energies
Article
Selection of Inertial and Power Curtailment Control Methods
for Wind Power Plants to Enhance Frequency Stability
SungHoon Lim 1 , Seung-Mook Baek 2, * and Jung-Wook Park 1, *
1
2
*
School of Electrical & Electronic Engineering, Yonsei University, Seoul 03722, Korea; sam9489@yonsei.ac.kr
Division of Electrical, Electronic & Control Engineering, Kongju National University, Cheonan 31080, Korea
Correspondence: smbaek@kongju.ac.kr (S.-M.B.); jungpark@yonsei.ac.kr (J.-W.P.)
Abstract: As renewable energy penetrates the power system, system operators are required to curtail
output power from generation units to balance the power supply and demand. However, large
curtailment from wind power plants (WPPs) may instantly cause excessive output power decrement,
causing system frequency to drop significantly before reaching its nominal value. In order to solve
this problem, this paper proposes a cooperative control framework to determine the operation of
WPPs in two control methods, which are the stepwise inertial control (SIC) method and the curtailed
control method. The proposed framework first determines the WPPs to operate in the curtailed
control method to provide the required power curtailment. Next, it determines the WPPs to operate
in the SIC method considering their releasable kinetic energy to provide an effective inertial response
and compensate for the sudden excessive output power decrement caused by other WPPs operated
in the curtailed control method. Therefore, each WPP is operated in one of two control methods to
provide required power curtailment while reducing the sudden excessive output power decrement.
To verify the effectiveness of the proposed cooperative control framework, several case studies are
carried out on the practical South Korea electric power system.
Citation: Lim, S.; Baek, S.-M.; Park,
J.-W. Selection of Inertial and Power
Curtailment Control Methods for
Keywords: frequency stability; power curtailment; stepwise inertial control; supply and demand;
wind power plant
Wind Power Plants to Enhance
Frequency Stability. Energies 2022, 15,
2630. https://doi.org/10.3390/
en15072630
1. Introduction
Academic Editors: Ziad M. Ali,
Worldwide, many countries are installing the high penetration of renewable energy,
especially wind and solar, for the transition to renewable energy sources. According to
the report from the International Renewable Energy Agency (IRENA) [1], the capacity
of renewable energy in 2020 was 291.7 GW in the United States, 100.6 GW in Canada,
55.4 GW in France, 47.4 GW in the United Kingdom, 32.9 GW in Sweden, 894.9 GW in
China, 103.5 GW in Japan, and 134.3 GW in India. In particular, the renewable energy
capacity of South Korea in 2020 was 21 GW, and the wind and photovoltaic take the largest
portion. Moreover, the South Korean government has planned to increase the renewable
energy generation rate to 20% by 2030. To do so, they are planning to install wind power
plants (WPPs) and photovoltaic up to 17.7 and 5.7 GW, respectively, until 2030.
However, many studies have reported that high renewable energy penetration may
cause several stability problems [2–6]. In terms of frequency stability, the renewable energy
penetration replaces the conventional synchronous generators (SGs) participating in various
ancillary services, resulting in various frequency stability problems. For example, when a
large disturbance occurs in the power system, conventional SGs provide power reserve and
inertial response (IR) to support the frequency stability. However, distributed generators
(DGs) normally operate on the maximum power point tracking (MPPT) control method,
which cannot provide additional frequency stability support. Therefore, the penetration of
DGs operating on this control method decreases the frequency stability supports.
For the power system with a low wind power penetration level (WPPL), the WPPs
operating in the MPPT control method have caused a minor frequency stability problem.
Omar Abdel-Rahim and Shady H.
E. Abdel Aleem
Received: 7 March 2022
Accepted: 31 March 2022
Published: 3 April 2022
Publisher’s Note: MDPI stays neutral
with regard to jurisdictional claims in
published maps and institutional affiliations.
Copyright: © 2022 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).
Energies 2022, 15, 2630. https://doi.org/10.3390/en15072630
https://www.mdpi.com/journal/energies
Energies 2022, 15, x FOR PEER REVIEW
Energies 2022, 15, 2630
2 of 15
penetration of DGs operating on this control method decreases the frequency stability
2 of 14
supports.
For the power system with a low wind power penetration level (WPPL), the WPPs
operating in the MPPT control method have caused a minor frequency stability problem.
However,asasWPPL
WPPLincreases,
increases,WPPs
WPPsoperating
operatingon
onthis
thiscontrol
controlmethod
methodare
arecausing
causingsevere
severe
However,
frequency
stability
problems
[7].
Moreover,
the
variability
and
uncertainty
of
renewable
frequency stability problems [7]. Moreover, the variability and uncertainty of renewable
energyresources
resources
are
causing
severe
problems
in the
power
supply
demand.
Thereenergy
are
causing
severe
problems
in the
power
supply
and and
demand.
Therefore,
fore,
the system
operator
may require
to operate
in different
methods
other
the
system
operator
may require
WPPs WPPs
to operate
in different
controlcontrol
methods
other than
than
the MPPT
to provide
frequency
anda curtail
the
MPPT
controlcontrol
methodmethod
to provide
frequency
stabilitystability
supportsupport
[8,9] and[8,9]
curtail
certaina
certain of
amount
power tothe
maintain
the power
[10]. methods
The control
methods
for
amount
power of
to maintain
power balance
[10].balance
The control
for WPPs
other
WPPs
other
than
the
MPPT
control
method
are
summarized
in
Figure
1.
than the MPPT control method are summarized in Figure 1.
Wind power
plants
MPPT control
method
- No frequency
stability support
Virtual inertial
control method
Frequency-based
inertial control
method
- Frequency stability support
(Inertial response)
Stepwise inertial
control method
Curtailed control
method
Proportional
curtailment
control method
- Frequency stability support
(Power reserve)
- Balancing power supply and
demand
Constant
curtailment
control method
Figure 1. Classification of WPP control methods.
Figure 1. Classification of WPP control methods.
Tosolve
solvethe
the frequency
frequency stability
researchers
have
deTo
stability problem
problemcaused
causedby
byhigh
highWPPL,
WPPL,
researchers
have
veloped
a
control
method
for
WPPs
to
provide
IR.
In
order
to
provide
IR
by
WPPs,
there
developed a control method for WPPs to provide IR. In order to provide IR by WPPs, there
aretwo
twotypes
typesofofvirtual
virtualinertial
inertialcontrol
control(VIC)
(VIC)methods,
methods,which
whichare
arethe
thefrequency-based
frequency-based
are
inertial
control
(FBIC)
method
and
the
stepwise
inertial
control
(SIC)
inertial control (FBIC) method and the stepwise inertial control (SIC) methodmethod
[11–13].[11–13].
These
These
two control
methods
thethe
IRkinetic
by the kinetic
from
the WPPs.
However,
two
control
methods
provideprovide
the IR by
energy energy
from the
WPPs.
However,
while
while
the provides
former provides
the IR
system frequency
change,
latter is indethe
former
the IR based
onbased
systemonfrequency
change, the
latter isthe
independent
of
pendent
the system
frequency
change
provides to
the
according
to Therefore,
its control
the
systemof
frequency
change
and provides
theand
IR according
itsIR
control
scheme.
scheme.
Therefore,
the SIC
method
is applied
more variously for applications.
the
SIC method
is applied
more
variously
for applications.
On
the
other
hand,
WPPs
operated
by
the
curtailed
can provide
On the other hand, WPPs operated by the curtailed
controlcontrol
methodmethod
can provide
power
power when
reserve
when a disturbance
in the
powerFor
system.
For the control
curtailed
control
reserve
a disturbance
occurs inoccurs
the power
system.
the curtailed
method,
method,
there are proportional
curtailment
control
andcurtailment
constant curtailment
conthere
are proportional
curtailment
control (PCC)
and(PCC)
constant
control (CCC)
trol (CCC)
methods
[14].curtails
The former
curtails
the output
of WPP
according
to its
methods
[14].
The former
the output
power
of WPPpower
according
to its
proportional
coefficient.
Therefore,
depending
ondepending
the wind speed,
size
of thethe
power
curtailment
proportional
coefficient.
Therefore,
on thethe
wind
speed,
size of
the power
differs.
