International Journal of Applied Power Engineering (IJAPE)

Vol. 11, No. 4, December 2022, pp. 304~318

ISSN: 2252-8792, DOI: 10.11591/ijape.v11.i4.pp304-318  304

Power loss minimization assessment of a doubly fed induction

generator with variable core resistance for wind turbines
operation

Crescent Onyebuchi Omeje1, Damian Benneth Nnadi2

1
Department of Electrical/Electronic Engineering, University of Port Harcourt, Choba, Port Harcourt, Rivers State, Nigeria
2
Department of Electrical Engineering, University of Nigeria, Nsukka Enugu State, Nigeria

Article Info ABSTRACT

Article history: This paper assesses the efficiency level and power loss minimization of a
doubly fed induction generator (DFIG). A modified DFIG equivalent circuit
Received Oct 1, 2022 with multi-core resistance connected in parallel was adopted. State-space
Revised Oct 30, 2022 differential equations of the DFIG was developed incorporating iron and
Accepted Nov 7, 2022 copper loss components while a minimum flux linkage that aids in the
minimization of the overall losses was derived. Simulation results showed
that losses were minimized when the equivalent core resistances were
Keywords: connected in parallel with minimum permissible current flow. The results
obtained during a transient disturbance showed that at different core
Doubly fed induction generator resistance values of Rfe = 0.75Ω and 0.25Ω, different efficiency values of
Hessian matrix 83.45% and 41.21% were realized. An unconstrained optimization test
Power loss modeling carried out on the DFIG variable parameters showed that the DFIG power
Simulations loss model was controllable with a positive definite value of 691.9801 and
Transient stability 2.9156〖e〗^(+5) for the leading principal determinants of the Hessian
matrix. All simulation processes were achieved in MATLAB/Simulink 2020.
This is an open access article under the CC BY-SA license.

Corresponding Author:
Crescent Onyebuchi Omeje
Department of Electrical/Electronic Engineering, University of Port Harcourt
East-West Road, PMB, 5323 Choba, Port Harcourt Rivers State, Nigeria
Email: [email protected]

1. INTRODUCTION
Wind energy in addition to photovoltaic energy source is the fastest growing energy with an annual
growth rate of 30% and with a predictable penetration of 12% of the global electricity demand [1]. The use of
power electronic converters allows for a variable speed operation of the wind turbine where the wind energy
conversion system extracts maximum power from the turbine during peak operation as reported in [2]. One
of the energy generation systems commercially available in the wind energy market is the doubly fed
induction generator (DFIG) and its numerous advantages are enumerated in [2], [3]. To produce a maximum
energy, an efficient DFIG with higher power rating is a precondition for wind energy conversion system
(WECS) though active power loss during the operation is the main drawback for efficiency optimization and
life expectancy’s challenge. Variable speed wind turbines are usually more effective due to their improved
efficiency in capturing more wind power and their innate ability to achieve higher quality of power at
optimum wind speed [4]. DFIG implementation is increasing in leaps and bounds as a consequence of a
reduced mechanical stress and noise in addition to the flexible control of active and reactive power which is
based on the back-to-back power electronics converter sandwiched between the induction machine and the
power grid [5]. Active power generation at the stator and rotor terminals is effectively controlled using the
rotor side converter (RSC). The regulation of the stator side active and reactive power is independently

Journal homepage: http://ijape.iaescore.com

Int J Appl Power Eng ISSN: 2252-8792  305

achieved with the RSC. However, the fluctuations in power output from the DFIG-based wind generation in
response to the variations in wind speed adversely affect the needed power quality [6]. Therefore, with an
increased penetration of the DFIG in power system operation, inertia of the power system is reduced. An
increased capacity in grid integration of this variable power poses an impactful challenge on power system
stability [7]. In line with the system stability is the minimization of the overall system losses under different
wind speed and reactive power conditions. Therefore, to improve on the system overall efficiency, the
reactive power flow is regulated with a minimized system electrical loss on the generator and on the power
electronic converters.
In this paper, a modified equivalent circuit of the DFIG with variable core resistance values for loss
minimization was adopted. The copper and iron losses of the DFIG were modeled as a function of the rotor
dq-axes currents and stator flux. At optimum condition, the stator d-axis flux that minimizes the total DFIG
loss was derived. The simulation results obtained showed that the efficiency values of 83.45% and 41.21%
were obtained when the equivalent paralleled core resistance values of the DFIG were kept at 0.75Ω and
0.25Ω. The power loss at a slight transient disturbance peaked at 202.277 kW and reduced to 17.2058 kW
under steady state for 0.75Ω net resistance. Similarly, at a net core resistance value of 0.25Ω, the power loss
during a transient disturbance peaked at 757.325 kW and reduced to 23.7597 kW. The variations in the power
loss at different net-core resistance accounted for the difference in the efficiency values as presented in the
preceding simulation waveforms.
This paper is organized as follows: Section 1 is the introduction; Section 2 is the reviewed literature
on DFIG loss control; Section 3 illustrates the wind turbine model of doubly fed induction generator;
Section 4 presents the DFIG mathematical modeling and power loss minimization scheme; and Section 5
contains the simulation results. Section 6 is the conclusion.

