Transient Analysis of Variable-Speed Wind Turbines

during a Converter Control Malfunction

R. Melício, V. M. F. Mendes, J. P. S. Catalão

Abstract--This paper is on variable-speed wind turbines with Accurate modeling and control of wind turbines have high
permanent magnet synchronous generator. A three-mass drive priority in the research activities all over the world. At the
train model and two different topologies for the power-electronic moment, substantial documentation exists on modeling and
converters are considered, respectively a two-level and a
control issues for the doubly fed induction generator (DFIG)
multilevel converter. A control strategy, based on fractional-
order controllers, is proposed for the wind turbines. Simulation wind turbine. But this is not the case for wind turbines with
results are presented to illustrate the behavior of the wind PMSG and full-power converter [4].
turbines during a converter control malfunction, considering the Previous papers were mainly focused on the transient
fractional-order controllers. Finally, conclusions are duly drawn. analysis of variable-speed wind turbines at external grid faults
[5, 6]. However, little attention has been given to the
Keywords: computer simulation, power electronics, transient possibility of internal abnormal operating conditions.
analysis, wind turbines. The influence of a pitch control malfunction, on the quality
of the energy injected into the grid, was studied in [7].
I. INTRODUCTION
This paper focuses on the transient analysis of wind turbines

A S the penetration level of wind power increases into the

power systems, the overall performance of the electric
grid will increasingly be affected by the characteristics of
with PMSG and full-power converters, considering: (i) a
three-mass drive train model; (ii) two different topologies for
power-electronic converters, respectively two-level and
wind turbines. One of the major concerns related to the high multilevel converters; (iii) a fractional-order control strategy;
penetration level of the wind turbines is the impact on power (iv) a converter control malfunction; (v) the bending
system stability [1]. Also, network operators have to ensure flexibility of the blades. Simulation results for the converter
that consumer power quality is not deteriorated. Hence, the control malfunction ascertain the performance of wind
total harmonic distortion (THD) should be kept as low as turbines equipped with fractional-order controllers.
possible, improving the quality of the energy injected into the
electric grid [2]. Power-electronic converters have been II. MODELING
developed for integrating wind power with the electric grid.
The use of power-electronic converters allows not only for A. Wind Turbine
variable-speed operation of a wind turbine, but also for The mechanical power of the turbine is given by:
enhancement on power extraction [3].
1
In a variable-speed wind turbine with full-power converter, Ptt = ρ π R 2 u 3c p (1)
2
the wind turbine is directly connected to the generator and the
generator is completely decoupled from the electric grid. Of where ρ is the air density, R is the radius of the area covered
all the generators used in wind turbines, the permanent magnet by the blades, u is the wind speed value, and c p is the power
synchronous generator (PMSG) is the one with a significant
coefficient.
advantage: it is stable and secure under normal operating
The computation of the power coefficient requires the use
conditions; and comparing with a wound synchronous
of blade element theory and the knowledge of blade geometry.
generator, it is smaller and does not need a direct current
In this paper, the numerical approximation developed in [8] is
power source for field excitation.
followed, where the power coefficient is given by:
18.4
⎛ 151 ⎞ − λi
The work of R. Melício was supported by the Fundação para a Ciência e a c p = 0.73 ⎜⎜ − 0.58 θ − 0.002θ 2.14 − 13.2 ⎟⎟ e (2)
Tecnologia (FCT) under Post-Doctoral grant (SFRH/BPD/68585/2010). ⎝ λi ⎠
R. Melício is with Center for Innovation in Electrical and Energy Engineering
(CIEEE), IST, Lisbon, and Centre for Aerospace Science and Technologies
1
(CAST), UBI, Covilha, Portugal. λi = (3)
V. M. F. Mendes is with ISEL, Lisbon, and CAST-UBI, Covilha, Portugal. 1 0.003
−
J. P. S. Catalão is with UBI, Covilha, and CIEEE-IST, Lisbon, Portugal (λ − 0.02 θ) ( θ3 + 1)
(e-mail of corresponding author: catalao@ubi.pt).
where θ is the pitch angle of the rotor blades and λ is the tip
Paper submitted to the International Conference on Power Systems
Transients (IPST2011) in Delft, the Netherlands June 14-17, 2011
speed ratio.
 The global maximum for the power coefficient is at null B. Three-Mass Drive Train Model
pitch angle and it is equal to: With the increase in size of the wind turbines, one question
c p max (λ opt (0), 0) = 0.4412 (4) arises whether long flexible blades have an important impact
on the transient analysis of wind energy systems during a fault
corresponding to an optimal tip speed ratio at null pitch angle [11].
equal to: To determine the dynamic properties of the blade, finite
λ opt (0) = 7.057 (5) element techniques may be used but this approach cannot
easily be implemented in power systems analysis programs.
The conversion of wind energy into mechanical energy Hence, to avoid the use of the finite element approach it is
over the rotor of a wind turbine is influenced by various forces necessary to simplify the rotor dynamics as much as possible.
acting on the blades and on the tower of the wind turbine (e.g. One way to achieve this is represented in Fig. 1, where the
centrifugal, gravity and varying aerodynamic forces acting on blade analysis is represented as a simple torsional system.
blades, gyroscopic forces acting on the tower), introducing Since the blade bending occurs at a significant distance
mechanical effects influencing the energy conversion [9]. from the joint between the blade and the hub, the blade can be
Those mechanical effects have been modeled by eigenswings split in two parts, OA and AB. The blade sections OA1, OA2
mainly due to the following phenomena: asymmetry in the and OA3 establish the moment of inertia of the hub and the
turbine, vortex tower interaction, and eigenswing in the rigid section blade and have the moment of inertia J h , the rest
blades. of the blade sections A1B1, A2B2 and A3B3 are the effective
The mechanical power over the rotor of the wind turbine flexible blade section and have the moment of inertia J b [12].
has been modeled, using the mechanical eigenswings [10], as
The configuration of the three-mass drive train model is
a set of harmonic terms multiplied by the power associated
shown in Fig. 2. The equations for the three-mass model are
with the energy capture from the wind by the blades, given by:
based on the torsional version of the second law of Newton,
⎡ 3
⎛ 2 ⎞ ⎤ deriving the state equation for the rotor angular speed at the
Pt = Ptt ⎢1 + ∑ Ak ⎜⎜ ∑ a km g km (t ) ⎟⎟ hk (t )⎥ (6) wind turbine and for the rotor angular speed at the generator.
⎢⎣ k =1 ⎝ m =1 ⎠ ⎥⎦

