alat tambahan pada wind turbin

8/10/2019 alat tambahan pada Wind Turbin

1/10

General study of the control principles and dynamic fault behaviour of

variable-speed wind turbine and wind farm generic models

Tareq Saber Abuaisha*

Institute for Electrical Power Supply with Integration of Renewable Energies, Technische Universitt Darmstadt, Landgraf-Georg-Strae 4, 64283 Darmstadt,

Germany

a r t i c l e i n f o

Article history:

Received 9 July 2013

Accepted 9 January 2014

Available online 28 February 2014

Keywords:

Generic model

Doubly-fed induction generator

Fully-rated converter

Wind turbine

Wind farm

Synchronous induction machine

a b s t r a c t

The interest towards generic models or sometimes also called standard models of wind turbine gener-

ators (WTGs) is signicantly increasing. Mainly due to their improved power quality, better controlla-

bility and higher power extraction capability, variable-speed wind turbines driving a synchronous or an

induction machine are capturing the global market. Throughout this paper, dynamic modelling and

performance analysis of the generic models of the variable-speed W TGs, namely the doubly-fed in-

duction generator and the fully-rated converter based WTGs, are achieved using integration between

Matlab/Simulink and PSCAD/EMTDC simulation platforms. Later on, the performance of type-4 wind

turbine driving a permanent magnet synchronous machine is analysed during fault and then compared

with the case when driving a wound rotor induction machine. The differences in control principles and

dynamic fault behaviour are highlighted. Afterwards, investigations on wind farm level are accom-

plished. A case study during which the developed generic models and the generic model of the variable-

speed machine are compared is conducted. Different arrangements for the construction of the generic

wind farm are considered.

2014 Elsevier Ltd. All rights reserved.

1. Introduction

With a signicant 20% renewable energy power share (including

wind energy, hydro-power, photo-voltaics and others), today Ger-

manyis already well on the way tomeeting the 2020 targetof a 35%

renewable energy share. Sharing by 40% of all renewable energy

sources, wind is already the dominant driver of renewable electric

power generation and will be the motor for future growth [1].

In fact the study of a large system such as the European network

involves the simulation of many interconnected systems, which are

operated by many different transmission and distribution system

operators, and comprise wind turbines and wind farms from

multiple manufacturers and technologies.This variety of wind turbine manufacturers existing in the global

market would inevitably result in different manufacturer-specic

wind turbine generator (WTG) models based on different ap-

proaches and technologies. These differences would naturally

result in different complexity and thus accuracy of the respective

models. Additionally, access to these models requires a non-

disclosure agreement between the dynamic model user and the

turbine manufacturer[2].

In response to these challenges, western electricity coordinating

council (WE-CC) has initiated the development of generic positive

sequence WTG dynamic models which are suitable for grid plan-

ning studies[3]. A generic model refers to a non-proprietary dy-

namic model that can be used to represent wind turbine generators

(WTGs) with similar physical and control topology, using only

different parametrization to represent a specic vendor equipment

[3,4].

Compared to xed-speed WTGs, doubly-fed induction gener-

ator (DFIG) and fully rated converter (FRC) wind turbine generators

are ultimately dominating the global market due to their higherenergy efciency, exible performance, higher power extraction

capability and better controllability[5,6].

In order todene the intended use and limitations and to insure

a healthy development of the generic models; further research,

investigation and improvement are highly recommended.

In this paper, a general study on wind turbine and wind farm

generic models is achieved. Atrst, the three-phase generic models

of type-3 (DFIG) and type-4 (FRC) are implemented and simulated.

In order to analyse the dynamic behaviour of the generic models, a

balanced three-phase fault at the wind turbine generator terminal

is enforced. The simulation results are then compared with

* Tel.: 49 (0)176 32262014.

E-mail addresses: [email protected],eng_abumosab@hotmail.

com.

Contents lists available atScienceDirect

Renewable Energy

j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m/ l o c a t e / r e n e n e

http://dx.doi.org/10.1016/j.renene.2014.01.004

0960-1481/

2014 Elsevier Ltd. All rights reserved.

Renewable Energy 68 (2014) 245e254
mailto:[email protected]:[email protected]:[email protected]://www.sciencedirect.com/science/journal/09601481http://www.elsevier.com/locate/renenehttp://dx.doi.org/10.1016/j.renene.2014.01.004http://dx.doi.org/10.1016/j.renene.2014.01.004http://dx.doi.org/10.1016/j.renene.2014.01.004http://dx.doi.org/10.1016/j.renene.2014.01.004http://dx.doi.org/10.1016/j.renene.2014.01.004http://dx.doi.org/10.1016/j.renene.2014.01.004http://www.elsevier.com/locate/renenehttp://www.sciencedirect.com/science/journal/09601481http://crossmark.crossref.org/dialog/?doi=10.1016/j.renene.2014.01.004&domain=pdfmailto:[email protected]:[email protected]:[email protected]


2/10

manufacturer-specic models in order to analyse the physical

behaviour and characteristics. Later on, the implemented models

were added to the Matlab/Simulink library to serve as basis for

future research and development.

Afterwards, the performance of type-4 wind turbine when it is

driving a permanent magnet synchronous machine (PMSM) is

investigated during fault and compared with the case when it is

driving a wound rotor induction machine (WRIM). The differences

in control principles and dynamic fault behaviour are highlighted.

