www.ietdl.

org
Published in IET Renewable Power Generation
Received on 9th December 2009
Revised on 7th May 2010
doi: 10.1049/iet-rpg.2009.0207

ISSN 1752-1416

One-power-point operation for variable speed wind/

tidal stream turbines with synchronous generators
H.-B. Zhang1 J. Fletcher1 N. Greeves2 S.J. Finney1 B.W. Williams1
1
Department of Electronics and Electrical Engineering, University of Strathclyde, Glasgow G1 1XW, UK
2
Senior electronics manager, Proven Energy Ltd., Scotland KA3 5LH, UK
E-mail: [email protected]

Abstract: This study presents a new operating scheme for variable speed wind/tidal stream turbines employing synchronous
generators. Current maximum power tracking schemes for variable speed wind energy conversion systems rely on periodic
comparison of the output power for guiding the direction of maximum power tracking or an online algorithm which
continuously provides a generator torque/speed reference corresponding to the maximum power status at different wind speed
conditions. The proposed scheme utilises only one-power-point on the maximum power curve. Once the information for the
maximum power status of a local wind speed is known, then the wind energy conversion system implements maximum
power tracking and constant power control for different wind speed conditions using simple control. The operation strategy is
applicable to tidal stream turbines. Simulation results demonstrate the concept.

D duty cycle
Nomenclature
DN duty cycle for rated power control
vi/ii converter input voltage/current J turbine inertia
Pm captured wind power Ra generator resistance
r air density Ld/Lq generator stator inductance
A turbine swept area
Cp power coefficient
vw wind speed 1 Introduction
Cpmax maximum power coefficient
Wind and tidal stream energy conversion possesses immense
k constant, k ¼ 1/2(rACpmax) potential for generating electrical power and wind industry
l tip speed ratio experience provides an invaluable reference for growing
b blade pitch angle tidal stream turbine exploitation [1, 2]. The electrical
v turbine rotational speed control topologies and energy converting strategies of wind
turbines can be applied to tidal stream turbines due to the
r blade radius similar control principles. Both involve fluid flow but tidal
Pe boost converter input electrical power turbines are located in the marine environment while wind
c magnetic flux turbines face air flow [3 – 7].
Various types of maximum power operating schemes exist
ig generator current
for variable speed wind energy conversion systems (WECS)
h generator/rectifier set efficiency [8 – 17]. Generally, WECS with synchronous generators
vN/vC boost converter input voltage at rated/cut-in [18 – 25] implement maximum power tracking with an
wind speed instantaneous torque or speed control reference for different
PN wind energy conversion system rated power wind speed conditions, with the assistance of a generator
rotating speed sensor or estimator. The control algorithms
vbase/ibase the one-power-point voltage/current may be complex and extra hardware cost is added due to
L boost converter inductor inductance the speed sensor/estimator.
vo dc link voltage Perturbation and observation (P&O) techniques use a simple
T switching period control concept and can remove the necessity for speed sensors
or estimators. Such techniques regulate the direction for
ton/toff boost converter switch on/off time maximum power tracking by periodic comparison of output
i∗ reference current power. Usually multiple steps are needed before the systems

IET Renew. Power Gener., 2011, Vol. 5, Iss. 1, pp. 99 –108 99

doi: 10.1049/iet-rpg.2009.0207 & The Institution of Engineering and Technology 2010
www.ietdl.org

Fig. 1 Wind energy conversion system with a synchronous generator

reaches steady state and attention is necessary to ensure the Cp is a non-linear function of tip speed ratio (TSR), l, and the
maximum power tracking ability under transient wind turbine blade pitch angle b. Here b ¼ 08 is set for maximum
conditions [26, 27]. power operation. TSR is given by
An online training method for maximum power tracking
based on the P&O technique has been proposed [28]. After the vr
l= (2)
training process, a lookup table containing reference currents vw
of different wind speeds is populated and then directly applied
to the power tracking control. However, training the WECS For each wind speed, there is a certain turbine rotating speed
over the full wind speed working range is time-consuming, which extracts maximum wind power. For rotating speeds
and the presented dynamic maximum power tracking response above or below this rotating speed, the turbine output
performance for transient wind changes was not satisfactory. power is reduced. Fig. 2 shows a typical Cp-TSR curve
Different from the control schemes in the literature, this which can be obtained from field-tested data [29, 30], for
paper presents a so-called ‘one-power-point’ (OPP) control which the optimised TSR TSRop is 6 and the corresponding
scheme for wind/tidal energy conversion systems with maximum Cp is Cpmax ¼ 0.398.
synchronous generators. Owing to control similarities, the
analysis and simulations of the proposed approach to 2.2 Extracted electrical power
WECS utilising the topology, shown in Fig. 1, are readily
applied to tidal stream energy conversion systems. The extracted electrical power at the boost converter side is
In order to implement maximum power control for the
WECS at variable wind speeds, the OPP approach requires pe = vi × ii (3)
the WECS to be operated at one wind speed to achieve the
corresponding maximum power status, and it has the For synchronous generators, with the smoothing effect of
following features: capacitor C1 , the rectifier output voltage can be expressed
as [28]
† Only the maximum power status of the WECS for one
local wind speed needs to be obtained. The measuring
vi = w(c, ig ) × v (4)
methods can be any of the conventional methods and wind
speed training is not necessary.
† Once the maximum status for the WECS for one wind For synchronous generators, the magnetic flux c is fixed by
speed, that is, the output voltage vi and the current ii at the field current which is held at rated value while for
rectifier side, is known, the WECS can be initialised with permanent magnet generators the flux is constant. Therefore
these two parameters and then both maximum power w is only related to the generator current ig . If the voltage
tracking and rated power control can be implemented. drop caused by ig due to the generator impedance can be
† The one particular maximum power point is measured only ignored, then w can be considered as a constant, whence vi
once and then same parameters are applied to the same type of is proportional to v [18, 21].
WECS to implement simple control without knowledge of the
speed, which is convenient for massive application.
† WECS using the initialised parameters are free of
maintenance and relocation. The WECS can be shut down
and restarted at any wind speed within the working range,
with the same control capability.

