Content uploaded by Tomislav Baškarad
Author content
All content in this area was uploaded by Tomislav Baškarad on Nov 19, 2018
Content may be subject to copyright.
Analysis of the Wind Generation Impact on Inertial
and Primary Frequency Response of the Croatian
Electric Power System
Tomislav Baškarad, Igor Kuzle, Josip Đaković and Perica Ilak
Department of Energy and Power Systems
University of Zagreb Faculty of Electrical Engineering and Computing
Zagreb, Croatia
tomislav.baskarad@fer.hr, igor.kuzle@fer.hr, josip.djakovic@fer.hr, perica.ilak@fer.hr
Abstract—This research examines the impact of wind
penetration in the Croatian power systems on system frequency
stability. The main focus of the research is the simulation of
system frequency response on the mathematical model of
Croatian power system. With these models the possibility of
providing wind inertial response and primary frequency
response will be examined. Characteristics of frequency
response such as the value of frequency nadir, the steady-state
frequency deviation and the rate of change of frequency will be
compared for all investigated cases. This paper uses low-order
system frequency response models to simulate power system
frequency changes. Also, most of the block parameters are
provided and are accurately calibrated to model “real life”
processes. The models are developed and implemented in the
MATLAB/Simulink and tested on “real life” case in Croatian
power system.
Keywords—frequency response, inertial response, primary
frequency regulation, Croatian power system, wind generation
I. INTRODUCTION
A mismatch between generation and demand causes
frequency deviations, so in a power system the balance of
production and consumption must always be satisfied. The
frequency is a common factor for the entire system, therefore
a power imbalance at one point is reflected throughout the
system by a change in frequency. Some generators therefore
provide active power to address imbalance of generation and
production [1]. A speed-droop governor of each generating
unit provides the primary speed control function, while in
absence of the speed-droop governor system frequency
response to a power imbalance depend on the inertia and
damping constant. When large imbalances occur between
generation and demand, the inertia of the rotating masses of
synchronous generators slows down the frequency change
thus helping the power system to achieve better frequency
response [2]. However, increase in number of wind power
plants that utilize variable speed generators that substitute
conventional synchronous generators has led to a decrease in
system inertia [3]. This decreases system stability in response
to disturbances that causes power swings.
This issue has been tackled by many research studies with
abundant literature, from which the following stands out. The
work presented in [4] evaluates the impact of wind generation
providing primary frequency response and synthetic inertial
response on a large interconnection. In [5], an assessment of
needed frequency response reserve is made on the case of
Great Britain transmission grid for different loads and wind
power penetration. It was concluded if the wind turbines do
not contribute to the overall system inertia, the substantial
additional reserves for primary frequency control are required
in case of increased wind generation. The project "Future
system inertia" in [6] was conducted by the Nordic Analysis
Group in order to observe power system behaviour related to
the system inertia and to determine what the current level of
kinetic energy is in the power system, and expected levels of
kinetic energy in the future. The researchers investigated the
different future scenarios where a large part of conventional
power plants has been displaced with power plants with a low
inertia constant and concluded it is possible that the kinetic
energy in the Nordic synchronous area can become lower
than 100 GWs. Similar project [7] has been carried out for the
small isolated Gran Canaria power system, the island in Spain
with most wind power capacity. The obtained results show
that using implemented inertia support techniques, the value
of frequency nadir can be also improved and thus the overall
system frequency stability.
Taking into account the achievements of previous research,
this paper presents research under which the mathematical
model for the research of Croatian power system stability is
developed. Main focus is on inertial response and primary
frequency control, and the impact of increased wind
penetration on it. Mathematical model is based on static and
dynamic parameters identified in the Croatian electric power
system (EPS).
This paper is organized as follows: In Section II, data
about the Croatian power system specifics are provided;
definition of frequency response and provision of primary
frequency regulation is given in Section III; the system
modelling is presented in Section IV; in Section V, the
simulation results are presented; the conclusions are given in
Section VI.
II. SPECIFICS OF CROATIAN ELECTRIC POWER SYSTEM
Croatian EPS consists of a transmission and distribution
network, the generation facilities and plants and electricity
customers. For safe and high-quality electricity supply and
electricity exchange, Croatia’s EPS is connected to the EPS’s
of neighbouring countries and systems of other ENTSO-E
members, which together constitute the synchronous network
of continental Europe. Total installed generating capacity in
2017 is 4777 MW (Table 1).