To show resemblance to other research efforts which are con-

ducted in this eld; a comparison between two different wind

farms is conducted in a case study. While the rst wind farm

consists of four wind turbines which are based on the implemented

generic models, the second wind farm is based on the generic

model of the variable-speed machine which is implemented in

Refs.[7,8]. Different arrangements for the construction of the rst

wind farm from generic wind turbine generators are considered.

2. Power control strategies in variable-speed WTGs

There are primarily two ways to control the power generated by

variable-speed WTGs. The rst is the aerodynamic power control

either by controlling the wind turbine blades or by yawing the

complete nacelle, and the second is the electrical power control

employed in the power electronics converters[9].

The use of these two control strategies enables on one hand the

extraction of maximum wind power during low wind speeds and

on the other hand, the reduction of the mechanical stress on the

shafts during high wind speeds[10,11].

2.1. Aerodynamic power control

The power extracted fromthe wind varies with the cube of wind

speed, but the wind turbine can only extract a fraction of this

amount which is given in Ref.[12]by

P 1

2rAv3wCpl;bW; (1)

Wind turbines cannot extract all the kinetic energy from the

wind, and the maximum extracted energy occurs with Cp 0.59,

and is termed as Betz limit [12]. In practice, this powercoefcient is

less than this value and also varies with the tip speed ratio l as

shown inFig. 1.

The idea of the aerodynamic power control is to achieve

maximum power tracking. In order to accomplish this task, the

turbine must operate at the peak ofCp/lcurve for all relevant wind

speed values. This curve is unique to a particular design of wind

turbine and thus it is given by the wind turbine manufacturer[13].

Fig. 1shows a typical Cp/l curve. Since Pmax occurs at Cp,max, the

wind turbine should follow the dashed curve to choose themaximum power coefcient valueCp,maxfor all relevant operating

wind turbine speeds.

Currently, pitch control is the most widely used control option

especially for multi-megawatt WTGs. Almost all variable-speed

wind turbine generation technologies use pitch control. When

wind speed is below the rated speed, it is used to maximize the

energy capture. And when it is above the rated speed, it is used to

reduce the mechanical stress on the gearbox and the shafts[10,11].

2.2. Electrical power control

Fig. 2 shows the per-unit mechanical output power versus speed

characteristics also in per-unit of a WTG for different wind speeds

and zero pitch angle. The dashed-dotted black curve is termed as

optimal tracking curve; which is the total amount of power exists in

the wind. Whilst, the dashed green (in web version) curve is known

as real tracking curve which is the real power extracted from the

wind by the wind turbine.

Fig. 2 compares between the real and the optimal power

tracking, where the green (in web version) points on the curve

represent the following[14]:

e u1 is the minimum wind speed at which thewind turbine comes

into operation (typically between 3 and 4 m/s)

e uris the speed at which the wind turbine will produce its rated

output power

e u2is the maximum operating wind speed

As stated in equation(1), at wind speeds above the cut-in speed

and till the rated wind speed is reached, the wind turbine output

Fig. 1. Power coefcientCpas a function of tip-speed ratio l for different blade pitchanglesb.

Fig. 2. Turbine speed [p.u] versus the turbine output power [p.u] at different wind

speeds and zero pitch angle b.

T.S. Abuaisha / Renewable Energy 68 (2014) 245e254246


3/10

power increases due to a cubic relationship with wind speed.

Moreover, the green (in web version) curve inFig. 2clearly shows

the link between variable-speed operation (and thus speed control)

and the amount of power that can be extracted from the wind.

3. Modelling and simulation of variable-speed WTGs

Fig. 3shows the dynamic model structure of type-3 and type-4WTGs including the individual subsystems together with the

commands in between, as rstly described by WECC[2,4].

The generic dynamic models are represented by a specic

number of interactions among different subsystems to achieve the

functionality of a typical WTG. This functionality should include the

independent control of active (torque) and reactive power and the

control of generator speed and blade pitch angles, in order to

achieve maximum wind power extraction[4].

These generic models are suitable for representing individual

WTGs or the equivalent of a wind power plant where the wind

speed is assumed to remain constant during the simulation time

frame[15].

3.1. Proposed generator/converter subsystem

To show an example on the implemented modelling of variable-

speed generic models, the proposed generator/converter subsys-

tem is comprehensively described in Fig. 4. The physical concepts

behind the proposed subsystem are presented in Section 4,along

with the modelling and simulation of wind farms.

Ref.[16] proved that the generator dynamics and power elec-

tronics can be emulated by a regulated current source. In this paper,

the proposed subsystem is represented by a regulated current

source which injects proportional three-phase currents (Iaref, Ibrefand Icref) into the power system based on the control commands

(Ipcmd& Eqcmd) from the converter control subsystem.

Thereafter, it gives both the measured active and reactive power

(Pgen & Qgen) to the converter control subsystem as feedback and

also the measured active power (Pgen) to the wind turbine sub-

system.

As shown in Fig. 4, the desired instantaneous active and reactive

power (Pgen & Qgen) are calculated using equations(2) and (3) as

stated in Refs. [12,16], and then fed into the corresponding sub-

systems.Fig. 5shows the instantaneous active and reactive power

measured at the WTG terminal when the input voltage falls to 50%

from its rated value. It is also seen that due to the voltage dip, the

active power has also dropped to a comparable value. Additionally,

the reactive powerhas increased to support the voltage as required.