2. METHODOLOGY AND REVIEWED LITERATURE ON DFIG POWER LOSS CONTROL

Research procedures and methods applied in this paper involved mathematical modeling of the doubly
fed induction generator at variable core resistances to determine its effects on the overall power loss of the
system. Stability test using the Sylvester criterion method and data acquisition derived from simulation
parameters were parts of the methods applied in the empirical analysis of the system controllability. In [8]–[10],
direct power control (DPC) and direct torque control of the DFIG was proposed as a conventional control
scheme. However, large torque and power ripples are the two major draw backs associated with this method.
The various methods involving back stepping direct power control strategy for DFIG have been presented in
[11]–[14]. Although a detailed analysis was presented but emphasis on variable core resistance and minimum
flux at which the DFIG power loss is minimized was not presented.
In [15], [16], a sliding mode approach in minimizing copper loss was presented in a nonlinear
control application while a coordinated predictive control of the DFIG using non-Gaussian wind power
predictive distribution was reported in [17]. The reports concentrated on regulating the extracted power from
the wind turbine while incorporating machine loss minimization technique for iron-loss. In [18], [19],
minimization of the active and reactive power ripples using the direct power control of matrix converter fed
DFIG was presented without reference to the DFIG core losses minimization.
This paper therefore analyzed a complete loss minimization of the DFIG using the parameter
variation that involves the core resistance and stator flux. A complete DFIG loss equation was modeled and
derived from a modified equivalent circuit involving the variable core resistance while a minimal flux
equation that minimized the DFIG losses was also derived. The Sylvester criterion using the Hessian matrix
was applied to determine if the DFIG total power loss was minimized within the context of the variable core
resistance values chosen. Therefore, with the leading principal determinants positive for Req = Rfe = 0.25Ω
and 0.75Ω, it showed that the overall total loss of the DFIG is controllable and so is minimized.

3. WIND TURBINE MODEL OF DOUBLY FED INDUCTION GENERATOR.

The schematic diagram of the DFIG wind turbines operation is presented in Figure 1. The DFIG
stator is directly connected to the grid through the grid side converter GSC and its rotor terminal is connected
to the rotor side converter RSC through the slip rings. The wind turbine is modeled with reference to optimal
power tracking to provide a maximum energy capture from the wind. The aerodynamic model of the wind
turbine gives a coupling between the wind speed and the mechanical torque produced by the wind turbine.
Aerodynamic is a science of physical laws that deals with the behavior of objects in airflow and forces that
are produced by this airflow. The aerodynamic power equation extracted from the wind turbines is given by
(1) as reported in [20]–[22].

Power loss minimization assessment of a doubly fed induction generator … (Crescent Onyebuchi Omeje)
306  ISSN: 2252-8792

1
Pm = × ρ × A × V 3 × Cp × (λ, β) (1)
2

Where Pm (W) is the mechanical power of the turbine, ρ (Kg/m 3) is the air density, A (m2) is the area
covered by the rotor turbine, V (m/s) is the wind speed upstream of the rotor and C p is the performance
coefficient or power coefficient. The power coefficient is a function of the pitch angle (β) of the rotor blades.
The tip speed ratio is the ratio between the blade tip speed and wind speed upstream of the rotor. Altering the
pitch angle implies slightly rotating the turbine blades along the horizontal or vertical axis. C p which
represents the wind turbine power coefficient is given by (2).
12.5
1.16 −
Cp (λ, β) = 22 × ( − 0.004β − 0.05) e λi
(2)
λi

1 1 0.035
= − (3)
λi λ +0.08β β3 +1

Where β represents the blade pitch angle and λ represents the tip speed ratio which is given by (4).
ω r Rr
λ = (4)
V

The dynamic model wind turbine is associated with the rotor speed ω r and gear box ratio ηg by (5).

𝑜𝑝𝑡𝑖𝑚𝑢𝑚
𝜔𝑟 = η𝑔 × 𝜔𝑟 (5)

The exact dynamic model of the torque equation for the generator is given by (6).

Pm πρR3
Tm = = × Vw2 × Ct (λ) (6)
ωr 2

Where 𝜌 = air density (Kgm-3); R = radius of the turbine (m); Vw= wind speed (MS-1); Ct (λ) = Torque
coefficient (pu). The wind turbine characteristics are presented in Figure 2.
The plot of power coefficient against the tip speed ratio at varying blade pitch angle is represented
in Figure 2. It is observed that the power coefficient increases as the pitch angle decreases. Therefore, at a
zero-pitch angle, the maximum power coefficient is 0.5 which implies that the maximum power is tracked at
a zero-blade pitch and tip speed ratio value of ten as shown in Figure 2.