TABLE 1. Installed power capacity with reference to primary resources [8]
Primary source
Installed
capacity (MW)
Waste
6
Biomass
36
Other renewable
39
Solar
51
Hydro-pumped storage
281
Fossil hard coal
325
Hydro run-of-river and
pondage
421
Wind onshore
537
Fossil gas
743
Fossil oil
950
Hydro reservoir
1388
Total
4777
Power generation consists of 23 hydro power plants with total
installed capacity of 2090 MW (43.75%), 8 thermal power
plants with total installed capacity of 2018 MW (42.24%), 18
wind power plants with total installed capacity of 537 MW
(11.43%). The largest power plant is hydro power plant
Zakučac with four units of the total power of 522 MW, two
units of 135 MW, one of 144 MW, and one of 108 MW. The
maximum and the minimum system load in 2017 were 3079
MW and 1305 MW, respectively [9].
III. FREQUENCY RESPONSE OF EPS
The certain industry processes require stable frequency
and its constant value for maintaining the speeds of used
rotating machines. The frequency response of power systems
is divided into three phases based on the different response
time and this paper analyses first two phases, the inertial
response phase and primary frequency response phase. When
a disturbance occurs in a system, synchronous machines will
release or absorb kinetic energy to the network to oppose the
change of frequency [10]. This phase is called inertial
response, and it is inherent in the system due to rotating
characteristic of conventional power plants and loads as
motors, pumps etc [11].
Initial frequency response is a result of synchronous
machines inertia which is immediately followed by primary
frequency regulation. Goal of primary frequency regulation
is to balance production and consumption within 15-20
seconds following a disturbance and to stabilize frequency on
some value other than nominal [12]. Turbines regulators
increase or decrease the flow of working fluid through the
turbine and thus increase or decrease power plant power
output based on frequency change. Power system dynamic
response depends on power plant type, power system
regulation energy and inertia and type of connected
consumers [13]. The dynamic behaviour of a synchronous
turbine-generator can be described using the motion equation
(the swing equation [1]) of a rotating mass (1):
dωr
dt=1
2Hsys
(Pm-Pe) (1)
where ωr is the nominal rotor speed [p.u.], Pm is the
aggregated mechanical power of turbine-generators in the
EPS [p.u.], Pe is the aggregated electrical power of generators
in the EPS [p.u.], and Hsys is the system inertia constant
[MWs/MVA]. Equation (1) shows that an imbalance between
the mechanical and electrical power of a turbine-generator
results in frequency derivative. Including the frequency
dependence of load in (1) leads to:
dωr
dt=1
2Hsys
[(Pm-Pe) – DΔωr] (2)
where ∆ωr is the rotor speed deviation [p.u.], D is the load-
damping constant and represents sensitivity of power system
loads to system frequency [14]. The average inertia constant
for a power system Hsys is determined by the combined inertia
of all rotating synchronous generators connected to the
system and can be calculated according to:
Hsys=∑HiSi
n
i=1
∑Si
n
i=1
(3)
where Hi and Si are the inertia constant and the nominal power
of generator i [2]. The n is the number of rotating
synchronous generators connected to the system.
There are three important factors (Fig. 1) to consider when
analysing the response of the system to a frequency
disturbance; the rate of change of frequency (df/dt), the steady
state frequency deviation (Δfss) and the value of frequency
nadir (fmin). Fig. 1 shows graphical representation of
frequency response indicators. The rate of change of
frequency (ROCOF) will depend on the average system
inertia, imbalance between generation and demand and load-
damping factor [15]; the steady state frequency deviation is
defined as the difference between the nominal frequency and
its steady-state value and it will be affected by the equivalent
droop characteristic of the generators in the system and total
amount of primary reserves delivered during disturbance
[16]; the value of the frequency nadir is defined as the lowest
frequency during a disturbance and it is influenced by more
factors: the size of the power imbalance, the amount of the
stored energy in the rotating masses of generators and their
dynamic characteristics [17].