Pt 3

2,

Vdt Idt Vqt Iqt

MW (2)

Qt 3

2,

VqtIdtVdtIqt

Mvar (3)

The functions of the subsystems of type-4 generic model are

almost identical to those of type-3[7]. However, the wind turbine

subsystem of type-4 WTG employs a simplied mechanical model.This is mainly due tothe fact, that in type-4 generic model the fully-

rated converter completely decouples both the generator transients

and grid faults from affecting each other and thus smoother

connection to the grid is enabled[7,8].

The proposed generator/converter subsystem along with all the

other subsystems are implemented and simulated in Matlab/

Simulink based on the parameters provided by WECC and general

electric (GE) for type-3 and type-4 generic models [4,15,17]. In the

following subsection, the control strategies followed in the voltage

source converters of the generic models are explained. The FRC-

based wind turbine is chosen as a case study.

3.2. Control strategies of voltage source converters in type-4 genericmodel

The typical structure of a fully rated converter WTG as imple-

mented in PSCAD/EMTDC is depicted in Fig. 6. It consists of a

generator, two pulse width modulated (PWM) voltage source

converters (VSC) with a back-to-back DC link, wind turbine and a

control system. The inductor L is used to couple the grid-side VSC

with the grid.

The generatorcan have a woundeld, wound rotor synchronous

generator (WRSG) or can use permanent magnets, permanent

magnet synchronous generator (PMSG) to provide the rotating

magneticeld. Or it can even be an induction generator. A purpose

of thisresearch work is to compare between the performance of the

synchronous and induction machines in terms of control principles

and dynamic fault analysis.

When a WRSG is employed, the required DC excitation can be

taken from the DC link. However, one attraction of the PMSM is its

high efciency since no magnetizing or eld current is necessary to

provide the magnetic eld. The control system shown in Fig. 6is

used to insure the desired functionality of the FRC-based wind

turbine as follows:

Pitch angle (b) command is used to insure maximum wind

power extraction by the turbine blades. Since the wind speed is

Fig. 3. Dynamic model structure for variable-speed wind turbine generator technologies [4].

T.S. Abuaisha / Renewable Energy 68 (2014) 245e254 247


4/10

Fig. 4. Detailed description of the proposed generator/converter subsystem for variable-speed generic models.

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8-1

-0.5

0

0.5

1

1.5

time [sec]

Active&R

eactivePower[pu]

PgenQgen

(a) Instantaneous active and reactive power measured

at type-3 WTG terminal

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8-1

-0.5

0

0.5

1

1.5

time [sec]Active&R

eactivepower[pu]

PgenQgen

(b) Instantaneous active and reactive power measured

at type-4 WTG terminal

Fig. 5. Instantaneous active and reactive power (Pgen& Qgen) measured at WTG terminal during a 50% voltage dip.



5/10

assumed to be constant during the simulation time frame, this

command was not implemented.

Voltage signal (Vrc) command is intended to control the rotor-

side VSC. Voltage signal (Vgc) command is intended to control the grid-

side VSC.

The role of these two signals in controlling the VSCs depends on

the type of the employed machine i.e. synchronous or induction

machine as will be shown later.

The variable magnitude, variable frequency output voltage of

the generator (Vr) is converted to a xed magnitude, xed fre-

quency voltage (Vs) by the fully-rated converter consisting of a

rotor-side VSC, a grid-side VSC and a DC link capacitor. Both con-

verters are designed to handle the full rated power of the wind

turbine. The proposed control strategy is achieved through con-

trolling the modulation indices (Mg and Mr) of the PWM using

equation(4).

Mg kg,

Vgabc

Vdc

!;

Mr kr,Vrabc

Vdc

!;

(4)

whereMg and Mrare the modulation indices of the grid and ma-

chine side converters respectively;kgandkr are the proportionality

constants.

The above explained strategy is shown in Fig. 7. In order to

highlight the differences in control principles and dynamic fault

behaviour, a balanced three-phase to ground fault was caused atthe low voltage side of the transformer. Afterwards, the perfor-

mance of the FRC wind turbine driving a PMSM is analysed during

fault and then compared with the case when driving a WRIM.

Fig. 8 shows the simulation results of the FRC wind turbine

when it is driving a PMSM compared with the case when it is

driving a WRIM. Note that the fault duration is relatively high and it

is not in accordance with most grid codes. This is because the goalof this research is to highlight different machine control principles;

it is not in any case to test the low voltage ride through (LVRT)

capability of the wind turbine generator.

Due to the three-phase balanced fault, the voltage falls down

and the current rises up for both machines. This can be seen in

Fig. 8a and b respectively.

As a result to increased rotor currents, the DC-link voltage tends

to increase (see magnied parts ofFig. 8a). Thus, the grid side VSC

will try to balance the DC-link voltage by forcing it to decrease

again, hence the reactive power will increase negative increase or

capacitive effect. This control action can be clearly realized in

Fig. 8c. However due to the additional reactive power requirement

of the induction machine, the reactive power and in accordance the

active power curves tend to oscillate. This effect is shown in theright hand side ofFig. 8c.

Moreover, the machine side converter in the PMSM will control

the active and reactive power by forcing them to stabilize at their

reference values. This control action is obvious in left hand side of

Fig. 8d. In comparison, the WRIM absorbs reactive power

depending on its characteristics and operating point. Thus when a

reactive power control that does not match the machine require-

ment is forced, the machine can enter an unstable situation.