On the
other hand,
latter
curtails
constant
output
power.output power.
curtailment
differs.
On thethe
other
hand,
the latter
curtails
constant
Moreover,
Moreover,asasthe
theWPPL
WPPLincreases,
increases,maintaining
maintainingthe
thebalance
balancebetween
betweenpower
powersupply
supply
and
anddemand
demandisisbecoming
becomingmore
moreimportant
importantthan
thanever.
ever.Moreover,
Moreover,the
theamount
amountofofrequired
required
power
powercurtailment
curtailmentincreases
increasesfor
forhigh
highWPPL.
WPPL.However,
However,excessive
excessiveoutput
outputpower
powerdecrement
decrement
instantly
switchingfrom
fromthe
theMPPT
MPPTcontrol
controlmethod
methodto
instantlyoccurs
occursfrom
fromWPPs
WPPs in
in the process of switching
tothe
thecurtailed
curtailedcontrol
controlmethod.
method.As
Asaa result,
result, this causes the
the system
system frequency
frequencydip
dipbefore
before
reaching
its
nominal
value.
In
particular,
if
this
frequency
dip
is
beyond
the
dead-band
reaching its nominal value. In particular, if this frequency dip is beyond the dead-bandofof
the
it will
cause
other
SGsSGs
to compensate
for the
loss. loss.
This paper
proposes
thegovernor,
governor,
it will
cause
other
to compensate
forpower
the power
This paper
prothe
design
of
a
cooperative
control
framework,
which
determines
each
WPP
operation
poses the design of a cooperative control framework, which determines each WPP operaintion
curtailed
controlcontrol
and SIC
methods.
Therefore,
when system
require power
in curtailed
and
SIC methods.
Therefore,
when operators
system operators
require
curtailment
to maintain
the power
the WPPs
operating
by the former
the
power curtailment
to maintain
thebalance,
power balance,
the
WPPs operating
by theprovide
former prorequired
power
curtailment,
and
other
WPPs
operating
by
the
latter
compensate
for
the
vide the required power curtailment, and other WPPs operating by the latter compensate
excessive
output power
by IR. by IR.
for the excessive
outputdecrement
power decrement
This
paper
is
organized
as
follows.
This paper is organized as follows.Section
Section22describes
describesthe
theoperation
operationof
ofthe
theWPPs,
WPPs,
including
MPPT
control,
curtailed
control,
and
SIC
methods.
In
Section
3,
the
proposed
including MPPT control, curtailed control, and SIC methods. In Section 3, the proposed
cooperative
cooperativecontrol
controlframework
frameworkimplementation
implementationisisdescribed
describedwith
withits
itstheoretical
theoreticalanalysis.
analysis.
Section 4 describes the characteristics of the practical South Korea electric power system
and verifies the effectiveness of the proposed framework with several case studies using the
DIgSILENT PowerFactory® (Version 2018, DIgSILENT GmbH, Gomaringen, Germany) [15].
Finally, conclusions are given in Section 5.
2. Wind Power Plants Operation
2.1. Characteristics of Permanent Magnet Synchronous Generator and MPPT Control M
In this paper, a type-4 wind turbine generator, which is a permanent
mag
3 of 14
chronous generator (PMSG), is considered for wind power. Generally, PMSG co
a rotor side converter (RSC), DC-link circuit with a capacitor, and grid side c
2.(GSC)
Wind Power
Plants Operation
[16]. Moreover,
the PMSG control system provides a reference signal for p
2.1.
Characteristics
of
Permanent
Magnet
Synchronous
Generator
and MPPT Control
Methodon the po
trol, RSC control, and GSC
control
methods.
Furthermore,
depending
this paper, the
a type-4
wind
turbine
generator,
which
is a permanent
magnet
syntemIncondition,
active
power
reference
(Pref
) is determined
based
on MPPT
cont
chronous
generator
(PMSG),
is
considered
for
wind
power.
Generally,
PMSG
consists
and curtailed control methods.
of a rotor side converter (RSC), DC-link circuit with a capacitor, and grid side converter
Besides the power reference determined by each control method, the me
(GSC) [16]. Moreover, the PMSG control system provides a reference signal for pitch control,
power
from
wind
source
is obtained
and calculated
asthe power system
RSC
control,
andthe
GSC
control
methods.
Furthermore,
depending on
condition, the active power reference (Pref ) is determined based on MPPT control, VIC, and
1
3
curtailed control methods.
=
P
ρπR2Vwind
Cmethod,
mec
P (λ, β) the mechanical
Besides the power reference determined by2 each control
power from the wind source is obtained and calculated as
where Pmec is the mechanical power extracted from the wind, ρ is the air density
1
2 3
CPis
(λ,the
β) wind speed, and CP is(1)
Pmec of
= 46.5
ρπRm,
Vwind
kg/m3, R is the rotor radius
Vwind
the pow
2
cient based on tip speed ratio (λ) and pitch angle (β). Normally, WPPs are operate
where Pmec is the mechanical power extracted from the wind, ρ is the air density of
MPPT control method to provide maximum power in a steady state [17]. As s
1.225 kg/m3 , R is the rotor radius of 46.5 m, Vwind is the wind speed, and CP is the power
Figure 2,based
the active
power
reference
is determined
by theWPPs
MPPT
PMPPT, wh
coefficient
on tip speed
ratio
(λ) and pitch
angle (β). Normally,
arecurve,
operated
is between
minimum
speed
limit
min) and
speed lim
speed
(ωr)control
by
the MPPT
methodthe
to provide
maximum
power
in (ω
a steady
statemaximum
[17]. As shown
inMoreover,
Figure 2, thePMPPT
activeispower
reference
calculated
as is determined by the MPPT curve, PMPPT , when
rotor speed (ω r ) is between the minimum speed limit (ω min ) and maximum speed limit
(ω max ). Moreover, PMPPT is calculated as
πρR5C
PMPPT5 (ωr ) =
PMPPT (ωr ) =
πρR CP,opt
2λ3opt
P,opt
3
3 opt
2λ
× ωr 3 = kopt × ωr 3
× ωr = k opt × ωr 3
(2)
where CP,opt and λopt are the optimal CP and λ values determined by the MPPT
pape
where
CP,optrespectively,
and λopt are theand
optimal
CPthe
andcoefficient
λ values determined
by the curve.
MPPT control
method,
kopt is
of the MPPT
In this
method, respectively, and kopt is the coefficient of the MPPT curve. In this paper, CP,opt is
set to 0.447 with β at 0 and λopt at 7.2.
set to 0.447 with β at 0 and λopt at 7.2.
Active power (pu)
Energies 2022, 15, 2630
Figure
2. MPPT
and operational
characteristics
Figure
2. MPPT
curvecurve
and operational
characteristics
of WPPs.
of WPPs.
2.2. Curtailed Control Method
2.2. Curtailed Control Method
As system operators need to maintain the power balance, they may require WPPs to
operators
toconsidering
maintain the
power
balance,Asthey
may require
operateAs
in system
the curtailed
control need
method
the wind
condition.
mentioned
operate in
the
control
method
considering
wind
condition.
previously,
there
arecurtailed
CCC and PCC
methods
to curtail
output powerthe
from
WPPs.
However, As m
power
curtailment
byare
the CCC
formerand
method
only available
at a specific
output
power
previously,
there
PCCismethods
to curtail
output
power
from WPP
ever, power curtailment by the former method is only available at a specific outpu
Energies 2022, 15, x FOR PEER REVIEW
Energies 2022, 15, 2630
4 of 15
level [14]. Therefore, the latter method is preferably applied to curtail output 4power
of 14 from
WPPs. The curtailed power using the PCC method is defined as
P (ω ) = α ×P
(ω )
cur preferably
r
cur applied
MPPT tor curtail output power from
level [14]. Therefore, the latter method is
WPPs. The curtailed power using the PCC method is defined as
(3)
where Pcur is the power reference by the PCC method and αcur is the coefficient for the
curtailed power curve. Therefore,
switches
to
Pcur (ωr )when
= αcurWPP
× PMPPT
(ωr ) from the MPPT control method
(3)
the PCC method, the output power decreases by ΔPdown as
where Pcur is the power reference by the PCC method and αcur is the coefficient for the
Pdown (WPP
ωr ) =switches
PMPPT ( ωfrom
r ) − Pcur
r)
curtailed power curve. Therefore,Δwhen
the( ωMPPT
control method to
=
(1
−
α
)
×
P
(
ω
)
the PCC method, the output power decreases by ∆Pdown
as MPPT r
cur
(4)
As shown in Figure
3, the
∆Pdown
(ωoutput
− Pcur (ωr ) from point A to B when WPP is
r ) = Ppower
r )decreased
MPPT ( ωis
(4)
switched from the MPPT control method
to
the
PCC
method.