Figure 1. Frequency response indicators
A. Provision of primary frequency regulation
Nominal frequency value in Croatian EPS is 50 Hz. In
normal operating conditions, in interconnected operation,
permissible frequency deviation from the nominal value is
± 50 mHz. Momentary frequency deviation from the nominal
value shall not exceed ± 800 mHz, and frequency deviation
from the set value in temporary stationary state shall not
exceed ±180 mHz. Primary control shall activate if frequency
deviation from the set value exceeds ± 20 mHz. Primary
control reserve of 0% to 50% shall trip within 15 seconds,
while that of 50% to 100% shall trip within maximum
activation time. Primary frequency control in isolated system
operation shall secure that the momentary frequency value
during disturbance does not drop below 49.20 Hz [18].
IV. SYSTEM MODELLING
The basic equation on which the system model is based is
the swing equation (2). Fig. 2 shows the block diagram
representation of swing equation including the effect of load
damping [1].
Figure 2. Block diagram representation of swing equation
where ∆Pm is the change in generated power in the system
[p.u.], ∆PL is the change in the system load [p.u.], H is the
system inertia constant (in MWs/MVA), ∆ωr is the rotor
speed deviation (p.u.), s is the Laplace operator, D is the load
damping constant. The damping constant is expressed as a
percent change in load for one percent change in frequency
and in all case studies, it will be set to 1%.
In the Fig 3. model of Croatian EPS is depicted on which the
performance of the primary frequency regulation is explored.
The hydro power plants, thermal power plants and wind
power plants are represented with the blocks HPP, TPP, and
WPP, respectively.
Figure 3. Depiction of Croatian EPS model
The models of generating units with a hydraulic turbine,
steam turbine and wind turbine participating in primary
frequency regulation used in this research are from [1] and
[19] and are improved and expanded with: a neutral (dead)
zone block, a primary control band block and a power output
block (see Fig. 4-7). The block neutral zone is the scope
defined by the frequency limit within which the controller
does not respond. The primary control band block is a set
range of primary control reserves expressed through the value
of upper and lower limits of active power within which the
generating unit’s increase or decrease power output during
frequency deviation. The block power output represents
momentary power output of each hydro and thermal plant.
This paper provides the novel blocks easy to implement in
various control loops for primary frequency control. Also,
this paper provides most of the block parameters that are
accurately calibrated to model “real life” processes. In the
Fig. 4-7 novel blocks are shown for the various technologies:
hydro turbine, steam turbine, gas turbine and wind turbine.
Figure 4. The block diagram of a generating unit with a hydro turbine
Figure 5. The block diagram of a generating unit with a steam turbine
Figure 6. The block diagram of a generating unit with a gas turbine
Figure 7. The block diagram of a generating unit with a wind turbine
A requirement for a transient droop compensation block is
due to the fact when the gate is suddenly opened, the flow
does not change immediately due to water inertia and initial
change in power is the opposite to gate changes [20]. The
parameter TR is a reset time [s], and RT is a transient droop.
The water starting time, TW [s], is the time required to
accelerate the water in the penstock from standstill to the
velocity Ur at the rated turbine head Hr [21]..
TW = LQr
gAHr
(4)
where L is a length of conduit [m], A is a pipe area [m2], g is
an acceleration due to gravity [m/s2], Qr is a rated flow [m3/s].
In the turbine transfer function of the generating unit with a
steam turbine (Fig. 5), time constant TCH [s] of the control
valve modulates the response of steam flow to a change in
opening of control valve. The reheater time constant TRH [s]
is present due to substantial amount of steam that is held in
the reheater. FHP is the fraction of the power, extracted from
the high-pressure turbine. The time constant of the governor
Tg [s] is used to represent the delay of the governor i.e. the
time required to change the valve/gate position from the
moment of detection frequency deviation from the nominal
value. The droop R is used to ensure equitable load sharing
between generating units. The amount of additional power
that can be used during primary frequency regulation is
limited by the upper level set in the primary control band
block. That is a head room and defined as the difference
between the rated power Pr and generation power P. In the
block diagram of the generating unit with a gas turbine, TVP,
TFS, TCD, are time constants related to valve positioning, fuel
dynamics and compressor discharge, respectively.
The wind power assuming a constant speed of wind is
considered here with the transfer function model from [19]
and expanded by novel block as shown in Fig. 7. Therefore,
the output power – frequency dynamics block is a transfer
function that relates change of wind turbine output power to
the power deviation set by the frequency support mechanism.