Simulation results are presented to demonstrate the control

concept.

2 Control principle
2.1 Wind turbine aerodynamic characteristics

The mechanical power captured from the wind is expressed as

1
Pm = rACp v3w (1)
2 Fig. 2 Power coefficient against TSR

100 IET Renew. Power Gener., 2011, Vol. 5, Iss. 1, pp. 99 –108
& The Institution of Engineering and Technology 2010 doi: 10.1049/iet-rpg.2009.0207
 www.ietdl.org
2.3 OPP control principle With (v1 , i1) known in conjunction with (10), the voltage vN
for switching the WECS into constant power mode is
As shown in Fig. 2, the maximum wind power is extracted if
Cp can be ideally maintained at its maximum value Cpmax for 
vN = (PN × v21 )/i1
3
different wind speeds, then (1) can be written as (11)

Pm = kv3w (5) The input voltage vi can be online compared with vN to

determine the operation mode:
where k is a constant and k ¼ 1/2(rACpmax).
Equation (5) is a monotone increasing function, where Pm † if vi ≤ vN , the WECS operates in the maximum power
is proportional to the cube of the wind speed vw . Such a tracking mode;
function is shown in Fig. 3. † if vi . vN , the WECS operates in a constant power mode.
Suppose the power conversion efficiency of the turbine/
generator set is h and when the WECS attains steady state Thus, with the OPP, once a maximum power point is
at each wind speed obtained, the other power points to regulate a constant
power output can be directly calculated. Hereafter the first
Pe = Pm × h (6) power point is termed as Pbase (vbase , ibase).

That is 3 Control implementation

OPP operation of WECS needs one initial maximum power
vi × ii = hkv3w (7) point condition for a local wind speed to be obtained. Then
the derived parameters are initialised into the WECS system
For a constant TSR, with vi proportional to v in order to extract the maximum wind power and rated
power control at various wind speeds.
vw / v / vi (8)
3.1 Obtaining the OPP
Equations (7) and (8) give
The accuracy of Pbase (vbase , ibase) is critical for OPP operation
v2w / v2i / ii (9) and can be obtained by any conventional maximum power-
point tracking (MPPT) algorithm for initialising purpose.
That is, when a power point (v1 , i1) for a wind speed vw1 on For this particular power point, any wind speed within the
the maximum power curve is obtained, as shown in Fig. 3, working range can be used and the wind speed need not be
another power point (v2 , i2) for a wind speed vw2 on the known. For practical operation, the wind speed in the
same curve must fulfil the following relations middle of the working range is recommended, in order to
reduce the non-linearity relation effect in (4). Once the
 2 maximum power status for a local wind speed is obtained,
v the WECS can be initialised for routine control.
i2 = 2 × i1 (10)
v1 In terms of re-calibration of Pbase due to a significant
WECS fault or possible system parameter time drift,
In other words, once OPP on the maximum power curve is various sensorless on-field calibration strategies can be
known, and if the current ii can be controlled utilising (10), used. Several re-calibration control strategies are briefly
the WECS can always be operated on the curve in Fig. 3 introduced.
that extracts maximum power.
When the wind speed increases and the output power 3.1.1 Stable wind environments
reaches the rated power PN of WECS, the system switches † Mature P&O techniques without employing the speed
into a constant power control mode for system safety, until sensor/estimator can be used for calibration.
the cut-out speed is reached and other precautions are † Fix vbase . Perturb the WECS with different current values
applied to avoid damage. starting from ibase , and then P&O techniques can be used to
find a new optimised ibase . Vice versa if ibase is fixed.

3.1.2 Severe unstable wind environments: Run the

WECS with the old Pbase for some time, obtain the average
boost converter input voltage and set it as the new vbase .
Then train the WECS with the control strategy as proposed
in [28], until a proper input current value is found, which is
set as the new ibase . This method takes a longer time.

