ICEECA2019 - Paper - 127 Final
ICEECA2019 - Paper - 127 Final
ICEECA2019 - Paper - 127 Final
SCR has originally been used for HVDC systems [3], then
Abstract— This paper investigates the impacts of large extended to RES featuring similar architecture and control as
synchronverters (SV) integration on transient stability and HVDC devices [4, 5].
strength of electric power systems. Large Synchronverters account As a stability indicator, the strength of a power system at
for large amounts of electrical energy injected into the power grid the POI affects its ability to maintain its voltage and power
through power converters, as in PV and off shore wind power
sources. Potential stability problems may occur when large power
quality under variable generation levels of RES. Connecting
converters are connected to weak power systems (PES). An SV is RES to weaker portions of the grid may tremendously harm
an inverter that mimics a synchronous generator (SG). voltage stability and quality at the POI [6], in particular
Methodologically, the SV performances are contrasted to the transient stability. For instance, a major fault occurring in an
standard SG of like parameters in terms of system strength and interconnected PES through weak transmission lines may lead
transient stability. These criteria are quantified respectively by the to transient instability. The criteria used for transient stability
Short-circuit Ratio (SCR) and the critical clearing time (CCT). evaluation is the Critical Clearing Time. This is because, in AC
The penetration ratio of SV generation is increased to the power grids integrating RES, large disturbances are mainly
detriment of SG power. Due to their current controllers’ compensated by large conventional synchronous generators.
limitations, the SV participation in short-circuit currents is
reduced to nominal values, diminishing the network strength.
However, as electronic power converters lack inertia, their
Although SVs have similar properties as SGs and have variable increasing number diminishes the overall system inertia, which
virtual inertia, this study reveals many limitations that must be affects the dynamic and transient stability of the power grid [7].
taken into account when planning their integration into the Consequently, power system operators have set rules
network. In opposition, SGs enhance the transient stability margin requiring that RES actively take part in power regulation of the
even in weaker systems. Hence, strong buses in a PES indicated by grid, just like conventional SGs. Based on this idea, the concept
large SCR and CCT feature larger amounts of SG power of Synchronverters was proposed. The aim is to render grid-
compared to the power generated by SVs. connected inverters emulate the essential behavior of SGs,
including the droop mechanism and inertial properties [8]. As a
Keywords- synchronverter; network strength; inertia; SCR;
CCT.
result, the SV should have the ability of providing grid support
by automatically adjusting its active and reactive powers to
I. INTRODUCTION comply with frequency and voltage grid requirements [11].
Meanwhile, the inertial characteristics emulated by the SV
Electric power systems are undergoing significant changes
contribute to the total inertia of the grid, enhancing thereafter
in the generation mix and source types. Power electronic
the system strength. Hence, the SV provides a promising
converters have evolved in number and capacity, being an
technology for various applications, such as HVDC
essential equipment in renewable energy sources (RES), such
transmission [9], STATCOM [10], and wind power systems.
as wind farms, solar PV plants, storage batteries and HVDC
One important advantage of the SV is that some system
interconnexion. Although, mostly are equipped with fast-acting
controls, many stability issues may arise when large power parameters, such as inertia, can be suitably chosen to improve
the system dynamic performances. In comparison, SGs inertia
converters are connected to weak power grids [1]. System
are constant, and prime movers may have large time constants
strength is one of the important concerns in the integration of
in droop control loops.
RES.
In this paper, we investigate the transient stability of a
Technically, a weak AC system may be evaluated using
four-bus power system, considering five configurations of SV
several measures: low ratio of inductance over resistance, high
integrations. The SV parameters are adapted to the conventional
impedance, and low inertia [2]. The short-circuit ratio (SCR) at
SG parameters. Dynamic responses of the SV and the SG are
the point of interconnection (POI) has evolved as an indicator
compared under different operating conditions. The SV benefits
of system strength. It is defined as the ratio of system MVA
its operation frequency and voltage drooping mechanisms for
short-circuit capacity (Scc), to the MW rated power of the
load sharing. For each configuration, strength at POI is
interconnected device [3], which could absorb or inject power
evaluated by its SCR. Also, transient stability of the systems is
at the POI. The POI bus is considered strong if its SCR is above
assessed in terms of the CCT, for different levels of the SCR at
three. As in indicator of system strength, the
the load bus.
1
II. OVERVIEW OF THE SYNCHRONVERETR
TECHNOLOGY
In this Section, the synchronverter concept is briefly
reviewed. The synchronous generator (SG) is the dominant type
of generator in conventional electric power systems [12]. SGs
offer many advantages such as (i) supplying inertial response
out of the kinetic energy stored in the rotating masses; (ii) they
can supply and absorb reactive power; (iii) they have simple
control structures with guaranteed grid stabilizing responses.