Due to this fact, for the case of WRIM topology the machine side

VSCcannot perform control on reactive power, as a result the active

and reactive power at the machine terminal will not stabilize at

their reference values during the fault as seen in the right hand side

ofFig. 8, which is not the case for the PMSM shown in the left hand

side.The simulation results proved that when a WRIM is employed,

in this case the machine side VSC cannot perform control on

reactive power otherwise the machine can go unstable. Whereas

Fig. 6. Detailed structure of an FRC-based wind turbine as implemented in PSCAD/EMTDC.

Fig. 7. Calculation of the modulation indices in FRC-WTG generic model as implemented in PSCAD/EMTDC.



6/10

reactive power control, and thus an indirect voltage control, on

machine side VSC is possible for the FRC wind turbine driving a

PMSM. Note that the control principles applied on the case of FRC

wind turbine driving a WRIM, will be also followed for the case of

DFIG driving an induction machine.

4. Modelling and simulation of variable-speed wind farms

Typical wind farms may consist of tens to hundreds of identical

wind turbines. As a consequence, representing a wind farm withindividual wind turbines for power system stability studies in-

creases the complexity of the model and requires time consuming

simulations. With this background, aggregated representation of

wind farms is essential.

4.1. Case study: aggregated model of a wind farm

Ref. [7] proposed a generic model of the variable-speed ma-

chine. This aggregated model was built based on a third order

quasi-sinusoidal model. The quasi steady-state (QSS) model can be

derived from the full order model by assuming that the transformer

voltage in the stator winding can be neglected against the much

greater speed voltage, as shown in equation(5).

jusJs[

dJs

dt

(5)Based on this assumption and solving for the original equations

of the full order model, equations (6) and (7)will be obtained for

the quasi stationary model.

Stator voltage equation:

(6)

Rotor ux equation:

dJrdt

RrLr

jusur

Jr kRRris vr (7)

where:

L0 LsL2m

Lr

kr Lm

Lr

Fig. 8. Control principles of an FRC wind turbine driving a PMSM (left hand side) compared with the case when driving a WRIM (right hand side) during a balanced three-phase

fault and constant wind speed.



7/10

As shown in the phase estimation channel ofFig. 4, in the pro-

posed generator/converter model the stator ux angle is calculated

using equation(8).

8>>>>>>>>>>>>>:

Jsa

Z vsaRsisadt

; qf tan1Jsb

Jsa

Jsb

ZvsbRsisb

dt

(8)

The above two equations along with the equation of motion

constitute the third order model. Equation(7) can be rewritten in

dq reference frame to generate the direct and quadrature compo-

nents in equations(9) and (10) respectively.

dJrddt

RrLr

us urJrqkRRrisd vrd (9)

dJrqdt

RrLr

usurJrdkRRrisq vrq (10)

Equations(9) and (10)are used in the machine side converter tocontrol the voltage and torque (active power) of the machine. The

DC-link voltage is assumed to remain constant around its rated

value, and thus the grid side converter is not considered explicitly

in the quasi stationary model.

Simplifying the quasi stationary model will generate the

aggregated model of the variable-speed machine shown in Fig. 9.

This is a representation of the core control functions, but for a

complete implementation the required key elements have to be

added. The complete generic model of a variable-speed machine

along with all the employed key elements is shown inFig. 10.

The model as a whole (both with and without the inclusion of

the detailed model) constitutes the simplied model. The only

difference between the two options (including or excluding the

detailed model) is the level of simplication which one may chooseto adopt for a particular simulation. The generic model of the var-

iable speed machine shown inFig. 10operates as follows:

The reference values Uref, Pref& Qref are usually being deter-

mined from load-ow initialization. Though, they can be set to

empirical values based on experience.

The voltage channel will increase or decrease the reactive power

in-feed whenever the voltage exits a dead-band of 10% above or

below the rated value.

The reactive power channel will determine the desired current

command iQ,ref [p.u], while the active power channel will

determine the desired current command iP,ref[p.u].

Thereafter, reactive power priority is applied. Mainly, because

the generic model is designed to analyse fault situations.

After that, the user e or more appropriately the system operator

e chooses whether to include or exclude the generic model.

Based on that choice, the desired active and reactive power Q(t)

&P(t)will be calculated.

The generic model of the variable-speed machine is imple-

mentedusing Matlab/Simulink. A graphical user interface (GUI) has

been designed to control the whole operation dynamically during

simulation. The GUI will assign the values ofUref, Pref& Qref based

on the operators choice. It also gives the operator the ability to

include or exclude the detailed model and then determines the

initial conditions required and feeds them into the corresponding

transfer functions and integrators.

4.2. Variable-speed generic wind farm model

The proposed wind farm consists of four WTGs. Each of these

WTGs has a capacity of 1.5 MW and thus the whole wind farm

capacity is 6 MW. This wind farm will be built based on the generic

models of type-3 and type-4 WTGs developed in Section3.

All the different possible arrangements are studied, in the rstarrangement all the four WTGs are FRC-based. In the second

arrangement they are assumed to be DFIG-based. Whereas in the

last arrangement, two WTGs are DFIG-based and the other two are

FRC-based.

The generic models of wind turbines will be employed to

represent the four individual wind turbines within the wind farm.

Then the resulted wind farm will be simulated and the results will

be compared with that of the aggregated wind farm developed in

Section4.1.

The implemented wind farm in Matlab/Simulink is depicted in

Fig. 11while the comparison results are shown inFig.12. The input

voltage is subjected to 50% voltage dip due to a balanced three-

phase fault.

As shown in Fig. 12, the active power dip is deeper in theaggregated model, and the reactive powersupport is higheras well.