3.2 Boost converter control

For a given current reference, different current regulators,

such as hysteresis and PI controllers, can be employed for
control. In this paper, the control approach is by direct
calculation, without the need to tune a controller. Owing to
Fig. 3 Captured wind power curve when Cp ¼ Cpmax the wind characteristics, the boost converter is always

IET Renew. Power Gener., 2011, Vol. 5, Iss. 1, pp. 99 –108 101
doi: 10.1049/iet-rpg.2009.0207 & The Institution of Engineering and Technology 2010
www.ietdl.org
readily observed; then scenario 4.2 verifies OPP operation
with realistic wind profiles. To clarify the illustration,
although no particular wind speed knowledge is needed for
OPP operation, specific wind speeds are used in the
simulation. Tables 1 and 2 show the wind turbine and
control system simulation specifications.

4.1 OPP operation with linear wind speeds

Scenario 4.1 is divided into three. The first part is

initialisation and maximum power tracking performance,
the second is the restart process of the initialised WECS
Fig. 4 One switching period current through the boost converter with constant power control and the third scenario is OPP
inductor operation reliability evaluation.

operating in a continuous non-steady state. The inductor 4.1.1 Initialisation and maximum power tracking:
current within one switching period (T ) in such a state is The wind speed span is 6 – 11 m/s. WECS’s initialisation
shown in Fig. 4. and maximum power tracking performance are shown in
Suppose the rectifier output voltage is vi at the beginning of Fig. 5a. To obtain the maximum power operating status of
the switching period, and ton and toff are the IGBT g on and the WECS at a particular wind speed, a conventional MPPT
off times in this period, then method is used. In Fig. 5a, period 0 2 t1 is a state with
proper duty control at a stable wind speed of vw ¼ 8.5 m/s.
Di1 Consequently, the WECS is operating at maximum power
vi = L (12) output. During this period the average voltage vi and
ton
current ii for Pbase are vbase ¼ 446.82 V, ibase ¼ 8 A, with
Di2 TSR ¼ 6.04 and Cp ¼ 0.39785.
vo − vi = L (13) At time t1 , Pbase is initialised by the system and the OPP
toff
algorithm takes the control of the WECS. As shown in
where L is the inductor inductance and vo is the dc-link Fig. 5a (a), during the period 1 –3 s, the wind speed
voltage, which is maintained constant by the grid converter. increases from 8.5 to 11 m/s, then during the period 5 – 7 s,
Equations (12) and (13) gives the wind speed decreases from 11 to 6 m/s. At time 9 and
10 s, two step changes occur, one is from 6 to 8 m/s, the
Di other is from 8 to 11 m/s.
vi = vo × (1 − D) + L (14) Fig. 5a parts (b) – (e) shows the captured mechanical power
T Pm , wind turbine power coefficient Cp , boost converter input
voltage vi , and current ii , respectively. With OPP operation,
where D ¼ ton/T is the duty cycle for IGBT switch g.
Cp can be maintained at its maximum value during various
The reference current at vi is
stable wind conditions. Worse cases occur during the
 2 periods 1 – 3 and 5 – 7 s, when the wind speed varies from
∗ vi one stable state to another. At 7 s, the lowest Cp ¼ 0.392 is
i = × ibase (15)
vbase

So Table 1 Wind turbine specifications

 2 Items Specifications
vi
Di = × ibase − ii (16) air density 1.220 kg/m3
vbase
turbine blade radius 2.75 m
Substituting (16) into (14), the duty cycle, which can be swept area 23.8 m2
instantaneously calculated to control the WECS tracking the turbine inertia J 6 kg m2
maximum power at variable wind speeds, is maximum power at 11 m/s 7.68 kW
generator direct drive three-phase PMSG
stator resistance Ra 1.4 V
D = 1 − [vi − ((vi /vbase )2 × ibase − ii ) × L/T ]/vo (17) stator inductance (Ld , Lq) 5.8 mH, 5.8 mH
flux induced by magnets 2.6 Wb
When a WECS reaches its rated power PN , the control duty to pole pairs 6
maintain a constant power output is

DN = 1 − [vi − (PN /vi − ii ) × L/T ]/vo (18)

Table 2 Control parameters

Items Specifications
4 Simulation results switching frequency 10 kHz
capacitance (C1 , C2) 2000 mF
The simulations are implemented in Matlab. Two scenarios inductance (L) 12 mH
are considered. Scenario 4.1 is OPP operation with linear dc link voltage (vo) 690 V
wind speeds, for which the operation characteristics are

102 IET Renew. Power Gener., 2011, Vol. 5, Iss. 1, pp. 99 –108
& The Institution of Engineering and Technology 2010 doi: 10.1049/iet-rpg.2009.0207
 www.ietdl.org

