PCC). As the person skilled in the art will be aware, a tap changer is a device that is fitted to power transformers for regulation of the output voltage to required levels. This is normally achieved by changing the ratios of the transformers on the system by altering the number of turns in one winding of the appropriate transformer/s. Tap changers offer variable control to keep the supply voltage within set limits. On load tap changers generally consist of a diverter switch and a selector switch operating as a unit to effect transfer current from one voltage tap to the next.

. c. e e cce e * Consideration of the network of Figure 1 allows the following relationship to be developed Vat EVE_ ( g PL)+X(+Qg-QL) (I) From equation (1) can be seen that control of the voltage at the first busbar I (Vat) or the second busbar 2 (V2) can be by altering any of the variables and/or directly by changing the tap ratio of the OLTC. Vat or V2 can be affected through control of the generator's active power Pg. reactive power Qg, both active and reactive power or control of the OLTC at Vat and the generator's active power Pg and reactive power Qg.

Under the present example, it is assumed that the reactive power requirements of the system are met within the ratings of the wind farm equipment. Control of external power factor compensation and active power control is not described herein. Within the system described here, active power control is assumed to be used for both the dispatch of active power and a frequency limiting function as described for example in Ekanayake, J B. Holdsworth, L, Jenkins N. "Control of Doubly Fed Induction Generator (DFIG) Wind-turbines", IKE Power Engineering, Vol. 17, Issue 1, Feb. 2003.

The control strategy described here utilises a co-ordinated control of the Ol,TC and the generator's reactive power Qg. Two embodiments are described hereinafter showing an a.c. voltage control strategy within a transmission network and a distribution network. For the transmission network case the OLTC is used to extend the dynamic range of the DFIG.

For the distribution case the DFIG is used for dynamic a.c. voltage control whilst the OLTC is used for steady state a.c. voltage control. .

* . #* * ce ace . . . . - - A typical configuration of a DFIG based wind-turbine 200 is shown schematically in Figure 2. DFIG based wind-turbines utilise a wound rotor induction generator 202, where the rotor winding is fed through back-to- back variable frequency voltage source converters VSCI, VSC2. The DFIG 202 and converters VSCI, VSC2 are protected by voltage limits and an over- current 'crowbar' circuit 204. The converter system enables variable speed operation of the wind-turbine by de-coupling the power system electrical frequency and the rotor mechanical frequency. The models developed herein uses dynamic linear time invariant equations for the DFIG 202, the voltage source converters VSC1, VSC2. Thc a.c. system and the transformer in the synchronous d-q reference frame as are detailed below. The transformer tap changer is represented as a finite state machine.

Thc models described here are based on the following assumptions: 1. The equations are derived on the synchronous reference frame using direct (d) and quadrature (q) axis representation.

2. The q-axis is assumed to be 90 degrees ahead of the d-axis in the direction of rotation 3. The q component of the stator voltage (Vs) used within the model is equal to the real part of the generator busbar voltage, (the generator busbar voltage is shown as Vs in Figure 2).

The mathematical models for the DFIC and the Voltage Source Converters VSC1, VSC2 are derived with reference to Figure 2.