= (1 − αcur ) × PMPPT
(ωr ) However, the output power
cannot maintain the power at point B. This is because a difference exists between the meAs shown
in Figure
3, thepower
outputreference.
power is decreased
from
to B when
WPP
chanical
power
and active
Therefore,
thepoint
rotorAspeed
is accelerated
by
is switched from the MPPT control method to the PCC method. However, the output
the swing equation as
power cannot maintain the power at point B. This is because a difference exists between
the mechanical power and active power reference.
dω Therefore, the rotor speed is accelerated
2Hωr r = Pmec − Pref
(5)
by the swing equation as
dt
dωr
2Hωr
= Pmec − Pre f
(5)
where H is the inertia constant of dt
the PMSG. As a result, the rotor speed is accelerated
where
is theBinertia
theintersection
PMSG. As a result,
is accelerated
from H
point
to C, constant
which isofthe
of thethe
Prefrotor
andspeed
Pmec. Therefore,
it from
can be conpoint
B
to
C,
which
is
the
intersection
of
the
P
and
P
.
Therefore,
it
can
be
concluded
mec
ref
cluded that when WPP switches from the MPPT control method to the curtailed control
that
when the
WPP
switches
from
the MPPT control
method
topoint
the curtailed
control
method,
method,
active
power
immediately
decreases
from
A to B and
then
reaches point
the active power immediately decreases from point A to B and then reaches point C. In
C. In other words, when WPPs are required to curtail by ΔPcur, the unwanted power decother words, when WPPs are required to curtail by ∆Pcur , the unwanted power decrement
rement of ΔPdec occurs. Moreover, ΔPdec becomes significant as WPPL and the required
of ∆Pdec occurs. Moreover, ∆Pdec becomes significant as WPPL and the required power
power curtailment
curtailment
increases.increases.
MPPT curve
Mechanical power curve
A
ΔPdown
B
ΔPcur
ΔPloss
C
Curtailed
power curve
Figure3.3.Operational
Operational
characteristics
of curtailed
the curtailed
control
method
for WPP.
Figure
characteristics
of the
control
method
for WPP.
2.3.
Control
Method
2.3.Virtual
VirtualInertial
Inertial
Control
Method
While conventional SGs provide various frequency responses, such as power reserve
While conventional SGs provide various frequency responses, such as power reserve
and IR, renewable energy-based DGs are less capable of providing these frequency reand IR, renewable energy-based DGs are less capable of providing these frequency responses. However, recent studies have developed a VIC method for WPPs to provide IR
sponses.
However,
recent
studies
have
developed
a VIC
methodWPPs
for WPPs
to provide IR
with
the releasable
kinetic
energy
stored
in the
rotating rotor.
Therefore,
can provide
with the
releasable
kinetic
storedfrequency
in the rotating
WPPs can prosimilar
IR as
conventional
SGsenergy
and support
stabilityrotor.
whenTherefore,
a large disturbance,
vide
similar
IR
as
conventional
SGs
and
support
frequency
stability
when
a
large disturbsuch as a generation trip occurs in the power system [11–13].
ance,
as a generation
occurs
in the
power
system for
[11–13].
Assuch
mentioned
previously,trip
there
are FBIC
and
SIC methods
the WPP VIC method.
The WPPs
operated bypreviously,
the FBIC method
provide
IRSIC
based
on the frequency
deviation
As mentioned
there are
FBICthe
and
methods
for the WPP
VIC method.
and
of change
of frequency
change
(RoCoF)
[13]. On
other hand,
operated
Therate
WPPs
operated
by the FBIC
method
provide
the the
IR based
on theWPPs
frequency
deviation
by
the
SIC
method
provide
the
IR
independently
from
the
frequency
but
based
on
the
and rate of change of frequency change (RoCoF) [13]. On the other hand, WPPs operated
deceleration
acceleration
stage
as shown in Figure
4. As
P0 on the
ref is increased
by the SIC and
method
provide
the[11],
IR independently
from
thePfrequency
butfrom
based
deceleration and acceleration stage [11], as shown in Figure 4. As Pref is increased from P0
to Pup by the SIC method, the right term of the swing equation of Equation (5) beco
negative. As a result, the rotor speed starts to decelerate right after Pref is increased f
P0 to Pup. Then, to recover the rotor speed to ω0, the SIC method decreases the Pref from
Energies
2022,
15,
2630
of 14
Energies 2022, 15, x FOR PEER REVIEW
5 of515
to Pdown.
totoPup
right term
termof
ofthe
theswing
swingequation
equationofofEquation
Equation
becomes
Pupby
bythe
theSIC
SIC method,
method, the right
(5)(5)
becomes
negative.As
Asa aresult,
result,the
therotor
rotorspeed
speedstarts
startstotodecelerate
decelerateright
rightafter
afterPPrefref is
is increased
increased from P0
negative.
to.PThen,
up. Then,
to recover
rotor
speedtotoωω00,,the
the SIC
SIC method
refref
from
PupPup
toPP0 up
to recover
thethe
rotor
speed
method decreases
decreasesthe
thePP
from
Pdown..
totoPdown
Figure 4. Operational characteristics of the conventional SIC method. (a) Active power and
speed plane. (b) Active power and time plane.
Figure 4. Operational
characteristics
of been
the conventional
SIC method.
Activestudies
power and
rotor improved
the SICcharacteristics
method has
developed
[11],(a)many
FigureAfter
4. Operational
of the conventional
SICin
method.
(a) Active
power have
and rotor
speed plane. (b) Active power and time plane.
SIC method
provide
power
speed
plane. (b) to
Active
poweractive
and time
plane. more effectively and improve the frequency stab
[18,19].
While
themethod
SIC method
[11] increases
power
by 0.1
without
considering
After
the SIC
has been in
developed
in [11], many
studies
havepu
improved
this
After the SIC method has been developed in [11], many studies have improved this SIC
SIC method
to provide
active power
more
effectively
and improve
the frequency
WPPL
orprovide
wind
condition,
the
SIC
method
recently
developed
instability
[19]stability
increases
the po
method
toWhile
active
powerin
more
effectively
and improve
thewithout
frequency
[18,19].
[18,19].
the
SIC
method
[11]
increases
power
by
0.1
pu
considering
the
regarding
the
prevention
of secondary
frequency
dip while
providing
effective sup
While
the
method
in [11]
power
by
0.1
pu without
considering
WPPL
orSIC
wind
condition,
theincreases
SIC method
recently
developed
in [19]
increases the
the WPPL
power or
for
frequency
nadir.
Therefore,
this
SIC
method
increases
the
power
by
ΔP
SIC as
wind
condition,
the SIC method
recentlyfrequency
developed
[19] increases
powersupport
regarding
regarding
the prevention
of secondary
dipinwhile
providingthe
effective
the
secondary
frequency
whileincreases
providing
effective
support
for frequency
forprevention
frequency of
nadir.