Fig. 5 OPP linear wind speed operating characteristics

a Initialisation and maximum power tracking
b Restart and constant power control

reached. When the wind speed stabilises at 6 m/s, maximum t ¼ 4 s, the wind speed begins to drop, and stabilises at
Cp is recovered in less than 0.3 s. At time 9 and 10 s, the t ¼ 5 s at a speed of 8.5 m/s.
configured WECS reacts to the two step changes in less Before t1 , the dc-link is charged to and maintained at
than 0.7 and 0.4 s, respectively, with Cp converging to its 690 V, with turbine rotating at v ¼ 21.7 rad/s and
maximum value. Cp ¼ 0.1. At time t1 , the OPP algorithm takes control. The
In the maximum power tracking mode with slow variation WECS starts by outputting a short period of Pe ¼ PN by
wind, Cp performance during wind changes generally reflects converting the system kinetic energy and releasing the
the damping effect of the turbine shaft and capacitor C1 . energy in C1 . The maximum wind power is captured in 0.5 s.
During transient wind speed increase/decrease, the energy Cp performance in Fig. 5b (c) is a result of the torque
stored in the turbine shaft system and capacitor C1 causes difference between the driven wind and electrical torque
the turbine rotating speed to increase/decrease slightly controlled by OPP MPPT and rated power control,
slower than the ideal rotating speed, so the actual TSR respectively. After 1 s, the turbine rotating speed increases
deviates from TSRop and a diminished Cp performance is as the acceleration increases. The initial acceleration is
observed. At such times, Cp does not represent the extracted small and TSR decreases from TSRop . Then the
electrical power delivered to the grid converter, and the acceleration reaches such a value that Cp remains at a
performance can be optimised by the control system design. minimum for a short period. After that, the significant wind
mechanical torque increment due to the wind speed increase
4.1.2 Restart and constant power control: The causes a dramatic acceleration increment, thus TSR/Cp
maximum power at a wind speed of 9 m/s is set as the rated recovers towards TSRop/Cpmax . After rated power PN is
power, PN ¼ 4.2 kW, and vN ¼ 471.5 V is the voltage reached, the turbine rotating speed continues to increase due
reference for the WECS switched into rated power control. to the reducing current reference given by constant power
Fig. 5b parts (a) – (e) shows the simulated wind speed, control, and TSR increases from less than TSRop to greater
turbine rotating speed, power coefficient Cp , captured wind than TSRop . During this process, a portion of the maximum
power Pm and electrical power Pe and boost converter input extractable gust energy is converted into kinetic energy in
current ii , respectively. turbine rotating system and electrical energy in C1 . In the
The system is initialised with parameters obtained at rated power control mode, the least boost converter current
vw ¼ 8.5 m/s. The period 0 – t1 represents a stable state reference is given when turbine speed reaches its maximum.
that the WECS is restarted at the wind speed 6 m/s with When turbine rotating speed reaches a maximum, the gust
IGBT g in the off-state. At time 1.5 s, the wind speed speed decrease deviates TSR further from TSRop and then
increases from 6 m/s and in 2 s reaches 10 m/s, then at the least Cp ¼ 0.31 occurs. Subsequently, the wind

IET Renew. Power Gener., 2011, Vol. 5, Iss. 1, pp. 99 –108 103
doi: 10.1049/iet-rpg.2009.0207 & The Institution of Engineering and Technology 2010
www.ietdl.org
mechanical torque (a result from both wind speed decrease During gust speed increases, the speed reaches 9 m/s at
and diminished Cp performance) and increasing electrical t ¼ 3 s. The WECS is switched into constant power control
torque (due to the decreasing turbine rotating speed) causes at t ¼ 3.22 s due to the dampening effect of turbine rotating
Cp approaches Cpmax in 1 s, in the given wind profile. system and capacitor C1 , as shown in Fig. 5b (d).

Fig. 6 OPP reliability analysis

a Linearity relationship between vi and v
b With stator resistance 1.4 and 2.8 V
c With turbine inertia 6 and 12 kg m2