g km = sin ⎛⎜ m ωk (t ' ) dt ' + ϕ km ⎞⎟

t

∫
⎝ 0 ⎠
(7)

where Pt is the mechanical power of the wind turbine

disturbed by the mechanical eigenswings, m is the order of
the harmonic of a eigenswing, A k is the magnitude of the
eigenswing k, g km is the distribution of the m-order harmonic
in the eigenswing k, akm is the normalized magnitude of g km ,
hk is the modulation of eigenswing k, ωk is the
eigenfrequency of the eigenswing k, and ϕ km is the phase of
the m-order harmonic in the eigenswing k.
The frequency range of the wind turbine model with
mechanical eigenswings is from 0.1 to 10 Hz. The values used
on (6) and (7) for the calculation of Pt are given in Table I Fig. 1. Blade bending dynamics for the three-mass drive train model.
[10].
TABLE I
MECHANICAL EIGENSWINGS EXCITED IN THE WIND TURBINE

k Source Ak ωk hk m a km ϕkm

1 4/5 0
1 Asymmetry 0.01 ωt 1
2 1/5 π/2
Vortex 1 1/2 0
2 tower 0.08 3 ωt 1
interaction 2 1/2 π/2

3 Blades 0.15 9π 1/2 (g11+g21) 1 1 0

Fig. 2. Blade bending dynamics for the three-mass drive train model.
 C. Wind Speed ⎧ dμ
⎪ , ℜ(μ) > 0
The wind speed usually varies considerably and it has been dt μ
⎪
modeled in this paper as a sum of harmonics with frequency μ
a Dt = ⎨ 1, ℜ(μ) = 0 (9)
range 0.1–10 Hz [10]. ⎪ t
⎪ ∫ ( dτ) , ℜ(μ) < 0
−μ
The wind speed model upstream of the rotor is modeled in
⎩ a
this paper by:
where μ is the order of derivative or integrals, ℜ(μ) is the
⎡ ⎤
u(t ) = u0 ⎢1 +
⎣
∑
k
Ak sin ( ωk t )⎥
⎦
(8) real part of the μ . The mathematical definition of fractional
derivatives and integrals has been the subject of several
where u0 is the average wind speed and u is the wind speed descriptions. The most frequently encountered one is called
value with disturbance. Riemann–Liouville definition, in which the fractional-order
integral is given by:
D. Generator
1 t
The generator considered in this paper is a PMSG. The a Dt−μ f (t ) =
Γ(μ) ∫a
(t − τ) μ −1 f ( τ) dτ (10)
equations for modeling a PMSG can be found in the literature
[13]. In order to avoid demagnetization of permanent magnet while the definition of fractional-order derivatives is:
in the PMSG, a null stator current is imposed [14].
1 dn ⎡ t f ( τ) ⎤
∫
μ
a Dt f (t ) = dτ⎥ (11)
Γ( n − μ ) dt n ⎢⎣
E. Two-level Converter a (t − τ) μ − n +1
⎦
The two-level converter is an AC-DC-AC converter, with
six unidirectional commanded IGBTs used as a rectifier, and where:
∞
with the same number of unidirectional commanded IGBTs
used as an inverter. The rectifier is connected between the
Γ( x ) ≡ ∫0
y x −1 e − y dy (12)

PMSG and a capacitor bank. The inverter is connected is the Euler’s Gamma function, a and t are the limits of the
between this capacitor bank and a second order filter, which in operation, and μ is the number identifying the fractional
turn is connected to an electric grid. A three-phase active
order. In this paper, μ is assumed as a real number that
symmetrical circuit in series models the electric grid, as in [7].
The configuration of the wind energy system with two- satisfies the restrictions 0 < μ ≤ 1 . Also, it is assumed that
level converter is shown in Fig. 3. a = 0 . The following convention is used: 0 Dt−μ ≡ Dt−μ .
F. Multilevel Converter The differential equation of the fractional-order PI μ
The multilevel converter is an AC-DC-AC converter, with controller, 0 < μ < 1 , in time domain, is given by:
twelve unidirectional commanded IGBTs used as a rectifier,
and with the same number of unidirectional commanded u(t ) = K p e(t ) + K i Dt−μ e(t ) (13)
IGBTs used as an inverter. The rectifier is connected between
where K p is a proportional constant and K i is an integration
the PMSG and a capacitor bank. The inverter is connected
between this capacitor bank and a second order filter, which in constant. Taking μ = 1 in (13), a classical PI controller is
turn is connected to an electric grid. A three-phase active obtained. Hence, using Laplace transforms the transfer
symmetrical circuit in series models the electric grid, as in [7]. function of the fractional-order PI μ controller is given by:
The configuration of the wind energy system with
multilevel converter is shown in Fig. 4. G( s) = K p + Ki s − μ (14)

III. CONTROL STRATEGY B. Converters Control

Power converters are variable structure systems, because of
A. Fractional-Order Controller the on/off switching of their IGBTs. As mentioned previously,
A control strategy based on fractional-order PI μ the controllers used in the converters are respectively
controllers is studied for the variable-speed operation of wind proportional integral and fractional-order PI μ controllers.
turbines with PMSG and full-power converters, and its design Pulse width modulation (PWM) by space vector modulation
is more complex than that of classical PI controllers [15]. (SVM) associated with sliding mode is used for controlling
Fractional-order calculus used in mathematical models of the the converters.
systems can improve the design, properties and controlling The sliding mode control strategy presents attractive
abilities in dynamical systems [16]. features such as robustness to parametric uncertainties of the
The fractional-order differentiator can be denoted by a wind turbine and the generator as well as to electric grid
general operator a Dtμ [17], given by: disturbances [18].
Fig. 3. Wind energy system with two-level converter.

Fig. 4. Wind energy system with multilevel converter.

Although wind turbines achieve an excellent technical TABLE II

availability of about 98% on average, they have to face a high WIND ENERGY SYSTEM DATA
number of malfunctions [19]. A severe malfunction in the Blades moment of inertia 400×10³ kgm²
control can occur due to voltage sags and swells, harmonics,
Hub moment of inertia 19.2×10³ kgm²
frequency variation, phase steps, DC components and noise
[20]. Generator moment of inertia 16×10³ kgm²