Thc reduced order machine model in units of total power produced by the generator may be expressed as: jV,tS = Rs x ids - Is (2) tVqs = Rs x is + Aids a * , a c . . ce. .. a !lvdrRr X idr 5 X qr + dt ldr lVqr = Rr x iqr + s x dr + dt qr dd Kd X i, d 5 X qd + ,lt dd O Vqd = Rd X iql + S X dd + a) dt qd where ds Lss Lm Lm 1idS 1 dr = Lm Lrr pLm Idr (5) dd Lm {Lm Ldd _i-dd and _ Lss Lm Im iqS qr = Lm Lrr {Lm i-qr (6) qd Lm Lm Ldd _i-qd- ln equations (5) and (6): (3 = 1 + rm Lss = L, + Lm Lrr = Lr + pLm and Ldd = Id + {Lm /n From equations (5) and (6), the stator current can be derived in the Power Unit (pu) form (lpu = the total power produced by the generator) as: ! idS = (< ds m dr m tId -) = qs _ L m i-d _ Lm d qs -Lm x Iqr L-m X iqd Vd; _ L i _ m i I L s) LST Ls, L . - c: ..- as: e e e e e Using equations (3) to (6), the rotor currents can be derived in the following pu form: , X3 - RrX3 - RdX2 L,'' [ _ 1] - - - - - - [ ( 8) X2- RrX2 R,/X,- - L P[dd = -Vdr ±idr a dd q Lss svd5 + -pVqS [ where Xl =:LLrr-L-a]' X2 =:L'(3Lm-L]' 3 L dd L,,] 69, C72 = 2 3] and Cr3 = [ ] The pu electromagnetic torque (positive for a motor) is calculated using: Te = ilds X Is Is X ids (9) = VqS iqS + Vds ids Then, if Tmis the mechanical torque, which depends upon wind speed, the machine swing equation is given by: d0r = 1 X(T -T.) (10) The a.c. side of the grid side voltage source converter shown In Figure 2 may be represented as a voltage source, (VvSc2). If the circuit up to and including the three phase stator voltage (Vs) is included then this can be a a. a cat e a e a represented by an interposing impedance consisting of resistance (Ra) and inductance (La) as shown in Figure 3.

For convenience the model is transformed into the synchronous orthogonal reference frame rotating at the supply frequency As, (via the Park transformation). This technique is described for example in Tang Yifan, Xu L, "A flexible Active and Reactive Power Control Strategy for a Variable Speed Constant Frequency Generating System, IKE Transactions on Power Electronics, Vol 10, No 4, July 1995 and the explanation of the Park transformation therein is incorporated herein by reference. The skilled person is directed to read this reference. The transform of the system shown in Figure 3 is: d = - ; fad + As in ±Us --VVSC2 (1 1) d q =- ia -5 ia ±V5 --VVSC2 where ces is the supply frequency.

The current through the d.c. side capacitor is defined as: c dt (12) Using the power balancing equation: Pdc Paul + Pace ( 13) a: :: . : ::.

. . . . . Pact 2 (Vvscld ird + Vvscl9 ir' ) ( 14) Pac2 2 (VVSC2d fad + VVSC2' in' ) Pdc C d Vdc ( 1 5) C dt Vdc 2 (VVSC2d fad + VVSC2q inq) act ( 16) We now consider the relationship between the a.c. side and the d.c. side as a continuous system as follows: 1 o Vvsc, =-M d (17) VVSC2 = V(C M2 where M2 is the ratio between the peak d.c. side voltage and the peak-to- peak, phase-to-neutral a.c. side voltage at the grid side converter (VSC2) terminals. M2d and M2q are the modulation ratio of the grid side converter (M2) in the d-q reference frame.

The equations for the rotor side converter (VSC 1) can be derived in a similar method and assuming an ideal converter model, the a.c. and d.c. side systems can be expressed as: d 'rd 1 Rr /Lr tVs 6)r 1ird 1 1 VSC'd 1 VdC M, dt Li v L-(;-mr) -Rr /Lr Ltrq L Lvvsciq 2L LMiq . ce: I. I: . ë i:- .. : : A: c d tiad 1 -Ro/La we triads 1 iVVSC2d 1Vdc M2d1 (18) d' Liar 1 l -6)5 -R,7 /La gLinq] L LVvsc2q 2L LM29] dt 4C {(Mid ird + Ml q irq)+ (M2d ' ia + M2 ' i)} The system model represented by equation (18) is non-linear because of the existence of multiplication terms between the state variables (id, iq) and the input (Mu, Ma). The operation of the VSC converters VSC1, VSC2 requires the state variables (id) and (iq) to follow varying reference points to control the reactive and active power accordingly. In addition, the direct voltage level (Vile) also has to be maintained at a set value. However, the model has only two independent inputs Mu and Mq. Hence, an exact feedback linearisation technique is not applicable since the corresponding decoupling matrix is not square. This difficulty is resolved by dividing the control into two separate loops, an inner fast current loop and an outer slow d.c. voltage loop as shown in Xu L, Andersen B R. Cartwright P. "Control of VSC Transmission Systems under Unbalanced Network Conditions", IEEE T&D Conference, Dallas, 7th to 12th Sept. 2003.