Therefore,
this SICdip
by
m
mΔPSIC as
ΔPSICmethod
= Pthe
(ω0 ) −by
P0the
×(power
ω
−
ω
)
nadir. Therefore, this SIC method increases
power
∆P
as
0
min
T -lim
SIC
m
ΔPSIC = [ PT -lim (ω0 ) − P0 ]×(ω0m − ωmin
)
(6)
m
where PT-lim(ω0) is the
power
atT-lim
ω0(based
limit,
∆PSIC
= [P
ω0 ) − on
P0 ] the
× (torque
ω0m − ωmin
) and m is an index
(6) depen
where PT-lim(ω0) is the power at ω0 based on the torque limit, and m is an index depending
on
the WPPLs. Note that m is decreased as WPPL increases. This is because if each W
on the
WPPLs.
m isatdecreased
increases.
if each
WPP
where
PT-lim
(ω 0 )Note
is thethat
power
ω 0 basedasonWPPL
the torque
limit,This
andismbecause
is an index
depending
SIC in high WPPL, the power decrement after frequ
provides
thesame
same
amount
ofSICΔP
in
high
WPPL,
the
power
decrement
after
frequency
provides
the
amount
of
ΔP
on the WPPLs. Note that m is decreased as WPPL increases. This is because if each WPP
nadir
(point
C’Figure
in Figure
also becomes
result,
this
may cause o
nadir arrestment
arrestment
C’ in
5) also5)
becomes
high. As ahigh.
result,As
thisamay
cause
other
provides
the same(point
amount
of
∆P
SIC in high WPPL, the power decrement after frequency
conventional
SGs
with
a
slow
frequency
response
speed
to
compensate,
causing
a
secondconventional
SGs
with
slow
frequency
response
speed
compensate,
nadir
arrestment
(point
C’ain
Figure
5) also becomes
high.
As atoresult,
this maycausing
cause a sec
ary frequency
frequency
dip.
Therefore,
ΔPSIC ΔP
is decreased
by reducing
Equation
(6)causing
when a (6) w
other
conventional
SGs
with a slow
frequency
response
speed
toincompensate,
ary
dip.
Therefore,
SIC is decreased
by m
reducing
m in
Equation
WPPL increases.
After
frequency
nadir is arrested,
Pref is decreased frommpoint
C’ to D’, (6)
secondary
frequency
dip.
Therefore,
decreased
in Equation
SIC is is
WPPL increases.
After
frequency∆P
nadir
arrested,byPreducing
ref is decreased
from point C’ to
and
enters
the
acceleration
stage
to
recover
the
rotor
speed
to
ω
0.
when
WPPL
increases.
After
frequency
nadir
is
arrested,
P
is
decreased
from
point
C’ to
ref
and enters the acceleration stage to recover the rotor speed to ω0.
D’, and enters the acceleration stage to recover the rotor speed to ω 0 .
[
]
Deceleration stage: Aʹ - Bʹ - Cʹ - Dʹ
Acceleration
stage:
Dʹ - EʹAʹ
- Aʹ
Deceleration
stage:
- Bʹ - Cʹ - Dʹ
Pup
Pup
Torque
limit
Acceleration
stage: Dʹ Mechanical
Torque
limit
power curve
Eʹ - Aʹ
C'
B'
PSIC
B'
C'
D'
Mechanical
power
curve
MPPT
curve
E'
D'
A'
E'
min
MPPT curve
,
D
,
C
,
E
PSIC
PA'
ref
0
max
Rotor speed (pu)
Figure 5. Operational characteristics of the
recent SIC method.
,
,
,
min
D
C
E
0
Rotor speed (pu)
Figure5.5.Operational
Operational
characteristics
the SIC
recent
SIC method.
Figure
characteristics
of the of
recent
method.
Pref
max
Energies2022,
2022,15,
15,2630
x FOR PEER REVIEW
Energies
6 6ofof1514
Frequency (Hz)
Active power (MW)
Proposed Cooperative
Cooperative Control
Control Framework
3.3. Proposed
Framework
As
the
WPPL
increases,
power
As the WPPL increases, power curtailment
curtailment becomes
becomesessential
essentialtotobalance
balancethe
thepower
power
supply
and
demand.
However,
as
mentioned
previously,
the
increment
on
required
supply and demand. However, as mentioned previously, the increment on required
power
in afrequency
severe frequency
dipreaching
before
power curtailment
cansignificant
cause significant
ΔPdec, resulting
curtailment
can cause
∆Pdec , resulting
in a severe
dip before
reaching
itsvalue.
nominal
value. in
AsFigure
shown6,in
Figure
6, when
WPPs are
switched
from
the
its
nominal
As shown
when
WPPs
are switched
from
the MPPT
control
MPPT
control
method
to
the
curtailed
control
method,
in
order
to
maintain
the
power
method to the curtailed control method, in order to maintain the power balance and restore
balance
and restore
frequency
f0 to nominal
), theof
output
poweris
the
frequency
from the
f 0 to
nominalfrom
frequency
(fnorm ), frequency
the output(fnorm
power
the WPPs
. Asfrequency
a result, the
frequency
falls signifof the WPPs
is instantly
decreased
further
ΔPdecthe
instantly
decreased
further
by ∆Pdec
. As aby
result,
falls
significantly
before
icantly
before
reaching
f
norm
,
and
this
problem
becomes
severe
as
WPPL
is
increased.
Morereaching fnorm , and this problem becomes severe as WPPL is increased. Moreover, suppose
over,
supposefalls
the beyond
frequency
beyondof
the
of the governor
reaching
the
frequency
thefalls
dead-band
thedead-band
governor before
reaching fbefore
norm . In that case,
f
norm
.
In
that
case,
it
will
activate
the
governor
response
from
SGs
to
compensate
for the
it will activate the governor response from SGs to compensate for the power loss using
the
power
loss
using
the
primary
frequency
reserve.
primary frequency reserve.
Figure 6. Frequency dip occurrence due to excessive power decrement during power curtailment.
Figure 6. Frequency dip occurrence due to excessive power decrement during power curtailment.
(a) Output power of WPPs operated in curtailed control method. (b) Frequency response when
(a)
Output
power ofoccurs.
WPPs operated in curtailed control method. (b) Frequency response when power
power
curtailment
curtailment occurs.
In order to solve this frequency dip problem, the proposed cooperative control frameIn order to solve this frequency dip problem, the proposed cooperative control framework determines the overall WPPs operation in the SIC and curtailed control methods to
work determines the overall WPPs operation in the SIC and curtailed control methods to
provide the required power requirement while improving the frequency dip. The main
provide the required power requirement while improving the frequency dip. The main
reason for the frequency dip occurrence during the power curtailment is the sudden sigreason for the frequency dip occurrence during the power curtailment is the sudden signifinificant output power decrement from WPPs. To solve this problem, the proposed framecant output power decrement from WPPs. To solve this problem, the proposed framework
work operates some WPPs by the SIC method during the power curtailment. Therefore,
operates some WPPs by the SIC method during the power curtailment. Therefore, they
they instantly increase the power by providing IR for a short period to compensate for the
instantly increase the power by providing IR for a short period to compensate for the power
power decrement, and, after providing IR, they decrease the power back to its initial value
decrement, and, after providing IR, they decrease the power back to its initial value (see
(see Figure 5). As a result, the excessive power decrement of ΔPdec caused by the curtailed
Figure 5). As a result, the excessive power decrement of ∆Pdec caused by the curtailed concontrol method may be compensated while improving the frequency dip without disturbtrol method may be compensated while improving the frequency dip without disturbing
ing the frequency recovery to fnorm.
the frequency recovery to fnorm .
The overall procedure to implement the proposed cooperative control framework is
The overall procedure to implement the proposed cooperative control framework is
shown in Figure 7, and the detailed operations are explained below. Note that the SIC
shown in Figure 7, and the detailed operations are explained below. Note that the SIC
method in [19] is used for the VIC method, and CCC and PCC methods are considered for
method in [19] is used for the VIC method, and CCC and PCC methods are considered for
curtailed control methods in the proposed framework.
curtailed control methods in the proposed framework.
1. Stage I—As power curtailment is required to maintain the power balance, parameters
1.
Stage I—As power curtailment is required to maintain the power balance, parameters
including the iteration number (k) and the total sum of the power curtailment from
including the iteration number (k) and the total sum of the power curtailment from
WPPs (Pcur,tot) are initialized. Then, the proposed coordination control framework beWPPs (Pcur,tot ) are initialized. Then, the proposed coordination control framework
gins. In this stage, the framework firstly assigns the WPPs to be operated by the PCC
begins. In this stage, the framework firstly assigns the WPPs to be operated by the
method to provide the required power curtailment (Pcur,req). Considering the technical
PCC method to provide the required power curtailment (Pcur,req ). Considering the
operation limit [10], αcur is assumed to be 5%. Note that WPPs are assigned to be optechnical operation limit [10], αcur is assumed to be 5%. Note that WPPs are assigned
erated by this method until Pcur,tot becomes higher than Pcur,req.
to be operated by this method until Pcur,tot becomes higher than Pcur,req .