Stiffness 1.8×106 Nm
IV. SIMULATION Turbine rotor diameter 49 m
The mathematical models for the wind energy system with Tip speed 17.64-81.04 m/s
the two-level and multilevel converters were implemented in
Rotor speed 6.9-30.6 rpm
Matlab/Simulink. The wind energy system considered has a
turbine rated power of 900 kW. The time horizon considered Turbine rated power 900 kW

in the simulation is 10 s. For the fractional-order PI μ Electric generator efficiency 90 %

controllers, μ = 0.7 is assumed in this paper. Table II Number of pairs of poles 10
summarizes the wind energy system data. Stator and rotor resistances 0.042 pu
A converter control malfunction is assumed to occur
Stator reactances Xd 1.05 pu
between 7.00 and 7.02 s, imposing a momentary malfunction
on the vector selection for the inverter of the two-level and the Stator reactances Xq 0.75 pu
multilevel converters, simulated by a random selection of Electric grid resistance Rn 0.014 pu
vectors constrained to no short circuits on the converters.
Electric grid reactance Xn 0.175 pu
The voltage vdc at the capacitor bank for the two-level
Inertia constant H 2.2 s
converter with three-mass drive train model is shown in Fig. 5.
 3000 3000
↓v dc
2500 2500
↑ ↑ Converter control malfunction
Converter control malfunction
2000 2000
Voltage (V)

Voltage (V)
↓v C1

1500 1500

1000 1000
↑v C2

500 500

0 0
6.9 7.4 7.9 8.4 8.9 6.9 7.4 7.9 8.4 8.9
Time (s) Time (s)
Fig. 5. Voltage vdc for the two-level converter with three-mass drive train Fig. 7. Voltage vdc for the multilevel converter with three-mass drive train
model. model.

As expected during the malfunction the voltage vdc suffers The currents injected into the electric grid for the wind
a small increase, the capacitor bank is charged, but almost energy system with multilevel converter and with the three-
after the end of the malfunction it recovers to its normal value. mass drive train model are shown in Fig. 8.
The currents injected into the electric grid for the wind
energy system with two-level converter and with the three-
1500
mass drive train model are shown in Fig. 6.
1000
1500
500
Current (A)

1000
0
500
Current (A)