104 IET Renew. Power Gener., 2011, Vol. 5, Iss. 1, pp. 99 –108
& The Institution of Engineering and Technology 2010 doi: 10.1049/iet-rpg.2009.0207
 www.ietdl.org
4.1.3 OPP operation reliability evaluation: OPP performance are shown in Fig. 6c (stator resistance ¼ 1.4 V
operation reliability is analysed for the configured WECS in both cases). The solid line in Fig. 6c (b) and (c) shows
and non-linear vi against v relations. WECS operation with an inertia of 12 kg m2, whereas the
OPP reliability on the configured WECS: The linearity dashed line represents the case of an inertia of 6 kg m2. A
between vi and v relation and different system parameter higher turbine inertia damps the system dynamic response,
variation effects are evaluated and shown in Figs. 6a– c, which reduces WECS dynamic wind power capturing
respectively. capability. When a WECS with larger inertia reaches a
The linearity between vi and v relation on the configured stable state at a higher wind speed, more kinetic energy is
WECS: OPP operation is based on (4). The relation that the stored within the turbine shaft system. In Fig. 6c (d), during
boost converter input voltage vi is proportional to v must 5 – 7 s, the output electrical power of the larger inertia
be maintained by system design for accurate OPP operation. turbine is more than that from the smaller inertia system.
In [31], such a relation was presented for a 0.8 kW WECS This is because the kinetic energy stored in a higher wind
based on a study of generator/rectifier set where good speed status exceeds the captured wind energy difference by
linearity is observed when the generator electrical angular the difference in Cp . Fig. 6c shows that dynamic response
frequency is below 120 rad/s. The WECS dynamic response capability with OPP operation remains high in both inertia
due to the applied control algorithm also needs to be cases.
considered because it affects the charging/discharging of OPP operation under non-linear v against vi relations: Fig. 7
capacitor C1 , thus affects vi . So in this paper, the linearity shows the OPP Pbase selection strategy. The solid line is the
relation between vi and v is evaluated with OPP operation actual generator/rectifier output voltage against generator
on the WECS in a dynamic maximum power tracking rotating speed curve [31]. KC , KN and KOPP represent the
process, as shown in Fig. 6a. lines intersecting this curve at voltage vC , vN and vbase ,
Fig. 6a (a) is the designed wind profile which simulates respectively, where vC is the boost converter input voltage
different variation possibilities, namely, linear change, step at the cut-in wind speed. If Pbase is chosen in the region
change and non-linear speed change. Fig. 6a (b) shows the between vC and vN , least control error occurs at the cut-in
turbine rotating speed v and boost converter input voltage wind speed and the wind speed at which the WECS enters
vi curves, obtained with OPP operation on the WECS in the rated power control, most wind power is extracted between
dynamic maximum power tracking mode under the wind these two extreme wind conditions.
profile defined in Fig. 6a (a). With the defined scales, two OPP operation accuracy for non-linear v and vi inter-
curves superimpose each other. Fig. 6a (c) shows the scaled relation depends on the WECS characteristics. Electrical
errors in areas A, B and C of Fig. 6a (b), where the grey and mechanical system non-linearities both influence OPP
lines are the boost converter input voltage and the black operation accuracy, but may still remain high under certain
lines are the turbine rotating speed. Fig.6a demonstrates non-linear conditions. The region with low variation slope
OPP operation accuracy on the presented WECS for the at the top of the l – Cp curve around the (TSRop , Cpmax)
given profiles. point provides flexibility for maintaining high OPP wind
Generator/rectifier set power conversion efficiency h: power extraction capability.
Equation (7) is another relationship that affects accurate
OPP operation. The power conversion efficiency at different † Although Pbase is critical for OPP operation, the candidate
generator resistance conditions (due to system parameter region around (TSRop , Cpmax) provides a relative wide TSR
time drift and temperature) and different power output range for high-energy capturing efficiency.
levels needs to be evaluated. Fig. 6b shows the operating † If the generator/rectifier set is designed to operate in the
status of the WECS with the designed generator stator region that introduces a significant non-linear vi against v
resistance 1.4 V (dashed line), compared with the operating relation, optimisation of the l – Cp curve should be
status obtained when doubling the resistance to 2.8 V (solid considered.
line). † During stable or ramping wind conditions at transient wind
Fig. 6b (c) shows that a larger stator resistance leads to a speed changes within the linear vi against v region, a high
better Cp performance when the wind speed increases while percent of maximum wind power can still be extracted,
a decreased Cp performance when the wind speed
decreases. During wind speed increase, an increased
generator resistance dampens the generator current increase,
thus a greater torque difference between the electrical
torque and that of the driven wind occurs. Therefore the
turbine rotating speed increases with a greater acceleration,
leading to a better Cp performance. When the wind speed
decreases, an increased generator resistance dampens the
energy release, thus a worse Cp performance occurs. The
increased generator resistance does not significantly affect
the WECS output electrical power, as shown in Fig. 6b (d),
where the two output electrical power curves superimpose.
Owing to the reversed Cp performance during wind speed
variation conditions, an increased generator resistance also
does not significantly affect the WECS wind power
converting efficiency. The dynamic response capability with
OPP operation remains high in both stator resistance cases.
Turbine inertia variation: Turbine/generator/gearbox
mechanical replacement may change the system inertia. The
effects of inertia variation (doubling of 6 kg m2) on OPP Fig. 7 OPP Pbase selection strategy