Figure 1. Schematic diagram of a DFIG operated wind turbine

Int J Appl Power Eng, Vol. 11, No. 4, December 2022: 304-318
Int J Appl Power Eng ISSN: 2252-8792  307

Figure 2. A plot of wind turbine power coefficient against tip-speed ratio at varying pitch angle

4. DFIG MATHEMATICAL MODELING AND POWER LOSS MINIMIZATION SCHEME.

Stator and rotor voltage as well as flux equations as reported in [23], [24] are given in (7)–(14)

dλds
Vds = R s ids + − ωe λqs (7)
dt

dλqs
Vqs = R s iqs + + ωe λds (8)
dt

dλdr
Vdr = R r idr + − ωe λqr (9)
dt

dλqr
Vqr = R r iqr + + ωe λdr (10)
dt

λds = LLs ids + Lm idm (11)

λqs = LLs iqs + Lm iqm (12)

λdr = LLr idr + Lm idm (13)

λqr = LLr iqr + Lm iqm (14)

Modified voltage equations can be obtained by substituting (11)–(14) into (7)–(10).

dids didm
Vds = R s ids + LLs + Lm − ωe (LLs iqs + Lm iqm ) (15)
dt dt

diqs diqm
Vqs = R s iqs + LLs + Lm + ωe (LLs ids + Lm idm ) (16)
dt dt

didr didm
Vdr = R r idr + LLr + Lm − ωe (LLr iqr + Lm iqm ) (17)
dt dt

diqr diqm
Vqr = R r iqr + LLr + Lm + ωe (LLr idr + Lm idm ) (18)
dt dt

Modified equivalent circuit model of the DFIG with multiple paralleled core resistance is obtained from
(15)–(18) and presented in Figures 3 and 4. Branch currents are obtained from Figures 3 and 4 by applying
KCL as expressed in (19) and (20).

[−ωe Lm iqm ]
ids + idr = idm + = idm − Aiqm (19)
Req

Power loss minimization assessment of a doubly fed induction generator … (Crescent Onyebuchi Omeje)
308  ISSN: 2252-8792

[ωe Lm idm ]
iqs + iqr = iqm + = iqm + Aidm (20)
Req

A simultaneous solution of (19) and (20) by substitution gives rise to (21) and (22).

[(ids +idr ) +A(iqs + iqr )]

idm = (1 + A2 )
(21)

[(iqs +iqr ) +A(ids + idr )]

iqm = (1 + A2 )
(22)

ωe Lm
Where: A = . For vector controlled condition, 𝜆𝑞𝑠 = 0 and 𝜆𝑑𝑠 = 𝜆𝑠 substituting this into (11) and
Req
(12) gives rise to (23) and (24).
−Lm
iqs = i (23)
LLs qr

𝜆𝑠 −Lm idr
ids = (24)
LLs

Figure 3. D-axis equivalent circuit of a DFIG with multiple paralleled core resistance

Figure 4. Q-axis equivalent circuit of a DFIG with multiple paralleled core resistance

Therefore, the dq-axes magnetizing currents under vector control is obtained by substituting (23)
and (24) into (21) and (22) as re-presented in (25) and (26).

[λs +(LLs −Lm )(idr + Aiqr )]

idm = (25)
LLs (1 + A2 )

Int J Appl Power Eng, Vol. 11, No. 4, December 2022: 304-318
Int J Appl Power Eng ISSN: 2252-8792  309

[(LLs −Lm )(iqr − Aidr )−Aλs ]

iqm = (26)
LLs (1 + A2 )

The dq-axes magnetizing voltages are presented in (27) and (28).

Aωe Lm λs ωe Lm LLs (Aidr −iqr )

Vdm = −ωe Lm iqm = + = A1 Aλs + A2 (Aidr − iqr ) (27)
LLs (1 + A2 ) LLs (1 + A2 )

ωe Lm λs ωe Lm LLs (idr + Aiqr )

Vqm = −ωe Lm idm = + = A1 λs + A2 (idr + Aiqr ) (28)
LLs (1 + A2 ) LLs (1 + A2 )

ωe Lm
Where: A1 = andA2 = LLs A1 .
LLs (1 + A2 )
The voltage magnitude across the magnetizing branch is presented in (29).

2
2
Vm = Vdm 2
+ Vqm = A22 (1 + A2 )[i2dr + i2qr ] + 2A1 A2 (1 + A2 )λs idr + A21 (1 + A2 )λ2s (29)

Total core loss in terms of magnetizing voltage and equivalent core resistance is given by (30).

(Vm )2 1
P Total
core = = A22 (1 + A2 )(i2r ) + 2A1 A2 (1 + A2 )λs idr + A21 (1 + A2 )λ2s (30)
Req Req

Where: i2r = i2dr + i2qr and R eq = 𝑅𝑓𝑒 ‖𝑅𝑓𝑒 ‖𝑅𝑓𝑒 ‖𝑅𝑓𝑒……𝑛+1 are derived from Figures 3 and 4. The stator and
rotor copper losses with core losses which are the total resistive losses of the DFIG are presented in (31).

Total
PResistive = R s [i2ds + i2qs ] + R r [i2dr + i2qr ] + R fe [i2qfe + i2dfe ] (31)

Substituting (23) and (24) into (31) gives rise to (32).