One conrmed reason for this fact is that the two models apply

different approaches on the provision of reactive power support

during the voltage dip.

In the aggregated wind farm model, the proportional gain factor

between voltage and reactive power (K) is considered (refer to

Fig. 10). While, in the detailed model the proportional gain is not

considered and thus reactive power support will not be propor-

tional to the voltage dip[19].

The results of the comparison representa comparablebehaviour

between the aggregated and the generic wind farm models. But

specically, the smooth behaviour of the aggregated wind farm

model is better matched with the constellation of four FRC-based

machines than other constellations, this is because the FRC de-couples the generator transients to yield a smoother response.

5. Conclusions

Doubly fed induction generator (DFIG) and fully rated converter

(FRC) wind turbine generators offer many advantages over thexed

speed technology including exible performance and better

controllability. In this paper, the three-phase dynamic generic

models of type-3 (DFIG) and type-4 (FRC) have been implemented

and simulated.

The performance of the FRC wind turbine driving a permanent

magnet synchronous machine (PMSM) is analysed during fault and

then compared with the case when it is driving a wound rotor in-

duction machine (WRIM). Different control strategies in the voltage

Fig. 9. Core of the aggregated generator model of machine and current control

(reproduced from Ref. [5]).



8/10

source converters (VSCs) for the two mentioned topologies should

be applied.

The simulation results proved that when a WRIM is employed,

in this case the machine side VSC cannot perform control on

reactive power otherwise the machine can go unstable. Whereas

reactive power control, and thus an indirect voltage control, on

machine side VSC is possible for the FRC wind turbine driving a

PMSM.

Afterwards, a generic wind farm model consists of four wind

turbine generators (WTGs) is implemented and simulated. In order

to show resemblance to other research efforts which are conducted

in this eld, the simulation results are then compared with that of

the aggregated model of the variable speed machine implemented

in Ref.[7].

The detailed model represents a better match to the real phys-

ical behaviour of the machine, however, the increased complexity

Fig. 10. Quasi steady-state generic model of the variable speed machine (reproduced from Ref. [18]).

Fig. 11. Generic Matlab/Simulink model of a wind farm consisting of four WTGs (1.5 MW each).



9/10

will inevitably demand greater modelling efforts and larger simu-

lation time. Apart from that, the simulation results showed that a

comparable behaviour between the aggregated and the generic

wind farm model can be obtained.

Acknowledgements

The authors would like to thank Manitoba HVDC ResearchCenter for the technical conversations and the exchange of ideas.

They are also grateful to the anonymous reviewers for their valu-

able comments and suggestions to improve the quality of the paper.

Nomenclature

r density of air in kg/m3

A area swept by the rotor blades in m2

vw wind speed in m/s

b blade pitch angle in degrees

l ratio of the rotor tip speedurto the wind speed vwCp coefcient of performance of the turbine

v,i,J complex voltage, current and ux vectors

Rs,Rr stator, rotor winding resistancesLs,Lr,Lls,Llr stator, rotor winding self and leakage inductances

Lm magnetizing inductance

us,ur synchronous, rotor angular frequencies

P,Q active and reactive power

Vgabc three-phase complex voltages at the grid side converter

Vrabc three-phase complex voltages at the rotor side converter

qf stator-ux vector position

Superscripts/subscripts

s, r stator, rotor

d,q direct, quadrature axis component

a,b alpha, beta axis component

g, c grid, converter side value

n, ref nominal, reference value

Appendix A. Simulation parameters

Fig. 12. Comparing the active (left hand side) and reactive (right hand side) power of an aggregated wind farm model with a generic model consisting of four WTGs during a

balanced three-phase fault and constant wind speed.

Table A.1

Parameters of variable-speed WTGs (seeFigs. 3 and 4).

Parameter DFIG FRC

Nominal RMS voltage 690 V 690 V

Nominal power 1.5 MW 1.5 MW

Nominal frequency 50 Hz 50 HzGain ofux calculation channel 1 p.u 1 p.u

Lag time ofux calculation channel (Tk) 0.03 s 0.005 s

Lag time of injected current channel (Tg) 0.02 s 0.03 s

Equivalent reactance (X0

)a 1 p.u 1 p.u

a The rest of the parameters are as reported by GE and WECC [15,17].

Table A.2

Parameters used in PSCAD to simulate the FRC wind turbine (seeFig. 6).

Parameter PMSM WRIM

Machine parameters

Nominal RMS voltage 600 V 600 V

Nominal power 1.5 MW 1.5 MW

Nominal frequency 50 Hz 50 Hz

Stator resistance 0.017 p.u 0.02 p.u

Leakage reactance 0.064 p.u 0.1 p.u

Control parameters of grid side VSCPro po rti on al ga in of AC vo lta ge regulat or 0 .1 p. u 0 .1 p.u

Integral gain of AC voltage regulator 0.02 p.u 0.02 p.u

Pro po rti on al ga in of D C vol ta ge r egula to r 0 .8 p. u 0 .8 p.u

Integral gain of DC voltage regulator 0.1 p.u 0.1 p.u

Control parameters of machine side VSCePMSM

Reactive power regulator

Max quadrature current (Iq,max) 5 p.u

Min quadrature current (Iq,min) 5 p.u

Proportional gain (KpQ) 0.1 p.u

Integral gain (KiQ) 0.02 p.u

Active power regulator

Max direct current (Id,max) 5 p.u

Min direct current (Id,min) 5 p.u

Proportional gain (KpP) 0.1 p.u

Integral gain (KiP) 0.02 p.u



10/10

References

[1] EEG. German renewable energy sources act [Online],http://www.eeg-aktuell.de/; 2012.