IET Renew. Power Gener., 2011, Vol. 5, Iss. 1, pp. 99 –108 105
doi: 10.1049/iet-rpg.2009.0207 & The Institution of Engineering and Technology 2010
www.ietdl.org
although the turbine rotating speed deviates slightly from the three consecutive gusts. The peak gust speeds are designed to
optimal speed. be within the rated WECS output power PN , that is, the
WECS always operates in the maximum power tracking mode.
The WECS rarely operates continuously at extremely high The wind profile, turbine rotating speed, power coefficient,
wind speeds, which would introduce significant non-linearity input mechanical power, output electrical power and boost
between the vi against v relation, thus, the deterioration of converter input voltage and current are shown, respectively.
wind power capturing efficiency under significant non-linear At time t ¼ 0 s, the WECS is started and converges to the
vi against v conditions does not significantly affect the maximum Cp in 0.3 s, as shown in Fig. 8c. Cp variations can
general WECS wind power conversion efficiency. be observed due to instantaneous wind speed changes, during
which the turbine shaft system and the energy stored in
4.2 OPP operation with realistic wind profiles capacitor C1 damps the WECS reaction. During the period
1 – 11 s, the least Cp is 0.355.
In this scenario, the wind profiles are created according to Fig. 8h shows another wind profile. The WECS operates at
[32] and the following regulation conditions are used: an average wind speed of 6 m/s, and then starts at t ¼ 1 s a
gust causes the WECS power rating to be exceeded. This
† The rated power PN ¼ 4.2 kW of the WECS is reached at a gust lasts 2.5 s. At t ¼ 5 s, a ramp wind speed change
wind speed of 9 m/s. commences and in 3 s reaches 9 m/s and then stabilises.
† The WECS is initialised with a Pbase obtained at a wind Figs. 8i– 8n show the corresponding WECS performances
speed of 8.5 m/s. under this wind profile.
As shown in Figs. 8j– 8l, during the constant power control
The WECS performances are shown in Fig. 8. period, the difference between the input mechanical power
Figs. 8a–g illustrates the performances of the initialised and the output electrical power is stored in the rotating shaft
WECS under unstable wind conditions. The WECS is started system and capacitor C1 . When the gust speed drops, this
at an average wind speed of 6 m/s, which is succeeded by energy is released into the WECS, as shown in Figs. 8k and

Fig. 8 OPP operation with realistic wind profiles

a, h Wind profiles
b, i Turbine rotating speed
c, j Instantaneous power coefficient
d, k Captured wind power
e, l Converted electrical power
f, m Boost converter input voltage
g, n Boost converter input current