An OLTC 400 is modelled as shown in Fig. 4 and the steady state equations are derived thus: v2 = Fat ( 1 9) 3 Z (20) = 3 (21) a e e e e ece Eliminating V2 provides: I3 =_ V3 + Vat (22) Z aZ 1 =_ V3 + V' (23) aZ a2Z which may be represented as a steady state OLTC 500 as in Figure 5.

The OLTC is modelled as a finite state machine as shown in the State Chart of Figure 6. During a simulation the system is initialised at the nominal tap position and then responds to a tap up or a tap down event. When either of these events happens the tap position and the tap ratio are updated after roughly five seconds, (representing a practical OLTC). The steady state system of Figures 4 and 5 is then updated accordingly. It will be appreciated that in other embodiments periods other than 5 seconds may be used. For example roughly any of the following time periods (or any period in between may be used): 1 second, 2 seconds, 3 seconds, 4 seconds, 6 seconds, 7 seconds, 8 seconds, 9 seconds, 10 seconds or more.

An overview of the control system is shown in Figure 7. A controller 700 is used to control the rotor side converter (VSCI) to give speed and voltage control. The controller 700 has a voltage input 702, a current input 704, a first output 706 to an Online Tap Changing (OLTC) transformer and a second output 708 to the rotor side converter (VSCI).

A second controller 7 l 0 is used to control the grid side converter (VSC2) to control the direct voltage across the d.c. side capacitor 712, (with a reference of l.0pu). The d-component of the rotor current (id) is used to &. Be: I- I: . ::.- ':e. ::e vary the reactive power absorbed and generated by the DFIG as described in Muller S. Deicke M, De Doncker RW, "Doubly Fed Induction Generator Systems for Wind-turbines", IEEE Industry Applications Magazine, May/June 2002. The a.c. voltage error, the difference between the reference a.c. voltage and the measured a.c. voltage, together with a PI controller is used to obtain ird.

The optimum torque/speed curve is used, via a look up table and transfer function 'R', to provide a q-eomponent of rotor current reference (irq*).

The speed controller is implemented as described in Ekanayake J B. Holdsworth L, Wu X G. Jenkins N. "Dynamic Modelling of Doubly Fed Induction Generator Wind-turbines", IEEE Transactions on Power Systems, Vol 1, p 803-809, May 2003 and is shown in Fig. 8.

In some line commutated HVDC schemes the OLTC of the converter transformer(s) is used to ensure that the firing angle is controlledwithin a defined operating range for a given a.c. system voltage. Its operating point is conditioned for a potential external perturbation. For a line commutated converter this type of control strategy is implemented to reduce the possibility of commutation failures. A system of this type is decribed in Burgess R P. Ainsworth J D, Thanawala fI L: "Voltage/Var Control at The McNeill Back-to-back TIVDC Converter Station", CIGRE, 1990 Session, 26th Aug _ 15t Sept 1990 Under the present example, the OLTC of a local transformer is used to ensure the DFIG rotor current is controlled within a defined operating range, which is also a function of the transmitted power. This reduces the possibility of it exceeding a control limit for a change in system operating coalitions This allows dynamic voltage support during network disturbances. This is described for example in ELTRA, "Specifications for t e He.* Connecting Wind Farms to The Transmission Networks", ELTRA Transmission System Planning, Document No 74557, www.eltra.dk, 1999.

When applied to a DFIG connected to a Transmission System the following logic is implemented: IF "ir is negative" AND "ir is at limit" THEN "tap up" IF " ir is positive" AND " ire is at limit" THEN "tap down" In addition, the rotor current measurement for the OLTC controller is put through a low pass filter to ensure the OLTC does not respond to transients.