−500
0
−1000
−500
−1500
6.95 6.975 7 7.025 7.05 7.075
−1000 Time (s)
Fig. 8. Currents injected into the electric grid (multilevel converter and three-
−1500 mass drive train model).
6.95 6.975 7 7.025 7.05 7.075
Time (s)
Fig. 6. Currents injected into the electric grid (two-level converter and three- As expected during the malfunction they decrease, but
mass drive train model). almost after the end of the malfunction they recover to their
normal behavior.
As expected during the malfunction the currents injected It should be noted that, for the same fault conditions, the
into the electric grid decrease, but almost after the end of the transient response of the three-mass drive train model is larger
malfunction they recover to their normal behavior. than that of the one-mass or the two-mass model. Thus, the
The behavior of the voltage vdc at the capacitor banks for three-mass model provides a more realistic approach. The
the multilevel converter with three-mass drive train model is consideration of the bending flexibility of blades can influence
identical to the behavior with the two-level converter, as the wind turbine response during internal faults.
shown in Fig. 7.
Also, Fig. 7 shows the inherent neutral-point voltage V. CONCLUSIONS
balancing problem with the multilevel converters due to the This paper reports a study for PMSG variable-speed wind
capacitor bank voltage divider: one capacitor bank being turbines, considering a three-mass drive train models and
overcharged, and the other being discharged; additional two different topologies for the power-electronic converters.
information can be found in [21, 22]. Our study deals with the transient analysis during an internal
fault, namely a converter control malfunction. The [9] Z. X. Xing, Q. L. Zheng, X. J. Yao, and Y. J. Jing, "Integration of large
doubly-fed wind power generator system into grid," in: Proc. 8th Int.
contributions of this paper are threefold: ascertaining the
Conf. Electrical Machines and Systems, pp. 1000-1004, Sep. 2005.
transient behavior at an internal fault using two-level and [10] V. Akhmatov, H. Knudsen, and A. H. Nielsen, "Advanced simulation of
multilevel converters; using a control strategy based on windmills in the electric power supply," Int. Journal of Electr. Power
fractional-order controllers; and investigating the effects of Energy Syst., vol. 22, pp. 421-434, Aug. 2000.
[11] H. Li and Z. Chen, "Transient stability analysis of wind turbines with
the bending flexibility of blades. The three-mass drive train
induction generators considering blades and shaft flexibility," in Proc.
model, including both blades and shaft flexibilities, may be 33rd IEEE IECON, pp. 1604-1609, 2007.
more appropriate for the transient analysis of wind energy [12] G. Ramtharan and N. Jenkins, "Influence of rotor structural dynamics
systems. representations on the electrical transient performance of DFIG wind
turbines," Wind Energy, vol. 10, pp. 293-301, Jul.-Aug. 2007.
[13] C.-M. Ong, Dynamic Simulation of Electric Machinery: Using
VI. REFERENCES Matlab/Simulink. NJ: Prentice-Hall, 1998, pp. 259-350.
[1] N. R. Ullah and T. Thiringer, "Variable speed wind turbines for power [14] T. Senjyu, S. Tamaki, N. Urasaki, and K. Uezato, "Wind velocity