FWT, HWT, T1, and T2 are the gain, inertia and time constants
of wind turbine, respectively. The block virtual inertia
response is used to emulate inertial response by extracting
energy from the rotational mass.
The power plants which do not participate in primary
frequency regulation are modelled to change output power
proportionally regarding to the frequency deviation.
A. System parameters
The required parameters for the model are mostly obtained
from [22] and Croatian Transmission System Operator. The
number of hydro power plants and thermal power plants
participating in primary frequency regulation is 14 out of 23,
and 4 out of 8, respectively. Due to unavailability of the data
of hydro generating unit, the parameters reset time TR and
temporary droop RT are set to 6 s and 0.5, according to the
recommendation from [1]. The unavailable parameters for
thermal generating unit, the control valve time constant TCH
and reheater time constant TRH are set to value 0.25 s and 7 s,
respectively [1]. The time constant of the governor Tg is set to
0.2 s for the all units. The fraction of the power extracted from
the high-pressure turbine FHP is set to 0.35. Gas turbine
parameters are typical for these turbines and set as follows
[19]: TVP = 0.1 s, TFS = 0.4 s, TCD = 0.4 s. The gain, inertia,
droop and time constants of wind turbines are set as follows
[19]: FWT = 2.04, HWT = 1, RWT = 5% T1 = 14.11 s, T2 = 29.05
s. Other necessary parameters for hydro power plants and
thermal power plants are real parameters and given in the
tables 2-5.
TABLE 2. Parameters for the hydro power plants which do not participate in
primary frequency regulation
HPP
Electrical Power
(MW)
H
(s)
Zeleni Vir
2x0.9
2
Fužine
1x4.8
2
Lepenica
1x1.2
1.97
Ozalj
5x1.1
2
Kraljevac
2x20.8
1x12.8
2.7
2
Miljacka
3x6.4
1x4.8
2.48
2
Golubić
2x3.75
1.96
Jaruga
2x3.6
2
Varaždin
2x47
3
TABLE 3. Parameters for the hydro power plants which participate in
primary frequency regulation
HPP
Electrical Power
(MW)
H
(s)
R
(%)
Tw
(s)
Neutral
zone
(mHz)
Čakovec
2x39.9
0.72
4
1
20
Dubrava
2x39.9
1.05
4
1
20
Rijeka
2x18.4
3.22
5
1.05
50
Vinodol
3x30
2.75
5
1.52
20
Senj
3x72
2.7
4
0.82
20
Sklope
1x22.5
2.94
5
1.23
20
Gojak
3x18.5
4.45
4
3.7
20
Lešće
2x20.6
3.1
4
1.28
20
Peruća
2x30
3.6
1
1.3
50
Orlovac
2x79
2.6
4
2.9
20
Đale
2x20.4
3.2
4
2
5
Velebit
2x138
3.3
4
2.8
20
Zakučac
2x135
1x144
1x108
4.11
4.22
4.11
4
4
4
0.63
0.75
0.75
5
5
5
Dubrovnik
2x126
4.31
6
1.05
6
TABLE 4. Parameters for the thermal power plants which participate in
primary frequency regulation
TPP
Electrical
Power
(MW)
H
(s)
R
(%)
Neutral
zone
(mHz)
Rijeka
1x320
3.6
5
20
Plomin B
1x210
5.5
5
25
Zagreb TE-TO K, L
1x208
1x112
5
4
4
4
10
10
TE-TO Osijek A
1x45
5.04
8
5
TABLE 5. Parameters for the thermal power plants which do not participate
in primary frequency regulation
TPP
Electrical Power (MW)
H
(s)
Sisak
2x200
4
Plomin A
1x115
3.3
TE-TO Zagreb C
1x120
3.9
Jertovec
2x38
4
TE-TO Osijek B
2x25
5
EL-TO Zagreb
1x41
1x48
4
4
V. SIMULATIONS RESULTS
As previously mentioned, the objective of this study is to
assess the impact of wind generation on frequency response
of Croatian EPS, so three case studies will be analysed:
firstly, the frequency response without wind generation in the
system, then frequency response with wind generation but
without contributing to the system inertia and to primary
frequency regulation, and at the end, frequency response of
the system in case when wind power plants participate in
primary frequency regulation. The cases will be observed for
two scenarios: the maximum system load and minimum
system load in 2017. The system is assumed to be working in
steady state equilibrium before the power disturbance.