A specific embodiment of the invention is now described with reference to a distribution system.

On a distribution network the OLTCs are used to control the steady state network voltage at a given busbar. Adding the DFIG provides dynamic voltage control at the same busbar.

When applied to a DFIG connected to a Distribution System case the logic is amended thus: IF " ir is at limit" AND "AC Voltage Is Low" 1 HEN "tap up" IF "ir is at limit" AND "AC Voltage Is lligh" THEN "tap down" The implementation of this control strategy for a distribution system will require a relatively slow- on the order of seconds- data transmission system since the transformer and its OLTC could be located some distance from the turbine installation. It is also recognised that for a wind farm installation consisting of several DFIGs the individual operating parameters will not all I r 1

C

are ce. a be the same. Therefore, the rotor current may not be at the same operating point for all of the machines and the control system will require some arbitration when defining the signal "ire is at limit".

The AC voltage that is controlled by the controller / control strategy may be at any point in an electricity network connected to the wind farm and does not necessarily have to correspond to the output of the tap changing transformer. Thus, the AC voltage being High or Low in the above strategy may refer to this predetermined point in the electricity network.

An example of a system incorporating the proposed a.c. voltage control strategies described herein is provided by a DFIG rated at 2MW having the following induction wind-turbine model parameters (star equivalent circuit): Vbase = 690 V, Sbase = 2 MW, Chase = 2gfbase Abase = 50 Hz, Stator resistance (Rs) 0.00488pu, Stator leakage reactance (Xs) 0.09241pu, Rotor resistance (Rr): 0.00549pu, Rotor leakage reactance (Xr): 0. 09955pu, Double-cage resistance (Rat): 0.2696pu, Double-cage reactance (by): 0.0453pu, Airgap magnetising reactance (Xm): 3.95279 pu, Rotor to double- cage mutual reactance (Xrm): 0.02 pu, Lumped inertia constant (H): 3.5 s, Kopt =0.56, Kp2 =0.3, K,2 =05.

The control model parameters may for example be as follows: Cut-in speed = 1000 rpm, Speed limit = 1800 rpm, Shutdown Speed = 2000 rpm.

An exemplary Transmission System is provided by a 60MW wind farm connected to a Transmission System through a 33kV:132kV transformer, as shown in Figure 9.

An exemplary use of one embodiment of the present invention within a Distributed System provided by a 4MW wind farm connected within a ale I 1 1 1 Distributed System is shown in Figure 9, (which is connected to the Transmission System through a 20km, 1 lkV line section).

In an example of a both the transmission system and the distribution system S in use, the converters are initialised so that the d.c. side capacitor is fully charged. The mechanical torque is initially 0.3pu. The mechanical torque is stepped from 0.3pu to 0.8pu at t=lOs and the a.c. system voltage at the 1 lkV busbar is measured as l.Opu shortly after, (when the system is stable).

The voltage at the llkV busbar is controlled to l.Opu until the voltage I O changes.

In an example of a transmission system in use, consider a situation where the voltage steps from l.Opu to 1.05 pu at t=30s. The rotor current then exceeds the control limit and consequently the tap changer begins to tap up to reduce it. Each tap step takes 5-seconds and 4 tap steps are performed before the rotor current is back within its operating range and the measured 33kV busbar voltage equals the reference voltage. These results are shown in Fig. 1] . In an example of use of the distribution system, the voltage reference is steps from l.Opu to O.95pu at t=30s. The rotor current then exceeds the control limit whilst the 1 lkV busbar voltage is still 'high' and consequently the tap changer begins to tap down to reduce it. As in the transmission, each tap step takes 5-seconds and 4 tap steps are performed before the rotor current is back within its defined operating range and the measured llkV busbar voltage is approximately equal to the reference voltage. These results are shown in Figure 12.

An example of the dynamic performance of the DFIG is shown in Figure 13 in which a local a.c. system disturbance is applied by switching in a local cca C

C C

C C C C . . C load at t=25s. As shown in Figure 13 the voltage is quickly controlled by the DFTG based wind farm.