Rs (λs )2 Rs L2 2Rs Lm Rs L2 2
P Total
Resistive = + m
(idr )2 − λs idr + m
(iqr ) + R r i2r + R fe [i2qfe + i2dfe ] (32)
L2
Ls L2
Ls L2
Ls L2
s

Other types of power losses associated with the DFIG operated system are the RL-filter losses,
Frictional losses and power electronics converter losses as presented in (33)–(35) respectively.

PRL−Filter = R Filter (i2d−Filter + i2q−Filter ) (33)

PFrictiion = K Fr ω2mech (34)

PConverter = R Converter (i2d−Converter + i2q−Converlter ) (35)

The total losses of the DFIG scheme can be written in terms of the total core loss, resistive loss, RL-filter
losses, Frictional losses and power electronics converter losses as given in (36).
1
P Total Total Total
Loss = P core + P Resistive + PRL−Filter + PFrictiion + PConverter = [A22 (1 + A2 )i2r +
Req
Rs (λs )2 Rs L2 2R L
2A1 A2 λs idr (1 + A2 ) + A21 λ2s (1 + A2 )] + + m 2
ir − s2 m λs idr + R r i2r + R fe [i2qfe + i2dfe ] +
L2
Ls L2
Ls LLs
R Filter (i2d−Filter + i2q−Filter ) + K Fr ω2mech + R Converter (i2d−Converter + i2q−Converlter ) (36)

The flux at which Total Power loss is minimized is determined by setting the derivative of (36) to zero.

𝑑P Total
Loss 1 2λs R s 2R s Lm
=0= [2A1 A2 idr (1 + A2 ) + 2λs A21 (1 + A2 )] + 2 − i =0
𝑑λs R eq LLs L2Ls dr
[Rr Req Lm −(1+A2 )A1 A2 L2
Ls ]idr
λ smin = 2 2 (37)
A2
1 (1+A )LLs + Rs Req

The flux at which minimum loss occurred is as presented in (37). This implies that the minimum flux
increases with a proportionate rise in the rotor direct axis current and equivalent core resistance values.
Therefore, the overall total loss is minimized when λ smin is substituted for λ s in (36) that gives rise to (38).

Power loss minimization assessment of a doubly fed induction generator … (Crescent Onyebuchi Omeje)
310  ISSN: 2252-8792

1
P Total Total Total
Loss−min = P core−min + P Resistive−min + PRL−Filter + PFrictiion + PConverter = [A22 (1 +
Req
Rs (λ smin )2 Rs L2 2Rs Lm
A2 ) i2r + 2A1 A2 λ smin idr (1 + A2 ) + A21 λ2smin (1 + A2 )] + + m 2
ir − λ
smin dr i +
L2
Ls L2
Ls L2
Ls
R r i2r + R fe [i2qfe + i2dfe ] + R Filter (i2d−Filter + i2q−Filter ) + K Fr ω2mech + 2
R Converter (id−Converter +
i2q−Converlter ) (38)

The DFIG power loss minimization technique applied on (38) was actualized with the Sylvester’s
criterion for positive definite state using the Hessian matrix determinant test. The objective function to be
minimized is the overall total power loss already given in (38). The Hessian matrix is obtained by taking the
second derivatives of (38) in terms of stator flux and rotor current as presented (39)–(42).

d2 PTotal 1 2Rs
H11 = Loss−min
= [ 2(1 + A2 )A21 ] + (39)
d2 λsmin Req L2
s

d2 PTotal
Loss−min
H12 = = 0 (40)
dλsmin dir

d2 PTotal
Loss−min
H21 = = 0 (41)
dir dλsmin

d2 PTotal 1 2𝐿2𝑚Rs
H22 = Loss−min
= [ 2(1 + A2 )A22 ] + + 2𝑅𝑟 (42)
d2 i r Req L2
s

The Hessian matrix equation is presented in (43) for the DFIG total loss parameter stability test.

1 2Rs
H11 = [2(1 + A2 )A21 ] + H12 = 0
Req L2
s
H = [ 1 2L2
] (43)
m Rs
H21 = 0 H22 = [ 2(1 + A2 )A22 ] + + 2R r
Req L2
s

The total power loss is minimized if the leading principal determinants of (43) are positive and greater than
zero. The principal determinants are obtained with the DFIG parameters in Table 2 substituted to give the
following:

1 2R s
|H11 | = | [ 2(1 + A2 )A21 ] + 2 | = 691 . 9801
R eq Ls

1 2R s
[ 2(1 + A2 )A21 ] + 2 0
H H12 R eq Ls
| 11 | = || |
|
H21 H22 1 2 2
2L2m R s
0 [ 2(1 + A )A2 ] + + 2R r
R eq L2s
2
1 2 )A2 ]
2R s 2 2
2Lm R s
= 2 [([2(1 + A 1 + 2 ) ([2(1 + A )A 2 ] + + 2R r ) − 0]
R eq Ls L2s
= 2.9156𝑒 + 05