Croatian EPS and presented block diagrams are developed
and implemented in MATLAB/Simulink.
A. Case study 1: Frequency response of the system without
wind generation
This case study is designed and investigated to assess the
frequency response of the system when electricity power is
generated only by hydro and thermal power plants. Fig. 7
shows simulated frequency response for a step load
disturbance of 5% for the different system loads (maximum
load of 3079 MW (blue), and minimum load of 1305 MW
(red)). The calculated system inertia constant is Hsys = 3.71 s.
As can be seen in Fig. 7 if disturbance of 5% (196 MW) is
applied at t = 10 s, in both cases the system is capable to
maintain frequency above the minimum permissible value of
49.2 Hz. The frequency decline is stopped in t = 14.16 s, at
frequency value of 49.29 Hz for the first case (blue), and in
t = 13.46 s, at frequency value of 49.36 Hz for the second case
(red). The steady-state frequency value is approximately
49.84 Hz for both cases and it is above the minimum
permissible value of 49.82 Hz. The power system shows a
better frequency response in case of the minimum system
load because it is assumed that all power plants are still
operating but with reduced production, so the system inertia
constant remains the same as in the case of maximum system
load but there are more available primary control reserves.
Figure 7. Frequency response for a step load disturbance of 5% (196 MW)
at minimum system load (red) and at maximum system load (blue)
In the Fig. 8 frequency response for a step load disturbance
of 219 MW (5.58%) in case of the maximum system load is
shown. Since the frequency reaches the minimum allowable
value of 49.2 Hz, it can be concluded that maximum
allowable load disturbance for the system is 219 MW. That
percentage of the step load disturbance can be greater for the
case of the minimum system load, due to previously
mentioned reasons. The change in output power is shown in
Fig. 9. It can be seen primary control reserves equal to the
load disturbance are activated for t = 8 s.
Figure 8. Frequency response for a step load disturbance of 219 MW (5.58%)
at maximum system load
Figure 9. Change in output power of a system for a step load disturbance of
219 MW (5.58%)
B. Case study 2: Frequency response of the system in case
of power generation mix but no inertial and primary
regulation provided by wind power plants
This case study is designed and investigated to assess the
frequency response of the system in case of the power
generation mix. A simplification that has been made is that
wind speed v0 = 1 p.u. is constant during the frequency event.
Two situations can be observed:
• Situation 1: Wind generation of 537 MW displaces
the power generation from thermal units which do
not participate in the primary frequency regulation
Displaced thermal power plants are TPP Sisak, TPP Plomin
A, and TPP Osijek B. In this case, the system inertia constant
is reduced to Hsys = 3.14 s, but available primary control
reserves remain the same as in previously analysed case study
1. In this situation, it is similar behaviour for both scenarios, a
scenario of maximum and minimum system load, so only the
case of the maximum system load is analysed. Fig. 10 shows
the comparison of frequency responses for this and previously
case 1 if disturbance of 5% (196 MW) is applied at t = 10 s.
As it could be expected, a frequency nadir time is lessened,
from t = 14.16 s (blue), to t = 13.72 s (red) because it depends
on the inertial response of the power system and total system
inertia has changed for 15%. The change in total system inertia
did not have significantly impact on the value of frequency
nadir because its value is also determined by the capability of
providing primary frequency regulation that has not been
changed. The value of steady-state frequency remained the
same in both cases.
Figure 10. Comparison of frequency responses for a step load disturbance of
5% (196MW) at maximum system load
• Situation 2: Wind generation of 537 MW displaces
the power generation from thermal units which
participate in the primary frequency regulation
Displaced thermal power plants are TTP Rijeka, TPP Plomin
B, and three gas units from TPP Zagreb. In this case, the
system inertia constant is reduced to Hsys = 3.17 s, and
available primary control reserves are also reduced for 25%
due to fact that thermal units which participate in primary
frequency regulation are now out. Fig. 11-13 show the
frequency and power response of the system for this situation.
As it was expected, the frequency response of the system got
worse.