Embodiments having a plurality of wind-turbines and associated converters with controllers(900, 1000) (such as those shown in Figures 9 and 10) may have a supervising controller (902,1002) with which each of the controllers (900,1000) is connected. As will be seen from the Figures the supervising controller 902, 1002 has been X and Xn controllers 900, 1000 connected thereto.

In such embodiments it may be the supervising controller that determines when the tap of the tap changing transformer (OLTC) is changed. In a preferred embodiment the tap of the tap changing transformer is only changed when the total output of the wind-turbines cannot be maintained within predetermined limits by controlling the converters associated with the wind-turbines. Generally this would occur when each of the coriverters had saturated.

As can be seen in Figure 14 each tap setting of the tap changing transformer (OLTC) provides an envelope 1400 within which the converter VSCI can control the output of the wind-turbine and/or the tap changing transformer (OLTC). If the output cannot be controlled by the converter VSCI within this envelope then the envelope can be changed by changing the tap of the tap changing transformer OLTC - as represented by the change of black line.

Claims

ece e e ee e e e e e e e e e ace e e e ace e e e e e see e see e e CLAIMS

1. A wind farm comprising at least one wind turbine and the combination of a voltage controller and a tap changing transformer arranged to control a reactive power output of the wind farm when the wind farm is connected to an electricity network through the tap changing transformer, the tap changing transformer being arranged to transform an output voltage of the wind farm at a predetermined position in the network and the controller having a first input arranged to provide a measure of the output voltage of the wind farm at the predetermined position and a second input representative of a desired reference voltage set by a network operator, the controller being arranged, in use, to monitor the inputs and (a) generate a first output to the tap changing transformer to change its tap setting, thereby to keep the voltage at the predetermined position within predetermined limits, and/or (b) generate a second output to means for controlling the reactive power output of the wind farm.

2. A wind farm according to claim I in which the controller has an output arranged to control a converter.

3. A wind farm according to claim 2 in which the converter is a rotor side converter.

4. A wind farm according to claim 2 in which the converter is a voltage source grid side converter.

5. A wind farm according to any preceding claim in which the controller is arranged to compare the first input to the second input and, if the first input exceeds the second input, the controller is arranged, as a first step, to generate an output on its second output in an attempt to cause the first input to be within a predetermined limit of the second input and, if the . . . . , e e ece e ese . . . . e en- ë input to the input still exceeds the second input, the controller is arranged, as a second step, to generate an output on its first output in order to change the tap setting of the transformer.

6. A wind farm according to any preceding claim in which the controller is arranged to determine the reactive power of a wind turbine connected thereto and control the converter such that the reactive power output tends towards a predetermined value.

7. A wind farm according to any preceding claim in which the controller is arranged, in use, to use a dynamic linear time invariant model of the wind-turbine in order to generate the or each output.

8. A wind farm according to any of preceding claim which comprises a low pass filter arranged to filter the output current before it is input to the controller.

9. A wind farm according to any preceding claim in which the turbine comprises a Doubly Fed Induction Generator (DFIG).

10. A wind farm according to any preceding claim in which the controller is arranged to provide dynamic reactive power control.

11. A wind farm according to any preceding claim which comprises a pulse width modulator (PWM).

12. A wind farm according to claim 11 in which the PWM is driven by the controller and arranged to control the converter.

13. A wind farm according to any preceding claim which further comprises a second converter.

. :. : : : : < e

14. A wind farm according to claim 13 in which the second converter is a grid side converter arranged to control the direct voltage across a d.c. link.

15. A wind farm according to claim 14 in which the d.c. link is provided by a capacitor.

16. A wind farm according to any of claims 13 to 15 in which the second converter is controlled by a second controller.

17. A wind farm according to claim 16 in which the second controller is arranged to drive a PWM.

18. A wind farm according to claim 17 in which the PWM is arranged to drive the second converter.

19. A wind farm according to any preceding claim in which the wind farm is connected to the electricity network via a transmission system.