106 IET Renew. Power Gener., 2011, Vol. 5, Iss. 1, pp. 99 –108
& The Institution of Engineering and Technology 2010 doi: 10.1049/iet-rpg.2009.0207
 www.ietdl.org
8l. Although during the wind speed decrease the captured during this period (rectifier loss ignored). Doubling the
wind power is less than the rated power, the WECS is able generator resistance results in a 4.2% loss of electrical
to output a power PN for a short period. power capability, while the increased generator resistance
Cp performance evaluation during a constant power control does not significantly affect the WECS general wind
mode is not critical for the design of the variable speed WECS. conversion efficiency, as the shown 0.03% conversion
For example, the worst Cp ¼ 0.27 in this profile occurs during efficiency reduction. This is because an increased generator
gust speed decrease, whereas the WECS outputs the maximum resistance causes reverse Cp performance at wind speed
allowed power. Consequently, the Cp performance changing conditions.
improvement under this condition is less significant. For the wind profile in Fig. 8h, with a stable and ramp wind
Fig. 8 also shows the Cp performances of OPP operation on condition, the OPP operated WECS has a wind energy
the variable speed WECS during stable wind conditions and converting efficiency of 98.95%. Under such conditions, a
the ramp wind speed changes are high. Deviations of Cp lower inertia turbine improves the WECS wind energy
from the maximum value are still observed due to the conversion efficiency, as the shown 0.56% efficiency
damping effect of mechanical/electrical systems at large improvement with a turbine inertia of 1 kg m2.
instant wind variations. The lowest Cp values at such
instants in the presented wind profile are generally greater 5 Conclusion
than 0.362 and the OPP operation instantaneously reacts,
converging Cp to the maximum Cp for tracking the In this paper, an OPP-based operating strategy is proposed for
corresponding maximum wind power. variable speed wind/tidal stream turbines with synchronous
generators. The approach needs the renewable energy
4.2.1 OPP wind energy converting efficiency conversion system to be initialised with OPP on the
evaluation with simulation: In simulation, the WECS maximum power curve, where the initialisation is by any of
wind power converting efficiency is evaluated by the the conventional maximum power tracking methods at an
following equation unknown wind/tidal speed. The control method is based on
the generator/rectifier set output voltage being proportional
t 2 to turbine speed. Once initialised, the renewable energy
t1
Cp
ho = t2 (19) conversion system is used without knowledge of the wind/
t1
Cp max tidal or generator speed. Simulation results demonstrated
the OPP approach.
where t1 – t2 is the evaluating period, and Cp is the
instantaneous wind turbine power coefficient during the 6 Acknowledgment
period.
OPP wind energy converting efficiencies for different wind The support of the Technology Strategy Board and the
conditions are obtained under the following conditions and EPSRC is acknowledged under the Collaborative Research
are shown in Tables 3 and 4: and Development grant ‘Multi-terminal DC transmission
micro-grid system for interfacing multiple wind turbines’,
† Pbase is located at (TSR ¼ 6.04, Cp ¼ 0.39785) on the TP/8/LOW/9/I/Q3068G.
Cp_TSR curve, with a 99.96% wind energy conversion
efficiency at a linear wind speed of 8.5 m/s. 7 References
† The base, gust and ramp components are unchanged with a
random noise component [32]. 1 Khan, J., Moshref, A., Bhuyan, G.: ‘A generic outline for dynamic
† Turbine friction effects on the WECS are ignored. modeling of ocean wave and tidal current energy conversion systems’.
Proc. IEEE PES’09, Calgary, AB, July 2009, pp. 1– 6
2 Khan, J., Bhuyan, G., Moshref, A., Morison, K., Pease, J.H., Gurney, J.:
For the wind profile in Fig. 8a during the period 1 – 11 s, the ‘Ocean wave and tidal current conversion technologies and their
OPP wind conversion efficiency is 98.5%. Simulation results interaction with electrical networks’. Proc. IEEE PES’08 General
show that the generated average electrical power is 2157.4 W Meeting-Conversion and Delivery of Electrical Energy in the 21st
Century, Pittsburgh, PA, July 2008, pp. 1– 8
3 Bryans, A.G., Fox, B., Crossley, P.A., O’Malley, M.: ‘Impact of tidal
Table 3 OPP wind energy conversion efficiency generation on power system operation in Ireland’, IEEE Trans. Power
Syst., 2005, 20, (4), pp. 2034–2040
Wind Period Conditions Ra loss Conversion 4 King, J., Tryfonas, T.: ‘Tidal stream power technology-state of the art’.
profile (W) efficiency, % Proc. OCEANS’09, Bremen, May 2009, pp. 1 –8
5 Tuckey, A.M., Patterson, D.J., Swenson, J.: ‘A kinetic energy tidal
Fig. 8a 1–11 s J ¼ 6 kg m2, 95.5 98.50 generator in the northern territory-results’. Proc. 23rd Int. Conf. on
Ra ¼ 1.4 V IECON 97, New Orleans, LA, November 1997, vol. 2, pp. 937–942
J ¼ 6 kg m2, 185.0 98.47 6 Benelghali, S.E., Benbouzid, M.E.H., Charpentier, J.F.: ‘Modeling and
Ra ¼ 2.8 V control of a marine current turbine driven doubly-fed induction
generator’, IET Renew. Power Gener., 2010, 4, (1), pp. 1– 11
7 Benelghali, S.E., Benbouzid, M.E.H., Ahmed-Ali, T., Charpentier, J.F.:
‘High-order sliding mode control of DFIG-based marine current
Table 4 OPP wind energy conversion efficiency
turbine’. Proc. IEEE IECON’08, Orlando, Florida, November 2008,
pp. 1228– 1233
Wind Period Conditions Conversion efficiency,
8 Barakati, S.M., Kazerani, M., Aplevich, J.D.: ‘A mechanical speed-
profile % sensorless maximum power tracking control for a wind turbine system
including a matrix converter’. Proc. IEEE PES’08 General Meeting –
Fig. 8h 4– 8 s J ¼ 6 kg m2, 98.95 Conversion and Delivery of Electrical Energy in the 21st Century,
Ra ¼ 1.4 V Pittsburgh, PA, July 2008, pp. 1– 8
J ¼ 1 kg m2, 99.51 9 Valenciaga, F., Puleston, P.F.: ‘High-order sliding control for a wind
Ra ¼ 1.4 V energy conversion system based on a permanent magnet synchronous
generator’, IEEE Trans. Energy Convers., 2008, 23, (3), pp. 860 –867