For the case of the minimum system load, Fig. 11 shows that
for the equal step load disturbance of 5% (196 MW), the
frequency nadir value is 48.8 Hz and has become lower for
0.57 Hz comparing with case 1 (blue). The frequency nadir
time has significantly increased from t = 13.47 s, to t = 15.42
s due to reduction of total system inertia and available primary
reserves. This situation is not desirable for the system since
the value of frequency nadir fell below 49.2 Hz, the
underfrequency load shedding program would be activated.
Comparing the changes in output power at minimum system
load for an equal step load disturbance od 5% (196 MW), the
differences can also be observed (Fig. 12). It can be seen in
case of wind generation in system (red), the activation of
primary control reserves takes about 10 seconds longer than
in the case without wind generation in system (blue). Also, the
initial speed of activation of primary reserves is slower.
Figure 11. Comparison of frequency responses for a step load disturbance of
5% at minimum system load for case study 1 (blue) and case study 2 (red)
Figure 12. Change in output power for the case study 1 (blue) and case study
2 (red)
For the case of the maximum system load, analysis showed
even worse primary response. Conducted analysis has shown
the frequency nadir reaches the value of 49.2 Hz for a step
load disturbance of 3.01% (118 MW) and comparing with the
case shown in Fig. 8-9, a drastic change can be noticed. The
capability of the system to maintain the frequency within the
permitted limits is reduced as much as 46% with respect to the
maximum allowable power imbalance. In case study 1, the
system could withstand a step load disturbance of 219 MW,
but in this case, it can withstand a step load disturbance of 118
MW. Even though the step load disturbance is greater 46% in
case study 1, the system in that case has much faster activation
of primary reserves.
Beside the previously mentioned reasons explaining the
differences between these two cases, another reason is that
three gas units are displaced in case study 2. Gas turbines
dynamics significantly improve system frequency response
[23]. The gas turbines have a much faster response than hydro
turbines, but also, the hydro turbines reduce output power
immediately following a disturbance due to the water inertia
which contributes to worse frequency response while gas
turbines do not possess that characteristic.
C. Case study 3: Frequency response of the system in case
of the participation wind power plants in inertial and
primary frequency control
Although, the wind power plants do not contribute to the
inertial and frequency control in the Croatian EPS, this case
study is designed and investigated to assess the possible
improvement in frequency response of the system if wind
power plants did participate in inertial and primary frequency
control. Capability of variable speed wind turbines to support
and enhance frequency response can be achieved by adding a
supplementary control loop that is sensitive to changes in
system frequency [24]. As in previous case, and here is
assumed that wind speed v0 = 1 p.u. is constant during the
frequency event. Unlike the previously examined case, in this
case is analysed only the situation when wind generation of
537 MW displaces the power generation from thermal units
which participate in the primary frequency regulation.
However, it is necessary to leave the head room for the wind
turbine [25], leaving a margin for power increase, so the
power from wind generation is set to 483 MW, leaving 10%
of nominal power for additional power. The percent of
deloading is to be decided based on the allowable maximum
limit of wind rotor speed. Fig. 13 shows a comparison of
frequency response for a step load of 5% for all three
analysed cases; the case study 1 - without wind generation in
the system (green); the case study 2 - with wind generation in
the system but no inertial and primary regulation provided by
them (red); the case study 3 – with wind power plants
participation in inertial and primary frequency control (blue).
The comparison is conducted for case of the minimum system
load. A noticeable improvement in the frequency response
can be seen for the case in which wind power plants
participate in inertial and frequency control. The frequency
nadir value is 49.12 Hz (blue) and has become higher for 0.32
Hz comparing to the case in which wind power plants do not
participate in inertial and frequency control (red). It is worth
noting that maximum allowable step load disturbance for the
case without WPP participation in inertial and frequency
control is 130 MW, and in the case with WPP participation in
inertial and frequency control, is greater for about 42% (184
MW). However, the case with the best frequency response is
the base case (green) without any displacement of
conventional power plants.
Figure 13. Frequency response for the all three-analysed: case study 1
(green), case study 2 (red), case study 3 (blue)
Fig. 14 shows a comparison of the changes in output power of
all three analysed cases. A green one indicates the case study
1, a red one indicates the case study 2, and a blue one indicates
the case study 3. As it was expected, the worst case is when
WPP do not participate in inertia and frequency control (red).