The parameters substituted above showed that the DFIG power loss model was controllable with a
positive definite value of 691.9801 and 2.9156 e +5 for the leading principal determinants of the Hessian
matrix. For dynamic simulation analysis of the doubly fed induction generator under a load disturbance, the
state space model equations were derived with the aid of Figures 5 and 6 and presented in (44)–(57).
Applying the Kirchhoff’s current law at nodes A and B on the simplified DFIG dq-axes equivalent circuit
diagram presented in Figures 5 and 6 gives rise to (44)–(47).

ids + idr = idfe + idm (44)

iqs + iqr = iqfe + iqm (45)

idfe = ids + idr − idm (46)

Int J Appl Power Eng, Vol. 11, No. 4, December 2022: 304-318
Int J Appl Power Eng ISSN: 2252-8792  311

iqfe = iqs + iqr − iqm (47)

Figure 5. D-axis equivalent circuit of a DFIG with single core resistance

Figure 6. Q-axis equivalent circuit of a DFIG with single core resistance

Sets of current differential equations applied in the DFIG simulations are presented in (48) to (53).

dids 1
= (Vds − R s ids + ωe LLs iqs − R fe idfe ) (48)
dt LLs

diqs 1
= (Vqs − R s iqs − ωe LLs ids − R fe iqfe ) (49)
dt LLs

didr 1
= (Vdr − R r idr − ωe Lm iqm + ωe LLr iqr − R fe idfe ) (50)
dt LLr

diqr 1
= (Vqr − R r iqr − ωe Lm idm + ωe LLr idr − R fe iqfe ) (51)
dt LLr

didm 1
= (R fe idfe + ωe Lm iqm ) (52)
dt Lm

diqm 1
= (R fe iqfe − ωe Lm idm ) (53)
dt Lm

Similarly, derivatives of (46) and (47) for a change in the core current gave rise to (54) and (55).

didfe dids didr didm

= + − (54)
dt dt dt dt

diqfe diqs diqr diqm

= + − (55)
dt dt dt dt

Power loss minimization assessment of a doubly fed induction generator … (Crescent Onyebuchi Omeje)
312  ISSN: 2252-8792

The estimated stator active and reactive powers are expressed in (56).
3
Ps = [Vds ids + Vqs iqs ]
2
3 } (56)
Qs = [Vqs ids − Vds iqs ]
2

Also, the estimated rotor active and reactive powers are expressed in (57).
3
Pr = [Vdr idr + Vqr iqr ]
2
3 } (57)
Qr = [Vqr idr − Vdr iqr ]
2

The mechanical dynamics of the DFIG based on speed and torque are presented in (58)–(60).
3
Tem = × Lm (idm iqr − iqm idr ) (58)
2

dωr 1
= × (Te − TL − βωr ) (59)
dt J

1
𝜔𝑟 = ∫ J × (Te − TL − βωr ) (60)

The efficiency of the machine in terms of power output and total losses of the DFIG is expressed in (61).
Power Output Tem ×ωr
⟅ = = (Tem ×ωr ) + P Total
× 100% (61)
Power Output+Total Losses Loss

5. RESULTS AND DISCUSSION

The simulation parameters are presented in Tables 1 and 2. A plot of the derived flux linkage against
the d-axis rotor current at varying core resistance is shown in Figure 7(a). It is observed that the flux linkage
increases in linear proportion with the d-axis current with a corresponding rise in the slope and attains a
R
higher value with a reduced equivalent core-resistance (R eq = fe ). In Figure 7(b) an exponential rise in the
10
total core loss was observed. The deviation in the exponential curve with higher value of core loss (900 kW)
R
is more pronounced when R eq = fe which implies that more field current and circulating flux is produced at
10
a lower core-resistance. In Figure 8(a), a plot of the total resistive loss against current was presented at a
varying rotor resistance. It is also observed that as the resistance value is increased to R r = 40 × 2.25Ω, the
total resistive loss obtained peaked at 2000kW at a current value of 3.25A as against 3.5A got when the rotor
resistance was reduced which accounted for the higher current flow through the path of least resistance. In
Figure 8(b), it is observed that the DFIG total power loss rose to a value of 3250 kW at a reduced core
R
resistance value of R eq = fe and higher rotor resistance value of R r = 40 × 2.25Ω with a rise in slope. This
10
indicates that at reduced core resistance, more field current is drawn since more paralleled resistance is
needed to achieve this process whereas with an increased rotor resistance, more power is dissipated across
the rotor terminal. The dq-axes voltages for the grid side converter (GSC) and rotor side converter (RSC) are
presented in Figures 9(a) and 9(b). It is observed that the voltage magnitude is maintained at 450V during the
simulation period thereby ensuring system stability in supply. In Figures 10(a) and 10(b), dq-axes rotor currents
are shown at variable core resistance values of R fe = 0.75Ω and 0.25Ω Similarly, in Figures 11(a) and 11(b),
the plots obtained for the dq-axes core current showed that transient disturbance has more impacts on the
DFIG core winding at a reduced core resistance value of R fe = 0.25Ω and this accounted for its low
efficiency value. In Figures 12(a) and 12(b), a sharp deviation in the settling time of the dq-axes magnetizing
currents were observed at different core resistance values. The plots for the motor speed and electromechanical
torques for different core resistance values are shown in Figures 13(a) and 13(b). It is observed that the transient
response in speed and torque is reduced during the period of transient disturbance for R fe = 0.75Ω but more
oscillatory transient response was obtained after the disturbance was cleared. The power losses and power
outputs obtained during a transient disturbance at core resistance values of R fe = 0.75Ω and 0.25Ω are shown
in Figures 14(a) and 14(b). It is shown that more losses were obtained in Figure 14(b) with a value of
757.325 kW than in Figure 14(a) that dissipated 202.274 kW. This accounted for the large variation in their
efficiency value of 83.45% and 41.21%.