[2] Corporation, N.A.E.R.. Standard models for variable generation. Tech. Rep. 116-390 Village Blvd., Princeton, NJ 08540-5721, USA: NERC; 2010

[3] Asmine M, Brochu J, Fortmann J, Gagnon R, Kazachkov Y, Langlois CE, et al.Model validation for wind turbine generator models. IEEE Trans Power Syst2011;26(3):887e91.

[4] Ellis A, Kazachkov Y, Muljadi E, Pourbeik P, Sanchez-Gasca J. Description andtechnical specications for generic WTG models e a status report. In: IEEE/PESPower Systems Conference and Exposition (PSCE); 2011. pp. 524e9. WECCWorking Group on Dynamic Performance of Wind Power Generation & IEEEWorking Group on Dynamic Performance of Wind Power Generation .

[5] Hu J, Yuan X. VSC-based direct torque and reactive power control of doublyfed induction generator. Renew Energy 2012;40(1):13e23.

[6] Caliao ND. Dynamic modelling and control of fully rated converter windturbines. Renew Energy 2011;36(8):2287e97.

[7] Fortmann J, Engelhardt S, Kretschmann J, Felters C, Janssen M, Neumann T,et al. Generic simulation model for DFIG and full size converter based windturbines. In: 9th International Workshop on Large-Scale Integration of WindPower, Quebec/Canada; 2010.

[8] Fortmann J, Engelhardt S, Kretschmann J, Felters C, Erlich I. New genericmodel of DFG based wind turbines for RMS type simulation. IEEE Trans Energy

Convers; 2013 ISSN: 0885-8969:99.[9] Blaabjerg F, Chen Z, Teodorescu R, Iov F. Power electronics in wind turbine

systems. In: 5th International Power Electronics and Motion Control Confer-ence (IPEMC), vol. 1; 2006. pp. 1e11.

[10] Boukhezzar B, Lupu L, Siguerdidjane H, Hand M. Multivariable control strategyfor variable speed, variable pitch wind turbines. Renew Energy 2007;32(8):1273e87.

[11] Lee J, Son E, Hwang B, Lee S. Blade pitch angle control for aerodynamic per-formance optimization of a wind farm. Renew Energy 2013;54:124e30[AFORE].

[12] Machowski J, Bialek JW, Bumby JR. Power system dynamics: stability & con-trol; chap. 7. 2nd ed. UK, Poland: John Wiley & Sons Ltd; 2008. pp. 265e97.

[13] Price WW, Sanchez-Gasca JJ. Simplied wind turbine generator aerodynamicmodels for transient stability studies. In: IEEE PES Power Systems Conferenceand Exposition (PSCE 2006); 2006. pp. 986e92.

[14] Kusiak A, Verma A, Wei X. Wind turbine capacity frontier from SCADA. WindSyst Mag 2012;3(9):36e9.

[15] Behnke M, Ellis A, Kazachkov Y, McCoy T, Muljadi E, Price W, et al. Devel-opment and validation of WECC variable speed wind turbine dynamic modelsfor grid integration studies. In: AWEA Wind Power Conference. Los Angeles,California: National Renewable Energy Laboratory e NREL; 2007. pp. 1e5.

[16] Faria KJ. Doubly-fed induction generator based wind power plant models.Masters thesis. Austin, Texas, USA: The University of Texas at Austin; 2009.

[17] Clark K, Miller NW, Sanchez-Gasca JJ. Modelling of GE wind turbine-generators for grid studies. Tech. Rep. Schenectady, NY, 12345 USA: GeneralElectric International, Inc.; 2010

[18] van Custem T. Algorithmic and modelling improvements for simpliedsimulation, validation on industrial cases. Tech. Rep. 2. PEGASE; 2011. pp. 41e8.

[19] Abuaisha TS. Methods for the provision of dynamic models for distributedenergy resources. Masters thesis. Aachen, Germany: RWTH-Aachen Univer-sity; 2012.

Table A.3

Parameter list of the aggregated wind farm model in Matlab/Simulink (seeFig.10).

Name Value Unit Description

Active and reactive power control

Vn 20 kV Nominal voltage

Sn 2 MVA Nominal apparent powe r

Uref e kV Assigned automatically from GUI

Pref e MW Assigned automatically from GUI

Qref e

Mvar Assigned automatically from GUIDBmax 1.1 p.u Upper deadband limit

DBmin 0.9 p.u Upper deadband limit

Ku 2 p.u Gain in voltage channel

Imax 1 p.u Converter current limit

Tv 60 s First time constant in reactive power channel

Tq 20 s Second time constant in reactive power channel

Detailed model parameters (orange box)