2. Stage II—As Pcur,tot becomes larger than Pcur,req in the previous stage, the system oper2.
Stage II—As Pcur,tot becomes larger than Pcur,req in the previous stage, the system
ator needs to decrease the Pcur,tot to curtail the exact amount of Pcur,req. If WPPs curtail
operator needs to decrease the Pcur,tot to curtail the exact amount of Pcur,req . If WPPs
curtail more than Pcur,req , the frequency will not recover to fnorm but will converge to
value. Therefore WPPk is operated by the CCC method to curtail the exact amount of
insufficient power curtailment (ΔPcur,CCC). As a result, while WPP1 to WPPk-1 are operated by the PCC method with αcur of 5%, WPPk is operated by the CCC method with
ΔPcur,CCC to curtail the exact amount of Pcur,req.
3. Stage III—After determining the WPPs to be operated by the curtailment control
7 of 14
method (PCC and CCC methods), the other WPPs are determined to be operated by
the SIC method to compensate for the power decrement caused by other WPPs operated by PCC and CCC methods. To do so, the total available IR for WPPk+1 to WPPn
(ΔP
SIC,totvalue.
) is calculated
as WPPk is operated by the CCC method to curtail the exact
a lower
Therefore
Energies 2022, 15, 2630
amount of insufficient power curtailment
(∆Pcur,CCC ). As a result, while WPP1 to
n
WPPk− 1 are operated by the PCC
method
ΔPSIC,tot
= with
ΔPSIC,iαcur of 5%, WPPk is operated by
(7)the
i = k +1
CCC method with ∆Pcur,CCC to curtail the
exact amount of Pcur,req .
3.
Stage
III—After
determining
the the
WPPs
to be power
operated
by the curtailment
However,
if ΔPSIC,tot
is larger than
required
curtailment,
it will causecontrol
anmethod
(PCC
and
CCC
methods),
the
other
WPPs
are
determined
to be the
operated
other power imbalance. Therefore, ΔPSIC for each WPP is modified considering
reby the
SICcurtailment
method toas
compensate for the power decrement caused by other WPPs
quired
power
operated by PCC and CCC methods. To do so, the total available IR for WPPk+1 to
Pcur,req
m
m
WPPn (∆PSIC,tot ) isΔcalculated
PSIC _mod,i =  Pas
(8)
T -lim, i ( ω 0 ) − P0, i 
 × ( ω 0 − ω min )
Δ PSIC,tot
n
to WPP
provide
IR to compensate for the power (7)
Therefore, the other WPPs (WPP
∆PSIC,i
∆Pk+1SIC,tot
= n)∑
decrement without causing a power imbalance.i=k+1
Start
(Power curtailment required)
Calculate ΔPSIC for WPPk+1, …,
WPPn by Equation (6)
Initialize parameters
(k = 0 and ΔPcur,tot = 0)
ΔPcur,CCC =
Yes
k −1
Pcur,req −  ΔPcur,i
i =1
Calculate ΔPSIC ,tot
By Equation (7)
Is Pcur,req < Pcur,tot ?
No
k=k+1
Operate WPPk by
CCC method
Calculate ΔPSIC ,mod for
WPPk+1,…, WPPn
by Equation (8)
Operate WPPk+1, …, WPPn by
SIC method
Stage III : Determination of WPPs
(WPPk+1 to WPPn) to be operated by
SIC method
Stage II : Determination of WPPk to
be operated by CCC method
End
(Provide required power
curtailment with IR support)
WPP1 to WPPk-1: PCC method
WPPk: CCC method
WPPk+1 to WPPn: SIC method
Operate WPPk by PCC method
k
Pcur,tot =  ΔPcur,i
i =1
Stage I : Determination of WPPs
(WPP1 to WPPk-1) to be operated by
PCC method
Figure7.7.Implementation
Implementationof
of the
the proposed
proposed cooperative
Figure
cooperativecontrol
controlframework.
framework.
In summary,
amount
of power
is curtailed
from WPPs,itexcessive
However,
if ∆Pwhen
larger
than the
required
power curtailment,
will causepower
another
SIC,totaislarge
decrement
instantly
occurs
during
the
power
curtailment,
which
causes
a
significant
frepower imbalance. Therefore, ∆PSIC for each WPP is modified considering the required
quency
drop
before
the
frequency
recovers
to
f
norm
.
To
solve
this
problem
without
disturbpower curtailment as
m
∆PSIC_mod,i = [ PT-lim,i (ω0 ) − P0,i ] × (ω0m − ωmin
)
Pcur,req
∆PSIC,tot
(8)
Therefore, the other WPPs (WPPk+1 to WPPn ) provide IR to compensate for the power
decrement without causing a power imbalance.
In summary, when a large amount of power is curtailed from WPPs, excessive power
decrement instantly occurs during the power curtailment, which causes a significant
frequency drop before the frequency recovers to fnorm . To solve this problem without
disturbing the power balance, the proposed cooperative control framework operates WPPs
partially by the SIC method to provide instant frequency support. In particular, the curtailed
control method is first applied to WPPs with PCC methods, and then the CCC method
Energies 2022, 15, 2630
8 of 14
is applied to provide the exact power curtailment of Pcur,req . Moreover, the other WPPs
that are not operated by the curtailed control method are operated by the SIC method
considering the Pcur,req to compensate for the excessive power decrement.
4. Simulation Results
To verify the effectiveness of the proposed cooperative control framework, several
case studies are carried out on the practical South Korea electric power system using the
DIgSILENT PowerFactory® software to provide an effective solution for power curtailment.
4.1. Characteristics of South Korea Electric Power System
In the practical South Korea electric power system, there are about 400 SGs with a
power capacity of 145 GW. Moreover, the load demand and power supply for one day
during winter in 2020 used in the simulation are about 82.4 and 83.9 GW, respectively.
Furthermore, the load demand and the power generation considering the types of SGs are
given in Table 1 according to the provinces with six areas. Therefore, the South Korea electric
power system has regional characteristics. As given in Table 1, area 1, which includes the
capital Seoul of South Korea, has the largest load demand. However, it is observed that
the power generation in area 1 is much lower than the load demand. Therefore, power is
transmitted from the other areas through a high-voltage transmission line, such as 345 and
765 kV transmission lines. Moreover, the types of SGs and their roles are different among
areas. For example, nuclear power plants, which take charge of base load power plants,
are primarily located in areas 5 and 6. On the other hand, the coal power plants, which
take charge of the load-following power plant, are primarily located in area 4. Lastly, the
peaking power plants are practically located in areas 1 and 2.
Table 1. Load demand and power generation according to areas in winter of early 2020.
Area
No.
1
2
3
4
5
6
Area Name
Seoul/Gyeonggi
Incheon
Gangwon
Chungcheong
Jeolla
Gyeongsang
Power Generation
Load
Demand
(MW)
Nuclear
(MW)
Coal
(MW)
Combined Cycle
(MW)
Others
(MW)
Total
(MW)
26,115
7056
2615
14,096
8642
23,871
0
0
0
0
5201
11,791
0
4826
2820
16,886
1111
6786
9717
4697
0
1835
3637
3902
5214
0
1204
359
715
3242
14,931
9523
4024
19,080
10,664
25,721
For wind resources, the currently installed capacity of WPPs is only about 1000 MW
in the South Korea electric power system, which is much lower than conventional SGs.
However, the South Korean government has planned to install 17.7 GW of WPPs by 2030.
Therefore, in this paper, according to the South Korean government’s renewable energy
policy and basic plan for long-term electricity supply and demand [20], 20 WPPs shown
in Figure 8 are considered. Moreover, their capacity is given in Table 2, and the total
capacity is about 10 GW. Furthermore, since the simulation environment of the South Korea
electric power system is based on winter in early 2020, the wind speed scenario is based on
January 2020 and February 2020, as given in Table 3.