Int J Appl Power Eng, Vol. 11, No. 4, December 2022: 304-318
Int J Appl Power Eng ISSN: 2252-8792  313

In Figures 15(a) and 15(b), the stator active and reactive powers are presented. It is observed that a
rapid oscillation was obtained during a transient disturbance which is more pronounced when the core
resistance is kept at R fe = 0.75Ω. Conversely, in Figures 16(a) and 16(b), during a transient disturbance a
rapid transient oscillation was observed at R fe = 0.25Ω for rotor active and reactive power which implies
that more current is drawn when R fe = 0.25Ω.

Table 1. Simulation parameters of the turbine blade model

Parameters Values
Blade radius (m) 1
Maximum power coefficient (Cp) 0.5
Optimal tip speed ratio 10
Cut in speed (m/s) 4.5
Rated wind speed (m/s) 13.5
Minimum Pitch angle (degree) 0
Maximum Pitch angle (degree) 30

Table 2. Simulation parameters used for the DFIG dynamic model

Parameters Values
Rated power (MW) 15
Rated mechanical power (MW) 10
Supply Voltage (V) 400
Supply Frequency (Hz) 50
Stator resistance (pu) 0.25
Rotor resistance (pu) 0.45
Stator leakage inductance (pu) 0.0877
Rotor leakage inductance (pu) 0.077
Mutual inductance (pu) 0.955
Core resistance of modified circuit (Ω) 2.25
Core resistance of simplified circuit (pu) 0.25 and 0.75
Filter resistance (pu) 0.03
Filter inductance (pu) 0.25
Converter resistance (pu) 0.001
Motor inertia Kg-M2) 0.045
Coefficient of Viscosity (NMS) 0.00006
Pole pair 2
Load torque (NM) 20

(a) (b)

Figure 7. A plot of (a) flux linkage (V.S) against current (A) and
(b) total core loss (W) against flux linkage (V.S)

Power loss minimization assessment of a doubly fed induction generator … (Crescent Onyebuchi Omeje)
314  ISSN: 2252-8792

(a) (b)

Figure 8. A plot of (a) total resistive loss (W) against current (A) and
(b) DFIG total loss (W) against flux linkage (V.S)

(a) (b)

Figure 9. A plot of (a) dq-axes GSC voltage (V) against time (S) and
(b) dq-axes RSC voltage (V) against time (S)

(a) (b)

Figure 10. A plot of (a) dq-axes rotor current (A) for Rfe = 0.75Ω and
(b) dq-axes rotor current for Rfe = 0.25Ω

Int J Appl Power Eng, Vol. 11, No. 4, December 2022: 304-318
Int J Appl Power Eng ISSN: 2252-8792  315

(a) (b)

Figure 11. The plots obtained for (a) dq-axes core current (A) for Rfe = 0.75Ω. and
(b) dq-axes core currents for Rfe = 0.25Ω

(a) (b)

Figure 12. A sharp deviation in the settling time of (a) dq-axes magnetizing current (A) for Rfe = 0.75Ω and
(b) dq-axes magnetizing currents for Rfe = 0.25Ω

(a) (b)

Figure 13. A plot of (a) speed (Rad/Sec) and torque (Nm) for Rfe = 0.75Ω and (b) for Rfe = 0.25Ω

Power loss minimization assessment of a doubly fed induction generator … (Crescent Onyebuchi Omeje)
316  ISSN: 2252-8792

(a) (b)

Figure 14. A plot of power loss and power output (kW) for (a) Rfe=0.75Ω and (b) Rfe = 0.25Ω

(a) (b)

Figure 15. A plot of stator active and reactive power for (a) Rfe = 0.75Ω and (b) Rfe = 0.25Ω

(a) (b)

Figure 16. A plot of rotor active and reactive power for (a) Rfe = 0.75Ω. and (b) Rfe = 0.25Ω

Int J Appl Power Eng, Vol. 11, No. 4, December 2022: 304-318
Int J Appl Power Eng ISSN: 2252-8792  317

6. CONCLUSION
In this paper, power loss minimization of a DFIG has been presented using a modified DFIG
equivalent circuit with multi-core resistances connected in parallel. The impact of core arrangements on
dynamic power loss inside a DFIG was analyzed. The loss minimization model of the DFIG was developed
incorporating iron, copper loss components, RL-filter losses, Frictional losses and power electronics
converter losses with a minimum flux linkage to minimize the overall total power losses. Simulation results
showed that losses are minimized when the equivalent core resistances are connected in parallel with
minimum permissible current flow. This aids in regulating the magnitude of the reactive power between the
wind turbine and the DFIG. An unconstrained optimization carried out showed that the DFIG power loss
model analyzed using the Hessian matrix is positive definite and therefore is controllable. The results
obtained during a transient disturbance indicated that at different core resistance values of R fe = 0.75Ω and
0.25Ω different efficiency values of 83.45% and 41.20% were obtained due to the variation in the core
currents and core losses. It can be concluded from the simulation results that with the reduced core resistance
and increased flux linkage more current flows through the core and more losses are dissipated with a
consequent local saturation of the core which may be unavoidable.