IET Renew. Power Gener., 2011, Vol. 5, Iss. 1, pp. 99 –108 107
doi: 10.1049/iet-rpg.2009.0207 & The Institution of Engineering and Technology 2010
www.ietdl.org
10 Beltran, B., Ahmed-Ali, T., Benbouzid, M.E.H.: ‘Sliding mode power 24 Esmaili, R., Xu, L.Y.: ‘Sensorless control of permanent magnet
control of variable-speed wind energy conversion systems’, IEEE generator in wind turbine application’. Conf. Rec. 41st IEEE IAS
Trans. Energy Convers., 2008, 23, (2), pp. 551– 558 Annual Meeting, Tampa, FL, October 2006, vol. 4, pp. 2070–2075
11 Ochoa, L.F., Padiha-Feltrin, A., Harrison, G.P.: ‘Time-series-based 25 Wai, R.J., Lin, C.Y., Chang, Y.R.: ‘Novel maximum-power-extraction
maximization of distributed wind power generation integration’, IEEE algorithm for PMSG wind generation system’, IET Electr. Power
Trans. Energy Convers., 2008, 23, (3), pp. 968– 974 Appl., 2007, 1, (2), pp. 275– 283
12 Chen, Z., Guerrero, J.M., Blaabjerg, F.: ‘A review of the state of the art 26 Molina, M.G., Mercado, P.E.: ‘A new control strategy of variable speed
of power electronics for wind turbines’, IEEE Trans. Power Electron., wind turbine generator for three-phase grid-connected applications’.
2009, 24, (8), pp. 1859– 1875 Proc. IEEE/PES Transmission and Distribution Conference and
13 Shen, B., Mwinyiwiwa, B., Zhang, Y.Z., Ooi, B.-T.: ‘Sensorless Exposition: Latin America, Bogota, August 2008, pp. 1 –8
maximum power point tracking of wind by DFIG using rotor position 27 Koutroulis, E., Kalaitzakis, K.: ‘Design of a maximum power tracking
phase lock loop (PLL)’, IEEE Trans. Power Electron., 2009, 24, (4), system for wind-energy-conversion applications’, IEEE Trans. Ind.
pp. 942–951 Electron., 2006, 53, (2), pp. 486–494
14 Qiao, W., Zhou, W., Aller, J.M., Harley, R.G.: ‘Wind speed estimation 28 Wang, Q., Chang, L.C.: ‘An intelligent maximum power extraction
based sensorless output maximization control for a wind turbine driving algorithm for inverter-based variable speed wind turbine systems’,
a DFIG’, IEEE Trans. Power Electron., 2008, 23, (3), pp. 1156–1169 IEEE Trans. Power Electron., 2004, 19, (3), pp. 1242– 1249
15 Abo-Khalil, A.G., Lee, D.-C.: ‘MPPT control of wind generation 29 Wang, Q.: ‘Maximum wind energy extraction strategies using power
systems based on estimated wind speed using SVR’, IEEE Trans. Ind. electronic converters’. PhD thesis, University of New Brunswick,
Electron., 2008, 55, (3), pp. 1489–1490 Fredericton, NB, Canada, 2003
16 Carrasco, J.M., Franquelo, L.G., Bialasiewicz, J.T., et al.: ‘Power- 30 Burton, T., Sharpe, D., Jenkins, N., Bossanyi, E.: ‘Wind energy
electronic systems for the grid integration of renewable energy handbook’ (John Wiley & Sons, New-York, 2001)
sources: a survey’, IEEE Trans. Ind. Electron., 2006, 53, (4), 31 Knight, A.M., Peters, G.E.: ‘Simple wind energy controller for an
pp. 1002– 1016 expanded operating range’, IEEE Trans. Energy Convers., 2005, 20,
17 Beltran, B., Ahmed-Ali, T., Benbouzid, M.E.H.: ‘High-order sliding- (2), pp. 459– 466
mode control of variable-speed wind turbines’, IEEE Trans. Ind. 32 Anderson, P.M., Bose, A.: ‘Stability simulation of wind turbine
Electron., 2009, 56, (9), pp. 3314–3321 systems’, IEEE Trans. Power Apparatus and Syst., 1983, PAS-102,
18 Yazdani, A., Iravani, R.: ‘A neutral-point clamped converter system for (12), pp. 3791–3795
direct-drive variable-speed wind power unit’, IEEE Trans. Energy
Convers., 2006, 21, (2), pp. 596–607
19 Dai, J.Y., Xu, D.W., Wu, B.: ‘A novel control system for current source 8 Appendix
converter based variable speed PM wind power generators’. Proc. IEEE
PESC’07, Orlando, FL, June 2007, pp. 1852– 1857 8.1 Tidal stream turbine hydrodynamic
20 Chinchilla, M., Arnaltes, S., Burgos, J.C.: ‘Control of permanent- characteristics
magnet generators applied to variable-speed wind-energy systems
connected to the grid’, IEEE Trans. Energy Convers., 2006, 21, (1), The mechanical power captured from the tidal current is
pp. 130–135
21 Haque, M.E., Negnevitsky, M., Muttaqi, K.M.: ‘A novel control strategy
expressed as [3 –7]
for a variable speed wind turbine with a permanent magnet synchronous
generator’. Proc. IEEE IAS’08, Edmonton, Alta, October 2008, pp. 1– 8 1
22 Inoue, Y., Morimoto, S., Sanada, M.: ‘Control method for direct torque Pm = rACp v3t
2
controlled PMSG in wind power generation system’. Proc. IEMDC’09,
Miami, FL, 3 –6 May 2009, pp. 1231– 1238
23 Hai, N.T., Jang, S.-H., Park, H.-G., Lee, D.-C.: ‘Sensorless control of
where r is the fluid density (kg/m3); A is the tidal turbine
PM synchronous generators for micro wind turbines’. Proc. IEEE swept area (m2); Cp is the power coefficient of the tidal
PECon’08, Johor Bahru, December 2008, pp. 936–941 stream turbine and vt is the fluid speed (m/s).

108 IET Renew. Power Gener., 2011, Vol. 5, Iss. 1, pp. 99 –108
& The Institution of Engineering and Technology 2010 doi: 10.1049/iet-rpg.2009.0207