4.2. Case 1—Required Power Curtailment of 606 MW
As shown in Figure 9, because of the power imbalance, the initial center of inertia
(CoI) frequency [21] is 60.035 Hz, which is higher than fnorm of 60 Hz. Note that the CoI
frequency, fCOI , is calculated as
m
f COI =
∑ HSG,j SSG,j f SG,j
j =1
!
m
×
∑ HSG,j SSG,j
j =1
! −1
(9)
Energies 2022, 15, 2630
9 of 14
where HSG,j and SSG,j are the inertia constant and capacity of the j-th SGs, and fSG,j is the
measured frequency of the bus that is connected to the j-th SGs. However, there are many
SGs and buses in a practical large-scale power system, making it impossible to measure
Energies 2022, 15, x FOR PEER REVIEW
the frequency of every bus to obtain fCOI . Therefore, the frequency of SGs with the largest
capacity in each area is selected and measured to obtain fCOI .
9o
Figure
8. 8.
South
Korea
electric
power
system
with 20with
WPPs.
Figure
South
Korea
electric
power
system
20 WPPs.
Table
2. 2.
Hosting
capacity
of 20of
WPPs.
Table
Hosting
capacity
20 WPPs.
Capacity (MW)
WPP1
200.1
WPP11
1499.6
WPP3 WPP1 WPP
WPP
2 WPP
WPP
4
5 3
299 200.1 220.8299
167.9
299
WPP13
WPP14
WPP15
WPP12 154.1
WPP13
1499.6WPP11 878.6
WPP2
299
WPP12
119.6
1499.6
119.6
1499.6
WPP64
WPP
218.5
220.8
WPP16
WPP
1000.514
878.6
Capacity (MW)
WPP
WPP
WPP
57
WPP
6 8 WPP7WPPWPP
8 WPP
WPP
9
10 9
170.2
637.1
46
400.2
167.9
218.5
170.2
637.1
46
WPP17
WPP18
WPP19
WPP20
WPP
15
WPP278.3
16
WPP171499.6
WPP18 41.4
WPP19
1000.5
154.1 1000.5 1000.5 278.3 1499.6
WPP
400.2
WPP
41.4
Table 3. Wind speed of 20 WPPs for all cases.
Table 3. Wind speed of 20 WPPs for all cases.
Wind Speed (m/s)
Case 1
(January)
Case 2
(February)
WPP1
WPP2
WPP3
6.5
7.3
6.8
6.8
WPP11
Case 1
(January)
Case 2
(February)
7.4
6.9
Wind Speed (m/s)
WPP7
WPP8
WPP9
WPP10
WPP4 WPP5 WPP6 WPP7 WPP8 WPP9 WPP
WPP4
WPP5
WPP6
7.7
6.4
6.7
WPP1 WPP2 WPP3
8.7
8.2
8.7
6.8
Case 1
6.5 6.8 7.3 6.3 6.8 6.47.7
6.4
6.7
8.77.5 8.2 6.7 8.7
6.8
8
8.7
7.5
(January)7.2
WPPCase
WPP14
WPP15
WPP16
WPP17
WPP18
WPP19
WPP20
2WPP13
12
6.8
8
7.2
6.8
6.3
6.4
8.7
7.5
7.5
6.7
(February)
8
8.5
7.4
7.5
8.1
9
7.6
8.8
7.4
WPP11 WPP12 WPP13 WPP14 WPP15 WPP16 WPP17 WPP18 WPP19 WPP
8.6
8.2
7.9
8.2
7.1
9.1
6.7
8.4
7.4
Case 1
7.4
8
8.5
7.4
7.5
8.1
9
7.6
8.8
7.4
(January)
In
case2 1, due to wind conditions based on January, 606 MW is required for power
Case
6.9 the power
8.6 supply
8.2 and
7.9demand.
8.2 In order
7.1 to curtail
9.1 6066.7
8.4
7.4
curtailment to balance
MW, WPP
(February)
operations are determined using the proposed cooperative control framework shown in
4.2. Case 1—Required Power Curtailment of 606 MW
As shown in Figure 9, because of the power imbalance, the initial center of iner
(CoI) frequency [21] is 60.035 Hz, which is higher than fnorm of 60 Hz. Note that the C
Frequency (Hz)
Energies 2022, 15, 2630
Figure 7. As a result, WPPs (WPP1 to WPP9) are operated by the PCC method with an αcur
of 5%, and WPP10 is operated by the CCC method with a constant power curtailment of
69.8 MW. On the other hand, when power curtailment occurs from WPPs (WPP1 to
WPP10), other WPPs (WPP11 to WPP20) are operated by the SIC method. Note that, since
ΔPSIC,tot (210 MW) is much lower than Pcur,req (606 MW), ΔPSIC for WPPs (WPP11 to WPP20)
are not modified by Equation (8).
10 of 14
As shown in Figure 10, WPPs (WPP1 to WPP10) determined to be operated by the
curtailed control method are curtailed at 20 s. As a result, the imbalance of power supply
and demand is solved, and fCOI starts to decrease from 60.04 Hz to fnorm. However, while
Figure 7. As a606
result,
(WPP
WPP9an
) are
operated
theofPCC
method
with
an αcur
MW WPPs
is curtailed
from
additional
230by
MW
excessive
power
decrement
1 toWPPs,
occurs
inoperated
the powerby
system.
Therefore,
as shown
inconstant
Figure 9, fpower
COI drops
significantly of
to
of 5%, and WPP
the CCC
method
with a
curtailment
10 is
59.964
Hz hand,
before recovering
to 60
Hz. To solveoccurs
this problem,
the proposed
cooperative
69.8 MW. On the
other
when power
curtailment
from WPPs
(WPP1 to
WPP10 ),
control framework
operates WPP11 to WPP20 by the SIC method. Thus, they provide an IR
other WPPs (WPP
11 to WPP20 ) are operated by the SIC method. Note that, since ∆PSIC,tot
of 210 MW at 20 s to compensate for the power decrement. As a result, it clearly shows
(210 MW) is much lower than Pcur,req (606 MW), ∆PSIC for WPPs (WPP11 to WPP20 ) are not
that the fCOI dip is significantly improved to 59.983 Hz (see the dashed red line in Figure
modified by Equation
9). Table 4 (8).
summarizes the operation of 20 WPPs during the power curtailment.
Figure 9. Results of center of inertia frequency for case 1.
Figure 9. Results of center of inertia frequency for case 1.
As shown in Figure 10, WPPs (WPP1 to WPP10 ) determined to be operated by the
curtailed control method are curtailed at 20 s. As a result, the imbalance of power supply
and demand is solved, and fCOI starts to decrease from 60.04 Hz to fnorm . However, while
606 MW is curtailed from WPPs, an additional 230 MW of excessive power decrement
occurs in the power system. Therefore, as shown in Figure 9, fCOI drops significantly to
59.964 Hz before recovering to 60 Hz. To solve this problem, the proposed cooperative
control framework operates WPP11 to WPP20 by the SIC method. Thus, they provide an IR
of 210 MW at 20 s to compensate for the power decrement. As a result, it clearly shows
Energies 2022, 15, x FOR PEER REVIEW
11 of 15
that the fCOI dip is significantly improved to 59.983 Hz (see the dashed red line in Figure
9).
Table 4 summarizes the operation of 20 WPPs during the power curtailment.
Figure 10. Results of output power from WPPs for case 1. (a) WPP1, WPP5, WPP6, WPP9, WPP12,
Figure 10. Results of output power from WPPs for case 1. (a) WPP1 , WPP5 , WPP6 , WPP9 , WPP12 ,
WPP15, WPP20. (b) WPP2, WPP3, WPP4, WPP7, WPP8, WPP10, WPP18. (c) WPP11, WPP14, WPP16. (d)
WPP
, WPP20, . WPP
(b) WPP
WPP3output
, WPP4 ,power
WPP7 ,ofWPP
. (c) WPPcontrol
.
2 , Total
8 , WPP
10 , WPP
11 , WPPmethod.
14 , WPP16
WPP15
13, WPP17
19. (e)
WPPs
operated
by18curtailed
(f)
(d)
WPP
,
WPP
,
WPP
.