system stability enhancement," IEEE Trans. Energy Conversion, vol. 22, and position sensorless operation for PMSG wind generator," in:
pp. 52-60, Mar. 2007. Proc. 5th PEDS, pp. 787-792, Nov. 2003.
[2] J. M. Carrasco, L. G. Franquelo, J. T. Bialasiewicz, E. Galvan, R. C. P. [15] B. Arijit, D. Swagatam, A. Ajith, and D. Sambarta, "Design of fractional-
Guisado, A. M. Prats, J. I. Leon, and N. Moreno-Alfonso, "Power- order PI-lambda-D-mu-controllers with an improved differential
electronic systems for the grid integration of renewable energy sources: evolution," Eng. Appl. Artif. Intell., vol. 22, pp. 343-350, Mar. 2009.
A survey," IEEE Trans. Industrial Electronics, vol. 53, pp. 1002-1016, [16] C. Jun-Yi and C. Bing-Gang, "Design of fractional order controllers
Aug. 2006. based on particle swarm optimization," in: Proc. IEEE ICIEA 2006, pp.
[3] J. A. Baroudi, V. Dinavahi, and A. M. Knight, "A review of power 1-6, 2006.
converter topologies for wind generators," Renewable Energy, vol. 32, [17] A. J. Calderón, B. M. Vinagre, and V. Feliu, "Fractional order control
pp. 2369-2385, Nov. 2007. strategies for power electronic buck converters," Signal Processing,
[4] A. D. Hansen and G. Michalke, "Modelling and control of variable- vol. 86, pp. 2803-2819, Oct. 2006.
speed multi-pole permanent magnet synchronous generator wind [18] B. Beltran, T. Ahmed-Ali, and M. E. H. Benbouzid, "Sliding mode
turbine," Wind Energy, vol. 11, pp. 537-554, Sep.-Oct. 2008. power control of variable-speed wind energy conversion systems," IEEE
[5] C. Jauch, "Transient and dynamic control of a variable speed wind Trans. Energy Conversion, vol. 23, pp. 551-558, Jun. 2008.
turbine with synchronous generator," Wind Energy, vol. 10, pp. 247-269, [19] B. Hahn, M. Durstewitz, and K. Rohrig, Reliability of Wind Turbines,
May-Jun. 2007. Renewable Energies Knowledge, 2006. Available: www.renknow.net.
[6] A. H. Kasem, E. F. El-Saadany, H. H. El-Tamaly, and M. A. A. Wahab, [20] A. P. Martins, "A DFT-based phasor estimation method for power
"An improved fault ride-through strategy for doubly fed induction electronics converter control and protection under strong grid voltage
generator-based wind turbines," IET Renew. Power Gener., vol. 2, pp. perturbations," Eur. Trans. Electr. Power, vol. 19, pp. 1082-1097, Nov.
201-214, Dec. 2008. 2009.
[7] R. Melício, V. M. F. Mendes, and J. P. S. Catalão, "Computer simulation [21] A. Shukla, A. Ghosh, and A. Joshi, "Control schemes for DC capacitor
of wind power systems: power electronics and transient stability voltages equalization in diode-clamped multilevel inverter-based
analysis," in: Proc. IPST 2009, Kyoto, Japan, Jun. 3-6, 2009. DSTATCOM," IEEE Trans. Power Deliv., vol. 23, pp. 1139-1149,
[8] J. G. Slootweg, S. W. H. de Haan, H. Polinder, and W. L. Kling, Apr. 2008.
"General model for representing variable speed wind turbines in power [22] J.-S. Lai and Z. F. Peng, "Multilevel converters – A new breed of power
system dynamics simulations," IEEE Trans. Power Syst., vol. 28, pp. converters," IEEE Trans. Ind. Appl., vol. 32, pp. 509-517, May-Jun.
144-151, Feb. 2003. 1996.