Iimmax 5 p.u Upper limit applied to imaginary current

Iimmin 5 p.u Lower limit applied to imaginary current

Iremax 5 p.u U pper li mi t a pplied to r ea l c urrent

Iremin 5 p.u L ower l imi t a pplied t o real curr ent

Ti1 0.005 p.u Time delay applied in upper branch

Ti2 0.03 p.u Time delay applied in lower branch

X 0.5 p.u Equivalent instantaneous reactance

http://www.eeg-aktuell.de/http://www.eeg-aktuell.de/http://refhub.elsevier.com/S0960-1481(14)00033-0/sref2http://refhub.elsevier.com/S0960-1481(14)00033-0/sref2http://refhub.elsevier.com/S0960-1481(14)00033-0/sref3http://refhub.elsevier.com/S0960-1481(14)00033-0/sref3http://refhub.elsevier.com/S0960-1481(14)00033-0/sref3http://refhub.elsevier.com/S0960-1481(14)00033-0/sref3http://refhub.elsevier.com/S0960-1481(14)00033-0/sref4http://refhub.elsevier.com/S0960-1481(14)00033-0/sref4http://refhub.elsevier.com/S0960-1481(14)00033-0/sref4http://refhub.elsevier.com/S0960-1481(14)00033-0/sref4http://refhub.elsevier.com/S0960-1481(14)00033-0/sref4http://refhub.elsevier.com/S0960-1481(14)00033-0/sref4http://refhub.elsevier.com/S0960-1481(14)00033-0/sref4http://refhub.elsevier.com/S0960-1481(14)00033-0/sref4http://refhub.elsevier.com/S0960-1481(14)00033-0/sref4http://refhub.elsevier.com/S0960-1481(14)00033-0/sref5http://refhub.elsevier.com/S0960-1481(14)00033-0/sref5http://refhub.elsevier.com/S0960-1481(14)00033-0/sref5http://refhub.elsevier.com/S0960-1481(14)00033-0/sref5http://refhub.elsevier.com/S0960-1481(14)00033-0/sref6http://refhub.elsevier.com/S0960-1481(14)00033-0/sref6http://refhub.elsevier.com/S0960-1481(14)00033-0/sref6http://refhub.elsevier.com/S0960-1481(14)00033-0/sref7http://refhub.elsevier.com/S0960-1481(14)00033-0/sref7http://refhub.elsevier.com/S0960-1481(14)00033-0/sref7http://refhub.elsevier.com/S0960-1481(14)00033-0/sref7http://refhub.elsevier.com/S0960-1481(14)00033-0/sref7http://refhub.elsevier.com/S0960-1481(14)00033-0/sref8http://refhub.elsevier.com/S0960-1481(14)00033-0/sref8http://refhub.elsevier.com/S0960-1481(14)00033-0/sref8http://refhub.elsevier.com/S0960-1481(14)00033-0/sref8http://refhub.elsevier.com/S0960-1481(14)00033-0/sref9http://refhub.elsevier.com/S0960-1481(14)00033-0/sref9http://refhub.elsevier.com/S0960-1481(14)00033-0/sref9http://refhub.elsevier.com/S0960-1481(14)00033-0/sref9http://refhub.elsevier.com/S0960-1481(14)00033-0/sref9http://refhub.elsevier.com/S0960-1481(14)00033-0/sref10http://refhub.elsevier.com/S0960-1481(14)00033-0/sref10http://refhub.elsevier.com/S0960-1481(14)00033-0/sref10http://refhub.elsevier.com/S0960-1481(14)00033-0/sref10http://refhub.elsevier.com/S0960-1481(14)00033-0/sref11http://refhub.elsevier.com/S0960-1481(14)00033-0/sref11http://refhub.elsevier.com/S0960-1481(14)00033-0/sref11http://refhub.elsevier.com/S0960-1481(14)00033-0/sref11http://refhub.elsevier.com/S0960-1481(14)00033-0/sref12http://refhub.elsevier.com/S0960-1481(14)00033-0/sref12http://refhub.elsevier.com/S0960-1481(14)00033-0/sref12http://refhub.elsevier.com/S0960-1481(14)00033-0/sref13http://refhub.elsevier.com/S0960-1481(14)00033-0/sref13http://refhub.elsevier.com/S0960-1481(14)00033-0/sref13http://refhub.elsevier.com/S0960-1481(14)00033-0/sref13http://refhub.elsevier.com/S0960-1481(14)00033-0/sref13http://refhub.elsevier.com/S0960-1481(14)00033-0/sref13http://refhub.elsevier.com/S0960-1481(14)00033-0/sref14http://refhub.elsevier.com/S0960-1481(14)00033-0/sref14http://refhub.elsevier.com/S0960-1481(14)00033-0/sref14http://refhub.elsevier.com/S0960-1481(14)00033-0/sref15http://refhub.elsevier.com/S0960-1481(14)00033-0/sref15http://refhub.elsevier.com/S0960-1481(14)00033-0/sref15http://refhub.elsevier.com/S0960-1481(14)00033-0/sref15http://refhub.elsevier.com/S0960-1481(14)00033-0/sref15http://refhub.elsevier.com/S0960-1481(14)00033-0/sref15http://refhub.elsevier.com/S0960-1481(14)00033-0/sref16http://refhub.elsevier.com/S0960-1481(14)00033-0/sref16http://refhub.elsevier.com/S0960-1481(14)00033-0/sref16http://refhub.elsevier.com/S0960-1481(14)00033-0/sref16http://refhub.elsevier.com/S0960-1481(14)00033-0/sref17http://refhub.elsevier.com/S0960-1481(14)00033-0/sref17http://refhub.elsevier.com/S0960-1481(14)00033-0/sref17http://