(e)
Total
output
power
of
WPPs
operated
by
curtailed
control
method.
13
17
19
Total output power of WPPs operated by SIC method.
(f) Total output power of WPPs operated by SIC method.
Table 4. Summary of WPPs operation and numerical results for case 1.
WPP No.
Control
method
P0
(MW)
ΔPcur
(MW)
ΔPdec
(MW)
ΔPSIC
(MW)
Energies 2022, 15, 2630
11 of 14
Table 4. Summary of WPPs operation and numerical results for case 1.
WPP No.
Control
Method
P0
(MW)
∆Pcur
(MW)
∆Pdec
(MW)
∆PSIC
(MW)
WPP1
PCC
44.2
30.6
11.5
-
WPP2
PCC
93.4
64.3
24.4
-
WPP3
PCC
75.5
52.2
19.7
-
WPP4
PCC
81
55.7
21.2
-
WPP5
PCC
35.4
24.5
9.2
-
WPP6
PCC
52.8
36.5
13.7
-
WPP7
PCC
90
61.7
23.6
-
WPP8
PCC
282.1
193.6
73.9
-
WPP9
PCC
24.3
16.7
6.4
-
WPP10
CCC
101.1
69.8
26.4
-
WPP11
SIC
488.1
-
-
30.8
WPP12
SIC
49.2
-
-
3
WPP13
SIC
739.6
-
-
43.9
WPP14
SIC
286
-
-
17.5
WPP15
SIC
52.2
-
-
3.5
WPP16
SIC
427
-
-
28.4
WPP17
SIC
587.4
-
-
29.9
WPP18
SIC
98.1
-
-
5.1
WPP19
SIC
820.7
-
-
47.4
WPP20
SIC
13.5
-
-
0.9
Total
-
4441.6
605.6
230
210.4
4.3. Case 2—Required Power Curtailment of 337 MW
In case 2, the imbalance between power supply and demand raises the fCOI to 60.015 Hz.
Note that the fCOI exceeds fnorm by a smaller amount than that of case 1. This is because
the total power generated from the overall WPPs is smaller than case 1 due to the wind
conditions. In order to balance the power supply and demand, 337 MW is required for
power curtailment. To solve this problem, the proposed cooperative control framework
is applied to determine the 20 WPPs operation. In general, the power system is operated
by each area rather than operating the entire system as one large area. Therefore, the
proposed framework is applied in each area to determine the operation of WPPs for this
case study. However, since there are fewer WPPs in areas 1, 2, and 4, the operation of
WPPs in these areas is considered simultaneously. Thus, in order to curtail 337 MW, each
area is curtailed by 84.3 MW. As a result, the framework determines WPPs (WPP1 , WPP4 ,
and WPP5 ) to be operated by the curtailed control method (PCC method), which provides
power curtailment. Moreover, WPP2 , WPP6 , WPP11 , and WPP16 are operated by the CCC
method to balance the power supply and demand precisely for each area. Note that WPPs
are not operated by the PCC method for areas 5 and 6 since WPPs (WPP11 and WPP16 )
operated by the CCC method can provide the required curtailment for each area. Finally,
the remaining WPPs are operated by the SIC method to provide IR and compensate for the
instant power decrement during power curtailment.
As shown in Figure 11, fCOI is over 60 Hz during 0 s to 20 s due to a power imbalance
of 337 MW. After determining the WPPs operation using the proposed cooperative control
framework, WPPs are curtailed at 20 s to balance the power supply and demand, making
fCOI recover to fnorm , which is 60 Hz. However, as shown in Figure 12e, an additional power
are not operated by the PCC method for areas 5 and 6 since WPPs (WPP11 and WPP
operated by the CCC method can provide the required curtailment for each area. Final
the remaining WPPs are operated by the SIC method to provide IR and compensate f
the instant power decrement during power curtailment.
As shown in Figure 11, fCOI is over 60 Hz during 0 s to 20 s due to a12
power
of 14 imbalan
of 337 MW. After determining the WPPs operation using the proposed cooperative contr
framework, WPPs are curtailed at 20 s to balance the power supply and demand, maki
fCOI recover to fnorm, which is 60 Hz. However, as shown in Figure 12e, an additional pow
decrement of 428
MW occurs
during
Ascurtailment.
a result, a frequency
dipaoccurs,
decrement
of 428
MWpower
occurscurtailment.
during power
As a result,
frequency dip o
making fCOI decrease
to
59.952
Hz
before
reaching
60
Hz.
In
order
to
solve
this
problem,
to solve this pro
curs, making fCOI decrease to 59.952 Hz before reaching 60 Hz. In order
the proposed coordination
controlcoordination
framework additionally
operates
WPPs (WPP
3 , WPP
7,
lem, the proposed
control framework
additionally
operates
WPPs
(WP
WPP8 , WPP9 , WPP
,
WPP
,
WPP
,
WPP
,
WPP
,
WPP
,
WPP
,
WPP
,
and
WPP
17
18
19 17, WPP1820
, WPP9,13WPP10, 14WPP12, 15
WPP13, WPP
14, WPP
15, WPP
, )WPP19, a
WPP107, WPP812
by the SIC method
as) by
soon
power
curtailment
other WPPs.
As a other
result,WPPs. As
WPP20
theasSIC
method
as soon asoccurs
powerfrom
curtailment
occurs from
the frequency dip
is
increased
to
59.98
Hz.
Table
5
summarizes
the
operation
of
20
WPPs
result, the frequency dip is increased to 59.98 Hz. Table 5 summarizes the
operation of
during the power
curtailment.
WPPs
during the power curtailment.
Energies 2022, 15, 2630
Energies 2022, 15, x FOR PEER REVIEW
13 of 15
11. Results
of center
of inertia
frequency
for case 2.
Figure 11. ResultsFigure
of center
of inertia
frequency
for case
2.
Figure
12. Results of output power from WPPs for case 2. (a) WPP1, WPP4, WPP5, WPP6, WPP9
Figure 12. Results of output power from WPPs for case 2. (a) WPP1 , WPP4 , WPP5 , WPP6 , WPP9 ,
WPPWPP
20. (b) WPP2, WPP3, WPP7, WPP10, WPP12, WPP15, WPP18. (c) WPP8, WPP11, WPP14, WPP16. (d
20 . (b) WPP2 , WPP3 , WPP7 , WPP10 , WPP12 , WPP15 , WPP18 . (c) WPP8 , WPP11 , WPP14 , WPP16 .
WPP(d)
13, WPP17, WPP19. (e) Total output power of WPPs operated by curtailed control method. (f
WPP13 , WPP17 , WPP19 . (e) Total output power of WPPs operated by curtailed control method.
Total
power
of WPPs
operated
(f)output
Total output
power
of WPPs
operatedby
bySIC
SIC method.
method.
In summary, it is clearly shown from cases 1 and 2 that a large amount of power curtailTable 5. Summary of WPPs operation and numerical results for case 2.
ment causes instant power decrement during a switching process from the MPPT control
method to the curtailed
control method.
a significant
Control
P0 Moreover, this
ΔPcauses
cur
ΔPdecfrequency dip
ΔPSIC
WPP
beforeNo.
reaching fnorm . To solve this problem, the proposed cooperative control framework
method
(MW)
(MW)
(MW)
(MW)
determines WPP operation in three control methods, which are the PCC, CCC, and SIC
WPP1 Therefore,PCC
50.6
13.2
methods.
the proposed framework
operates35.1
WPPs that could
provide required -
WPP2
WPP3
WPP4
CCC
SIC
PCC
122.9
89.6
55.8
49.2
38.7
54.9
14.6
9.5
-
Energies 2022, 15, 2630
13 of 14
power curtailment for power balance while also providing IR to compensate for the instant
power decrement. The results show that frequency dip during the power curtailment is
significantly improved using the solution by the proposed framework.
Table 5. Summary of WPPs operation and numerical results for case 2.
WPP No.