REFERENCES
[1] M. Edrah, K. L. Lo, and O. Anaya-Lara, “Impacts of high penetration of DFIG wind turbines on rotor angle stability of power
systems,” IEEE Trans. Sustain. Energy, vol. 6, no. 3, pp. 759–766, 2015, doi: 10.1109/TSTE.2015.2412176.
[2] S. Tohidi and M. I. Behnam, “A comprehensive review of low voltage ride through of doubly fed induction wind generators,”
Renew. Sustain. Energy Rev., vol. 57, pp. 412–419, 2016, doi: 10.1016/j.rser.2015.12.155.
[3] V. Yaramasu, B. Wu, P. C. Sen, S. Kouro, and M. Narimani, “High-power wind energy conversion systems: State-of-the-art and
emerging technologies,” Proc. IEEE, vol. 103, no. 5, pp. 740–788, May 2015, doi: 10.1109/JPROC.2014.2378692.
[4] H. Jabbari Asl and J. Yoon, “Power capture optimization of variable-speed wind turbines using an output feedback controller,”
Renew. Energy, vol. 86, pp. 517–525, 2016, doi: 10.1016/j.renene.2015.08.040.
[5] T. Dinesh and R. Rajasekaran, “Independent operation of DFIG-based WECS using resonant feedback compensators under
unbalanced grid voltage conditions,” in 2015 International Conference on Innovations in Information, Embedded and
Communication Systems (ICIIECS), Mar. 2015, vol. 30, no. 7, pp. 1–8, doi: 10.1109/ICIIECS.2015.7192996.
[6] F. M. Ebrahimi, A. Khayatiyan, and E. Farjah, “A novel optimizing power control strategy for centralized wind farm control
system,” Renew. Energy, vol. 86, pp. 399–408, 2016, doi: 10.1016/j.renene.2015.07.101.
[7] M. H. Baloch, J. Wang, and G. S. Kaloi, “Stability and nonlinear controller analysis of wind energy conversion system with
random wind speed,” Int. J. Electr. Power Energy Syst., vol. 79, pp. 75–83, 2016, doi: 10.1016/j.ijepes.2016.01.018.
[8] H. Nian, P. Cheng, and Z. Q. Zhu, “Coordinated direct power control of dfig system without phase-locked loop under unbalanced
grid voltage conditions,” IEEE Trans. Power Electron., vol. 31, no. 4, pp. 2905–2918, 2016, doi: 10.1109/TPEL.2015.2453127.
[9] M. K. Bourdoulis and A. T. Alexandridis, “Direct power control of DFIG wind systems based on nonlinear modeling and
analysis,” IEEE J. Emerg. Sel. Top. Power Electron., vol. 2, no. 4, pp. 764–775, 2014, doi: 10.1109/JESTPE.2014.2345092.
[10] P. Xiong and D. Sun, “Backstepping-Based DPC Strategy of a Wind Turbine-Driven DFIG Under Normal and Harmonic Grid
Voltage,” IEEE Trans. Power Electron., vol. 31, no. 6, pp. 4216–4225, Jun. 2016, doi: 10.1109/TPEL.2015.2477442.
[11] X. Wang, D. Sun, and Z. Q. Zhu, “Resonant-Based Backstepping Direct Power Control Strategy for DFIG under Both Balanced
and Unbalanced Grid Conditions,” IEEE Trans. Ind. Appl., vol. 53, no. 5, pp. 4821–4830, 2017, doi: 10.1109/TIA.2017.2700280.
[12] M. El Azzaoui, H. Mahmoudi, and C. Ed-Dahmani, “Backstepping control of a Doubly Fed Induction Generator integrated to
wind power system,” Proc. 2016 Int. Conf. Electr. Inf. Technol. ICEIT 2016, pp. 306–311, 2016, doi:
10.1109/EITech.2016.7519611.
[13] B. Bossoufi, M. Karim, A. Lagrioui, M. Taoussi, and A. Derouich, “Observer backstepping control of DFIG-Generators for wind
turbines variable-speed: FPGA-based implementation,” Renew. Energy, vol. 81, pp. 903–917, Sep. 2015, doi:
10.1016/j.renene.2015.04.013.
[14] J. Koupeny, M. Siebrecht, S. Lücke, and A. Mertens, “Observer-based online parameter estimation of doubly fed induction
generators based on the gradient descent method,” Int. ETG Congr. 2015; Die Energiewende - Blueprints New Energy Age, 2015.
[15] Y. T. Weng and Y. Y. Hsu, “Sliding mode regulator for maximum power tracking and copper loss minimisation of a doubly fed
induction generator,” IET Renew. Power Gener., vol. 9, no. 4, pp. 297–305, 2015, doi: 10.1049/iet-rpg.2014.0125.
[16] I. K. Amin and M. N. Uddin, “Nonlinear control operation of DFIG based WECS with stability analysis,” 2017 IEEE Ind. Appl.
Soc. Annu. Meet. IAS 2017, vol. 2017-January, pp. 1–8, 2017, doi: 10.1109/IAS.2017.8101789.
[17] P. Kou, D. Liang, F. Gao, and L. Gao, “Coordinated Predictive Control of DFIG-Based Wind-Battery Hybrid Systems: Using
Non-Gaussian Wind Power Predictive Distributions,” IEEE Trans. Energy Convers., vol. 30, no. 2, pp. 681–695, 2015, doi:
10.1109/TEC.2015.2390912.
[18] A. Yousefi-Talouki, E. Pouresmaeil, and B. N. Jørgensen, “Active and reactive power ripple minimization in direct power control
of matrix converter-fed DFIG,” Int. J. Electr. Power Energy Syst., vol. 63, pp. 600–608, 2014, doi: 10.1016/j.ijepes.2014.06.041.
[19] P. K. Gayen, D. Chatterjee, and S. K. Goswami, “Stator side active and reactive power control with improved rotor position and
speed estimator of a grid connected DFIG (doubly-fed induction generator),” Energy, vol. 89, pp. 461–472, 2015, doi:
10.1016/j.energy.2015.05.111.
[20] K. E. Okedu, “Augmentation of DFIG and PMSG Wind Turbines Transient Performance Using Different Fault Current Limiters,”
Energies, vol. 15, no. 13, 2022, doi: 10.3390/en15134817.
[21] K. E. Okedu, “Enhancing the performance of DFIG variable speed wind turbine using a parallel integrated capacitor and modified
modulated braking resistor,” IET Gener. Transm. Distrib., vol. 13, no. 15, pp. 3378–3387, 2019, doi: 10.1049/iet-gtd.2019.0206.
[22] K. E. Okedu, “Improving the transient performance of DFIG wind turbine using pitch angle controller low pass filter timing and
network side connected damper circuitry,” IET Renew. Power Gener., vol. 14, no. 7, pp. 1219–1227, 2020, doi: 10.1049/iet-
rpg.2019.1124.