It can be seen the change in output power for the case in which
WPP participate in inertia and frequency control (blue), is
comparable during the first few seconds to the base case in
which electricity power is generated only by hydro and
thermal power plants (green). This is due to fast response of
wind turbine generators to adjust the active power to the pre-
defined values [26]. After the first 5 seconds, the green trace
shows a better response due to more available primary control
reserves from the conventional units that unlike the WPP
reserves, are not limited to 10%.
Figure 14. Comparison of change in output power of all three cases; case
study 1 (green), case study 2 (red), case study 3 (blue)
For convenience, a summary of all the obtained results is
shown in the tables 6-7.
TABLE 6. Obtained results for all three cases at the minimum system load
Minimum
system load
1305 MW
Case Study 1
Case Study 2
Case Study 3
Step load
disturbance
5%
fnadir = 49.36 Hz
fnadir = 48.8 Hz
fnadir = 49.12 Hz
Maximum
step load
disturbance
Pmax = 245 MW
(6.25%)
Pmax = 130 MW
(3.31%)
Pmax = 184 MW
(4.69%)
TABLE 7. Obtained results for all three cases at the maximum system load
Maximum
system load
3079 MW
Case Study 1
Case Study 2
Case Study 3
Step load
disturbance
5%
fnadir = 49.29 Hz
fnadir = 48.68 Hz
fnadir = 49.03 Hz
Maximum
step load
disturbance
Pmax = 219 MW
(5.58%)
Pmax = 118 MW
(3.01%)
Pmax = 178 MW
(4.54%)
VI. CONCLUSION
In this paper, the frequency response of Croatian power
system following a step loss of generation is analysed. Based
on the transfer functions of the generating units and identified
power plants parameters in the Croatian power system, the
mathematical model for the research frequency stability is
developed in MATLAB/Simulink with the novel blocks with
accurately determined parameters. Developed blocks are easy
to implement in various control loops for primary frequency
control. The investigated case studies included different
situations in the system regarding to wind power generation
and whether or not WPP participate in inertial and frequency
control. Several simulation scenarios are performed based on
the real state of power system from 2017. The obtained
results were analysed regarding to frequency specifics and
provisions of primary frequency control. The one scenario
shows that system would not have a major problem if WPP’s
displace thermal units which normally do not participate in
primary frequency regulation. However, the comparison
between a grid with and without wind power shows that
displacing conventional generation which take a part in
primary frequency regulation will lead to a deterioration of
the system frequency response. For that case, the results show
the capability of the system to maintain the instantaneous
frequency value above 49.2 Hz is reduced for about 46%
regarding to maximum allowable power imbalance. The
investigations of the third case provide evidence that
frequency nadir followed by generation loss at high levels of
wind generations can be well tolerated if WPP’s had inertial
and governor-like controls and results showed the
improvement of system response for 4.2% in case of
minimum system load, and for 2.0% in case of maximum
system load with respect to maximum allowable power
imbalance.
ACKNOWLEDGMENT
This work has been supported in part by the Croatian
Science Foundation under the project WINDLIPS – WIND
energy integration in Low Inertia Power System (grant No.
HRZZ-PAR-02-2017-03).
The work of the authors is a part of the H2020 project
CROSSBOW–CROSS Border management of variable
renewable energies and storage units enabling a transnational
Wholesale market (Grant No. 773430). This document has
been produced with the financial assistance of the European
Union. The contents of this document are the sole
responsibility of authors and can under no circumstances be
regarded as reflecting the position of the European Union.
REFERENCES
[1] P. S. Kundur, “Power System Stability and Control,” Library of
Congress Cataloging-in-Publication Data. p. 1179, 1994.
[2] C. Rahmann and A. Castillo, “Fast frequency response capability
of photovoltaic power plants: The necessity of new grid
requirements and definitions,” Energies, vol. 7, no. 10, pp. 6306–
6322, 2014.
[3] M. S. Seheda and O. B. Dudurych, "Issues of inertia response and
rate of change of frequency in power systems with different
penetration of variable speed wind turbines," 2016 Electric Power
Networks (EPNet), Szklarska Poreba, 2016, pp. 1-4.
[4] Y. C. Zhang, V. Gevorgian, and E. Ela, “Role of Wind Power in
Primary Frequency Response of an Interconnection Preprint,” Int.
Work. Large-Scale Integr. Wind Power Into Power Syst., no.
September, pp. 1–9, 2013.