Based on these proprieties, an inverter based on the
synchronverter controls is developed to mimic a SG [8].
A synchronverter consists of a power part, which is the same a)
as a conventional power electronic converter depicted in Fig.1a,
and an electronic part, which consists of the measurement and
the control circuits. The power part of the SV is the standard
hardware of a three-phase pulse width modulation (PWM)
inverter, with an LCL filter at the output of each phase.
Following the terminology defined in [8], the three inductors
with parameters Ls and Rs play the role of stator coils in a SG.
The input of the power part is the value of the desired
synchronous internal voltage e from the electronic part. The
three phase inverter then generates a high-frequency switching
voltage signal eabc [ea eb ec ]T which average value over a
switching period is e. The basic mathematical model of an SV
is given in Equations (1-6).
d b)
2H ( Pm Pe ) / (1)
dt Figure 1. Organization of the synchronverter.
d (2) a) the power part of SV, b) the electronic part of SV
dt
Pm Pref D p (n ) (3) in Fig.1.a. The DC-DC converter and the PV panel or any other
renewable source, presented in Fig.1.a, can be represented by
Pe Em (ia sin ib sin( 2 ) ic sin( 2 )) (4)
3 3 one equivalent constant voltage source noted VDC. It is noted
1 that there is a difference in hardware between a SV and SG:
Em ( Dq (Uref Um ) Qref Qe ) (5)
Ks some modest energy storage is required on the DC bus of the
Qe Em (ia cos ib cos( 2 ) ic cos( 2 )) (6) inverter to mimic the effect of rotor inertia.
3 3 The mechanical equation of the SV rotor is expressed by (1).
where H is the inertia constant, Pref and Qref are the given values The SV control can be divided into active loop control and
of active and reactive powers, Pe and Qe are the actual output reactive loop control as described by (3) and (5) respectively.
values of active power and reactive power, n and are the Its active loop includes active frequency droop control and
rated and actual values of electric angular velocity, is the inertial response control, which mainly realizes the function of
rotor angle, Dp and Dq are the droop coefficients of active and independent frequency response. The reactive loop consists of
reactive loops, Um and Uref are the actual and given values of reactive power-voltage droop control and end-voltage closed-
grid voltage amplitude, Em is the internal potential amplitude of loop control, which achieve automatic voltage regulation and
the SV, K is the integral coefficient. voltage amplitude control of the SV [8].
The electronic part of a three-phase SV, as shown in Fig.1.b,
includes the mathematical model (1-6) of a three-phase round-
III. SHORT -CIRCUIT R ATIO AND SYSTEM STRENGTH
rotor synchronous machine as the core. The SV controller
strategy processes the electric measurements (vabc and iabc) as The Short-circuit Ratio is an indicator of the system
well as the Pref and the Qref which purpose is to generate e, using strength, which is defined as the ratio of system MVA short-
the following expression , circuit power to the MW rating power of the interconnected
device [3]. The strength of a power system at the POI of the
e a b c E m [s in s in ( 2 ) s in ( 2 )]T (7) device is viewed as the ability of the system at the point to
3 3 maintain its voltage stability and quality. In this section, the
relationship between SCR and system strength at the POI of
The latter expression is passed through a PWM generation RES is explained.
block to generate six pulses to drive the power semiconductors For an HVDC interconnection, the SCR is used to quantify
the strength at the POI and its value determines how strong the
2
AC system to support the stable operation of HVDC converter.
Moreover, the RES can either be on grid or off grid. When
connected to the grid, RES are typically interfaced to the
distribution, far from the main grid. The points of
interconnection of these RES are generally weak, and voltage
stability problems are likely to occur [5,6]. Therefore, it is
necessary to measure the system strength at all possible
interconnection points. Like HVDC devices, the commonly
used SCR has been extended to quantify the strength at the POI
of RES. The SCR at bus i can be expressed as:
2
Sac ,i Vi 1
SCRi . (8)
Pd ,i Pd ,i Zi
where S ac , i is the short-circuit capacity of the system at bus i;
Pd ,i is the MW rated power of the RES or the load connected to
bus i; Vi and Z i are respectively the Thevenin voltage and
impedance viewed at bus i .