Power loss minimization assessment of a doubly fed induction generator … (Crescent Onyebuchi Omeje)
318  ISSN: 2252-8792

[23] A. Parida and D. Chatterjee, “Model-based loss minimisation scheme for wind solar hybrid generation system using (grid-
connected) doubly fed induction generator,” IET Electr. Power Appl., vol. 10, no. 6, pp. 548–559, 2016, doi: 10.1049/iet-
epa.2016.0040.
[24] G. S. Kaloi, J. Wang, and M. H. Baloch, “Active and reactive power control of the doubly fed induction generator based on wind
energy conversion system,” Energy Reports, vol. 2, pp. 194–200, 2016, doi: 10.1016/j.egyr.2016.08.001.

BIOGRAPHIES OF AUTHORS

Crescent Onyebuchi Omeje is a Senior Lecturer in the Department of Electrical

Electronic Engineering, University of Port Harcourt, Port Harcourt River State Nigeria. He
graduated from the University of Nigeria, Nsukka in the Department of Electrical Engineering.
He has published widely in local and international journals. His research work focuses on power
electronics, new energy conversion system, multilevel inverter applications, Electric motor
drives and Power systems modeling and simulation. He is a member of the Institute of
Electrical/Electronic Engineering (IEEE), a member of the Nigerian Society of Engineers (NSE)
and he is also a registered member of the Council for the Regulation of Engineering in Nigeria
(COREN). He can be contacted at email: [email protected].

Damian Benneth Nnadi is presently a Professor in the Department of Electrical

Engineering UNN. His university education was at Enugu State University of Science and
Technology (ESUT), Enugu from 1993 to 1999, there, he obtained a B.Eng.
(Electrical/Electronic). He got his M.Eng. degree also from Enugu State University of Science
and Technology (ESUT) from 2002 to 2004. He has also obtained a Ph. D. in Power Electronics
option from Electrical Engineering, University of Nigeria, Nsukka in 2014. He is a member of
the Nigerian Society of Engineers (NSE). He is the current financial secretary of Nsukka
Chapter of the Nigerian Institute of Electrical and Electronic Engineering (NIEEE) and he is
also a registered member of the Council for the Regulation of Engineering in Nigeria (COREN).
He is a member of the Industrial/Power Electronics and New/Renewable Energy Research
Group, Power system/High Voltage Research Group. He can be contacted at email:
[email protected].

Int J Appl Power Eng, Vol. 11, No. 4, December 2022: 304-318