[5] P. J. C. Vogler-Finck and W. G. Früh, “Evolution of primary
frequency control requirements in Great Britain with increasing
wind generation,” Int. J. Electr. Power Energy Syst., vol. 73, pp.
377–388, 2015.
[6] E. Ørum, M. Laasonen, and et al, “Future system inertia,” pp. 1–
58, 2015.
[7] J.S.Gonzalez, “Frequency support by variable speed wind turbines
in the power system of gran canaria.”
[8] Croatian Transmission System Operator, “CIGRE-Croatian
electric power system 2017.”
[9] Croatian Transmission System Operator, “Annual report for the
year ended 31 December 2017,” 2017.
[10] P. Tielens and D. van Hertem, “Grid Inertia and Frequency Control
in Power Systems with High Penetration of Renewables,” Status
Publ., 2012.
[11] S. Sharma, S. Huang and N. Sarma, "System Inertial Frequency
Response estimation and impact of renewable resources in
ERCOT interconnection," 2011 IEEE Power and Energy Society
General Meeting, Detroit, MI, USA, 2011, pp. 1-6.
[12] J. L. Agüero, M. C. Beroqui and F. Issouribehere, "Grid frequency
control. Secondary frequency control tuning taking into account
distributed primary frequency control," IEEE PES General
Meeting, Providence, RI, 2010, pp. 1-8.
[13] I. Kuzle, T. Tomisa and S. Tesnjak, "A mathematical model for
studying power system frequency changes," 2004 IEEE Africon.
7th Africon Conference in Africa (IEEE Cat. No.04CH37590),
Gaborone, 2004, pp. 761-764 Vol.2.
[14] P. M. Anderson and M. Mirheydar, “A low-order system
frequency response model,” IEEE Trans. Power Syst., vol. 5, no.
3, pp. 720–729, 1990.
[15] M. Shafiullah, H. Rahman, M. I. Hossain, and M. Ahsan, “The
study of dependency of power system stability on system inertia
constant for various contingencies,” in 1st International
Conference on Electrical Engineering and Information and
Communication Technology, ICEEICT 2014, 2014.
[16] F. Teng and G. Strbac, "Assessment of the Role and Value of
Frequency Response Support From Wind Plants," in IEEE
Transactions on Sustainable Energy, vol. 7, no. 2, pp. 586-595,
April 2016.
[17] Y. Sun, Z. Zhang, G. Li and J. Lin, "Review on frequency control
of power systems with wind power penetration," 2010
International Conference on Power System Technology,
Hangzhou, 2010, pp. 1-8.
[18] Croatian Transmission System Operator, “The Grid Code.”
[19] M. Krpan and I. Kuzle, "Inertial and primary frequency response
model of variable-speed wind turbines," in The Journal of
Engineering, vol. 2017, no. 13, pp. 844-848, 2017.
[20] R. A. Naghizadeh, S. Jazebi, and B. Vahidi, “Modeling hydro
power plants and tuning hydro governors as an educational
guideline,” Int. Rev. Model. Simulations, 2012.
[21] "Hydraulic turbine and turbine control models for system dynamic
studies," in IEEE Transactions on Power Systems, vol. 7, no. 1, pp.
167-179, Feb. 1992.
[22] I. Kuzle, “Identification of Dynamic Parameters Relevant for
Frequency Changes in Medium Developed Electric Power
System,” Faculty of Electrical Engineering and Computing, 2002.
[23] L. Meegahapola, "Characterisation of gas turbine dynamics during
frequency excursions in power networks," in IET Generation,
Transmission & Distribution, vol. 8, no. 10, pp. 1733-1743,
October 2014.
[24] G. C. Tarnowski, P. C. Kjar, P. E. Sorensen and J. Ostergaard,
"Variable speed wind turbines capability for temporary over-
production," 2009 IEEE Power & Energy Society General
Meeting, Calgary, AB, 2009, pp. 1-7.
[25] R. G. de Almeida and J. a P. Lopes, “Primary frequency control
participation provided by doubly fed induction wind generators,”
15th PSCC, Liege, 2005.
[26] M. Garmroodi, G. Verbič and D. J. Hill, "Frequency Support From
Wind Turbine Generators With a Time-Variable Droop
Characteristic," in IEEE Transactions on Sustainable Energy, vol.
9, no. 2, pp. 676-684, April 2018.