20. A wind farm according to claim 19 in which the controller is arranged to change the tap of the tap changing transformer when the rotor current in one or more of the turbines is at a predetermined limit.

21. A wind farm according to claim 20 in which the controller is arranged to cause the tap changing transformer to do at least one of the following: change down a tap when a rotor current in one or more of the turbines is negative; and tap up a tap when a rotor current in one or more of the turbines is positive.

22. A wind farm according to any preceding claim in which the wind farm is connected to the electricity network via a distribution system.

. c. . . q e. e

23. A wind farm according to claim 22 in which the controller is arranged to cause the tap changing transformer to change tap when the rotor current in one or more of the turbines is at a predetermined limit.

24. A wind farm according to claim 23 in which the controller is arrange to do at least one of the following: cause the tap changing transformer to change down a tap when the voltage at the predetermined point in the network is low; and change up a tap when the voltage at the predetermined point in the network is high.

25. A controller for use in the wind farm of any of claims I to 24.

26. A method of controlling a reactive power output of a wind farm when the wind farm is connected to an electricity network, the wind farm comprising at least one wind turbine, a voltage controller and a tapchanging transformer arranged to connect the wind farm to an electricity network, in use, the controller being arranged to compare a first input giving a measure of the output voltage of the wind farm at a predetermined position in an electricity network with a second input representative of the desired voltage at the predetermined point in the network and the method causing the controller generate a first output to the tap changing transformer to change its tap setting, thereby to keep the voltage at the predetermined position within predetermined limits, and/or (b) generate a second output to means for controlling the reactive power output of the wind farm.

27. A method according to claim 26 which tries to maintain the voltage at the predetermined point in the network within a predetermined limit of the desired voltage at the predetermined point in the network using the converter and if this cannot be achieved the method then alters the tap at which the tap changing transformer is set.

. . .

28. A method according to claim 26 or 27 which allows a plurality of generators to be controlled generally by a single tap changing transformer.

29. A method according to any of claims 26 to 28 which monitors the outputs of each of the generators being controlled.

30. A method according to any of claims 26 to 29 which further comprises providing a supervising controller to monitor the controllers of each of the generators.

31. A method according to any of claims 26 to 30 which implements a control strategy in which the tap of the tap changing controller is changed only when the voltage at the predetermined point in the network can not be maintained within a predetermined limit of the desired voltage at the predetermined point by altering any one of the converters.

32. A method according to any of claims 26 to 31 which alters the tap on the tap changing transformer only once substantially all of the generators have saturated.

33. A method according to any of claims 26 to 32 which models the tap changer transformer as a finite state machine.

34. A method according to any of claims 26 to 33 in which the controller changes the tap of the tap changing transformer when the rotor current in one or more of the turbines is at a predetermined limit.

35. A method according to claim 34 in which the controller causes the tap changing transformer to perform at least one of the following: change :e. *: ::: . . . . .

down a tap when a rotor current in one or more of the turbines is negative; and up a tap when a rotor current in one or more of the turbines is positive.

36. A method according to claim 34 in which the controller causes the tap changing transformer to perform at least one of the following: change down a tap when the voltage at the predetermined point in the network is low; and change up a tap when the voltage at the predetermined point in the network is high.

37. A method of controlling the output of a generator comprising measuring a parameter indicative of the output, comparing the measured parameter with at least one predetermined value and if the measured parameter falls outside a predetermined limit of the predetermined value controlling a tap changing transformer to try and bring the output back within the predetermined limit of the predetermined value.

38. A controller for use in a combination of a voltage controller and a tap changer substantially as described herein and as illustrated in the accompanying Figures.

39. A system comprising a wind-turbine, converter and tap changing converter substantially as described herein and as illustrated in the accompanying Figures.

40. A method of controlling the output of a generator as described herein and as illustrated in the accompanying Figures.

41. A windfarm substantially as described herein and as illustrated in the accompanying Figures.