refhub.elsevier.com/S0960-1481(14)00033-0/sref18http://refhub.elsevier.com/S0960-1481(14)00033-0/sref18http://refhub.elsevier.com/S0960-1481(14)00033-0/sref18http://refhub.elsevier.com/S0960-1481(14)00033-0/sref18http://refhub.elsevier.com/S0960-1481(14)00033-0/sref18http://refhub.elsevier.com/S0960-1481(14)00033-0/sref19http://refhub.elsevier.com/S0960-1481(14)00033-0/sref19http://refhub.elsevier.com/S0960-1481(14)00033-0/sref19http://refhub.elsevier.com/S0960-1481(14)00033-0/sref19http://refhub.elsevier.com/S0960-1481(14)00033-0/sref19http://refhub.elsevier.com/S0960-1481(14)00033-0/sref19http://refhub.elsevier.com/S0960-1481(14)00033-0/sref19http://refhub.elsevier.com/S0960-1481(14)00033-0/sref19http://refhub.elsevier.com/S0960-1481(14)00033-0/sref18http://refhub.elsevier.com/S0960-1481(14)00033-0/sref18http://refhub.elsevier.com/S0960-1481(14)00033-0/sref18http://refhub.elsevier.com/S0960-1481(14)00033-0/sref17http://refhub.elsevier.com/S0960-1481(14)00033-0/sref17http://refhub.elsevier.com/S0960-1481(14)00033-0/sref17http://refhub.elsevier.com/S0960-1481(14)00033-0/sref16http://refhub.elsevier.com/S0960-1481(14)00033-0/sref16http://refhub.elsevier.com/S0960-1481(14)00033-0/sref15http://refhub.elsevier.com/S0960-1481(14)00033-0/sref15http://refhub.elsevier.com/S0960-1481(14)00033-0/sref15http://refhub.elsevier.com/S0960-1481(14)00033-0/sref15http://refhub.elsevier.com/S0960-1481(14)00033-0/sref15http://refhub.elsevier.com/S0960-1481(14)00033-0/sref15http://refhub.elsevier.com/S0960-1481(14)00033-0/sref14http://refhub.elsevier.com/S0960-1481(14)00033-0/sref14http://refhub.elsevier.com/S0960-1481(14)00033-0/sref14http://refhub.elsevier.com/S0960-1481(14)00033-0/sref13http://refhub.elsevier.com/S0960-1481(14)00033-0/sref13http://refhub.elsevier.com/S0960-1481(14)00033-0/sref13http://refhub.elsevier.com/S0960-1481(14)00033-0/sref13http://refhub.elsevier.com/S0960-1481(14)00033-0/sref12http://refhub.elsevier.com/S0960-1481(14)00033-0/sref12http://refhub.elsevier.com/S0960-1481(14)00033-0/sref12http://refhub.elsevier.com/S0960-1481(14)00033-0/sref11http://refhub.elsevier.com/S0960-1481(14)00033-0/sref11http://refhub.elsevier.com/S0960-1481(14)00033-0/sref11http://refhub.elsevier.com/S0960-1481(14)00033-0/sref11http://refhub.elsevier.com/S0960-1481(14)00033-0/sref10http://refhub.elsevier.com/S0960-1481(14)00033-0/sref10http://refhub.elsevier.com/S0960-1481(14)00033-0/sref10http://refhub.elsevier.com/S0960-1481(14)00033-0/sref10http://refhub.elsevier.com/S0960-1481(14)00033-0/sref9http://refhub.elsevier.com/S0960-1481(14)00033-0/sref9http://refhub.elsevier.com/S0960-1481(14)00033-0/sref9http://refhub.elsevier.com/S0960-1481(14)00033-0/sref9http://refhub.elsevier.com/S0960-1481(14)00033-0/sref8http://refhub.elsevier.com/S0960-1481(14)00033-0/sref8http://refhub.elsevier.com/S0960-1481(14)00033-0/sref8http://refhub.elsevier.com/S0960-1481(14)00033-0/sref7http://refhub.elsevier.com/S0960-1481(14)00033-0/sref7http://refhub.elsevier.com/S0960-1481(14)00033-0/sref7http://refhub.elsevier.com/S0960-1481(14)00033-0/sref7http://refhub.elsevier.com/S0960-1481(14)00033-0/sref6http://refhub.elsevier.com/S0960-1481(14)00033-0/sref6http://refhub.elsevier.com/S0960-1481(14)00033-0/sref6http://refhub.elsevier.com/S0960-1481(14)00033-0/sref5http://refhub.elsevier.com/S0960-1481(14)00033-0/sref5http://refhub.elsevier.com/S0960-1481(14)00033-0/sref5http://refhub.elsevier.com/S0960-1481(14)00033-0/sref4http://refhub.elsevier.com/S0960-1481(14)00033-0/sref4http://refhub.elsevier.com/S0960-1481(14)00033-0/sref4http://refhub.elsevier.com/S0960-1481(14)00033-0/sref4http://refhub.elsevier.com/S0960-1481(14)00033-0/sref4http://refhub.elsevier.com/S0960-1481(14)00033-0/sref4http://refhub.elsevier.com/S0960-1481(14)00033-0/sref4http://refhub.elsevier.com/S0960-1481(14)00033-0/sref3http://refhub.elsevier.com/S0960-1481(14)00033-0/sref3http://refhub.elsevier.com/S0960-1481(14)00033-0/sref3http://refhub.elsevier.com/S0960-1481(14)00033-0/sref3http://refhub.elsevier.com/S0960-1481(14)00033-0/sref2http://refhub.elsevier.com/S0960-1481(14)00033-0/sref2http://www.eeg-aktuell.de/http://www.eeg-aktuell.de/

alat tambahan pada wind turbin

Documents