Control
Method
P0
(MW)
∆Pcur
(MW)
∆Pdec
(MW)
∆PSIC
(MW)
WPP1
PCC
50.6
35.1
13.2
-
WPP2
CCC
122.9
49.2
54.9
-
WPP3
SIC
89.6
-
-
9.5
WPP4
PCC
55.8
38.7
14.6
-
WPP5
PCC
33.8
23.7
8.8
-
WPP6
CCC
46.0
21.9
18.9
-
WPP7
SIC
90.0
-
-
9.7
WPP8
SIC
215.9
-
-
26.0
WPP9
SIC
15.6
-
-
1.7
WPP10
SIC
96.7
-
-
7.3
WPP11
CCC
395.8
84.3
180.3
-
WPP12
SIC
61.1
-
-
5.5
WPP13
SIC
664.0
-
-
84.4
WPP14
SIC
347.9
-
-
35.8
WPP15
SIC
68.2
-
-
7.4
WPP16
CCC
287.7
84.3
137.2
-
WPP17
SIC
607.2
-
-
58.3
WPP18
SIC
67.3
-
-
5.8
WPP19
SIC
713.8
-
-
92.9
WPP20
SIC
13.5
-
-
1.5
Total
-
4043.3
337.2
427.9
345.8
5. Conclusions
This paper proposed the new cooperative control framework between the curtailed
control and virtual inertial control (VIC) methods to minimize the instant power decrement
during the power curtailment and improve the frequency dip. To do so, this paper first
analyzed the power loss that occurs in the process of switching from the maximum power
point tracking control method to the curtailed control method, which caused a severe impact
on frequency stability during the frequency recovery. In order to solve this problem, the
proposed cooperative control framework determined the WPP operation in the proportional
curtailment control (PCC) and constant curtailment control (CCC) methods to provide
the required power curtailment. Then, it operated the rest of the WPPs by the stepwise
inertial control (SIC) method to provide an inertial response with the consideration of the
power balance.
The effectiveness of the proposed coordination framework was verified with several
case studies on the practical South Korea electric power system. The results show that the
proposed coordination framework successfully determined WPP operation in the PCC,
CCC, and SIC methods to provide the required power curtailment and compensate for the
excessive power decrement. Therefore, it is expected that the proposed framework would
provide a promising solution on power curtailment and enable the high penetration of
WPPs to the power system.
Energies 2022, 15, 2630
14 of 14
Author Contributions: The research was conducted in collaboration with all authors. S.L. wrote the
paper; S.L. performed the simulations; S.-M.B. and J.-W.P. supervised the paper. All authors have
read and agreed to the published version of the manuscript.
Funding: This work was supported in part by the National Research Foundation of Korea (NRF)
(grant number: 2020R1A3B2079407), the Ministry of Science and ICT (MSIT), Korea, and in part
by Basic Science Research Program through the NRF funded by the Ministry of Education (grant
number: 2020R1I1A3074996).
Conflicts of Interest: The authors declare no conflict of interest.
References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
Renewable Energy Statistics 2021. Available online: https://www.irena.org/publications/2021/Aug/Renewable-energystatistics-2021 (accessed on 5 October 2021).
Eto, J.H.; Berkeley, L.; Undrill, J.; Mackin, P.; Daschmans, R.; Williams, B.; Haney, B.; Hunt, R.; Ellis, J.; Illian, H.; et al. Use of
Frequency Response Metrics to Assess the Planning and Operating Requirements for Reliable Integration of Variable Renewable Generation;
Lawrence Berkeley National Laboratory (LBNL): Berkeley, CA, USA, 2010.
Kayikçi, M.; Milanovic, J.V. Dynamic Contribution of DFIG-Based Wind Plants to System Frequency Disturbances. IEEE Trans.
Power Syst. 2009, 24, 859–867. [CrossRef]
Vorobev, P.; Greenwood, D.M.; Bell, J.H.; Bialek, J.W.; Taylor, P.C.; Turitsyn, K. Deadbands, Droop, and Inertia Impact on Power
System Frequency Distribution. IEEE Trans. Power Syst. 2019, 34, 3098–3108. [CrossRef]
Kumar, G.V.; Sarojini, R.K.; Palanisamy, K.; Padmanaban, S.; Holm-Nielsen, J.B. Large Scale Renewable Energy Integration: Issues
and Solutions. Energies 2019, 12, 1996. [CrossRef]
Oyekale, J.; Petrollese, M.; Tola, V.; Cau, G. Impacts of Renewable Energy Resources on Effectiveness of Grid-Integrated Systems:
Succinct Review of Current Challenges and Potential Solution Strategies. Energies 2020, 13, 4856. [CrossRef]
Nguyen, H.T.; Member, S.; Yang, G.; Member, S.; Hejde, A. Combination of Synchronous Condenser and Synthetic Inertia for
Frequency Stability Enhancement in Low Inertia Systems. IEEE Trans. Sustain. 2019, 10, 997–1005. [CrossRef]
Yan, X.; Sun, X. Inertia and Droop Frequency Control Strategy of Doubly-Fed Induction Generator Based on Rotor Kinetic Energy
and Supercapacitor. Energies 2020, 13, 3697. [CrossRef]
Yang, D.; Li, J.; Zhang, X.; Hua, L. Frequency Support from a Variable-Speed Wind Turbine Generator Using Different Variable
Droop Characteristics. Energies 2020, 13, 4477. [CrossRef]
Cañas-Carretón, M.; Carrión, M. Generation Capacity Expansion Considering Reserve Provision by Wind Power Units. IEEE
Trans. Power Syst. 2020, 35, 4564–4573. [CrossRef]
Ullah, N.R.; Thiringer, T.; Karlsson, D. Temporary Primary Frequency Control Support by Variable Speed Wind Turbines—Potential
and Applications. IEEE Trans. Power Syst. 2008, 23, 601–612. [CrossRef]
Kang, M.; Muljadi, E.; Hur, K.; Kang, Y.C. Stable Adaptive Inertial Control of a Doubly-Fed Induction Generator. IEEE Trans.
Smart Grid 2016, 7, 2971–2979. [CrossRef]
Hu, J.; Sun, L.; Yuan, X.; Wang, S.; Chi, Y. Modeling of Type 3 Wind Turbines with df/dt Inertia Control for System Frequency
Response Study. IEEE Trans. Power Syst. 2017, 32, 2799–2809. [CrossRef]
Wang, Y.; Bayem, H.; Giralt-devant, M.; Silva, V.; Guillaud, X.; Francois, B. Methods for Assessing Available Wind Primary Power
Reserve. IEEE Trans. Sustain. Energy 2015, 6, 272–280. [CrossRef]
DIgSILENT. DIgSILENT PowerFactory 2018 User Manual; DIgSILENT: Gomaringen, Germany, 2018.
Deng, J.; Wang, J.; Li, S.; Zhang, H.; Peng, S.; Wang, T. Adaptive Damping Design of PMSG Integrated Power System with Virtual
Synchronous Generator Control. Energies 2020, 13, 2037. [CrossRef]
Zhang, X.L.; Huang, C.; Hao, S.P.; Chen, F.; Zhai, J.J. An Improved Adaptive-Torque-Gain MPPT Control for Direct-Driven PMSG
Wind Turbines Considering Wind Farm Turbulences. Energies 2016, 9, 977. [CrossRef]
Kang, M.; Kim, K.; Muljadi, E.; Park, J.W.; Kang, Y.C. Frequency Control Support of a Doubly-Fed Induction Generator Based on
the Torque Limit. IEEE Trans. Power Syst. 2016, 31, 4575–4583. [CrossRef]
Yang, D.; Kim, J.; Kang, Y.C.; Muljadi, E.; Zhang, N.; Hong, J.; Song, S.; Zheng, T. Temporary Frequency Support of a DFIG for
High Wind Power Penetration. IEEE Trans. Power Syst. 2018, 33, 3428–3437. [CrossRef]
Nam, H. Impact of Nuclear Phase-Out Policy and Energy Balance in 2029 Based on the 8th Basic Plan for Long-Term Electricity
Supply and Demand in South Korea. Renew. Sustain. Energy Rev. 2020, 122, 109723. [CrossRef]
Mujcinagic, A.; Kusljugic, M.; Nukic, E. Wind Inertial Response Based on the Center of Inertia Frequency of a Control Area.
Energies 2020, 13, 6177. [CrossRef]