From the expression of the SCR, it is clear the system
Figure 2. Four areas and four SV/SG interconnected generation
strength is highly dependent on the ratio of short-circuit
capacity to the power of the interconnected device at a specific
point. The short-circuit capacity may also be expressed as: - Case 4: Areas 1, 2 and 3 are SGs. Area 4 is an SV. This
case is designed as 3SG-1SV with SV power integration ratio
S ac V0 . I cc (9)
equals 25%.
where V0 and I cc are the pre-fault bus voltage magnitude - Case 5 named 4SG: Area 1, 2, 3 and 4 are SGs designated
and fault current, respectively. SG1, SG2, SG3 and SG4 with no SV power integration.
The generation sources are chosen to have identical generator,
with voltage, power and frequency ratings given respectively by
IV. THE TESTED FOUR AREA POWER SYSTEM V = 100 kV, Sn = 200 MVA, f = 50 Hz. The nominal current of
The synchronverter control strategy is applied to the system the SV and the SG is expressed by:
described in Fig. 2. The system is built based on S 200
In n 1.15kA (10)
MATLAB/Simulink toolbox. A series of time domain 3Vn 3.100
simulations are performed to investigate the transient stability For the SV parameters, they are firstly adapted to match
of the tested system, considering five configurations. The standard SG parameters [9], [12]: a 5% governor droop (1/Dp_SV
impact of the SV integration on load point strength and on its and 1/Dp_SG), inertia constant H=4 s, and 3% voltage regulation
transient stability is assessed in terms of the CCT and SCR. gain (1/DQ_SV and 1/DQ_SG).
The proposed tests consist of four control areas designated
as Area 1, Area 2, Area 3 and Area 4, feeding a common load V. SIMULATION RESULTS
area at POI, via AC lines as in Fig. 2. Two different generation The main advantage of the SVs is to feature the structure of
sources are studied: synchronverters and synchronous classic SGs, which controls are well known. The SV operation
generators. SVs are modeled by (1)-(7), SGs of the same power is verified on the test power system shown in Fig. 2. Dynamic
rating are equipped with power frequency and voltage reactive and transient performances of SV are compared with SG of
power controls described in (3) and (5), respectively. same capacity under different faults.
The study cases are defined as follows: A. Load share of the SV vs SG
An important mechanism for SVs and SGs for even load
- Case 1: Areas 1, 2, 3 and 4 are SVs designated as SV1, sharing is to adjust the real power delivered to the grid
SV2, SV3 and SV4. This case is named 4SV with SV according to the grid frequency. In this Section, the dynamic
integration ratio equals to 100%. performances of the cases 4SV and 3SV-1SG are compared.
- Case 2: SV in Area 1 has been replaced by an The responses of these cases to a load change at t=10 s are
equivalent SG (marked as SG1 in Fig.2). Areas 2, 3 and 4 are shown in Figs 3 and 4. In Fig.3, SV1, SV2, SV3 and SV1
kept as SVs. responses of the 4SV system are presented. Fig.4 depicts the
- Case 3: Areas 1 and 2 are SGs named SG1 and SG2, 3SV-1SG responses. The active powers, the speeds, the
respectively. voltages and the reactive powers dynamics are similar.
- This case is called 2SG- 2SV with SV power These results confirm that the SV performances are nearly
integration ratio equal to 50%. identical to SG ones, for steady state and small signal
stability. The SV and the SG coupled via an AC lines as shown
3
in Fig. 2 are set in a parallel arrangement and are initially at (a)
(a) (b)
nominal frequency f0 =1p.u with power outputs, respectively,
PSV1=PSG1=0.8 p.u (Fig.3.a and Fig.4.a). Fig.3.b and Fig.4.b
show, a slow down of the speed response to an active load
increase PL=+0.8 pu. The governors increase their power
output until they reach a new common operating frequency f’.
The frequency of the SG followed the SV frequency, and the
real and reactive powers tracked their set points. For example,
for the 3SV-1SG system, the amount of load picked up by each
unit depends on the droop characteristics, such that:
PL 0.8 (c)
(c) (d)
f f ' f 0 0.01 p.u . (11)
DSG 1 DSV 2 DSV 3 DSV 4 4 * 20
The SG1, SV2, SV3 and SV4 increased their real powers output
by 0.2 p.u
PSG1 f * DP SG1 0.01* 20 0.2 p.u (12)
PSV 2 PSV 3 PSV 4 f * DP SV 0.01*20 0.2 p.u (13)
The amount of load increase corresponds to a 0.1 % drop of the
PSV D Figure 3. Responses of the 4SV system to a +0.8 p.u step in the load.
frequency. Hence, P SV , where DP-SV and DP-SG1 are a) Responses of the SVs active powers in (p.u) b) Responses of the SV
PSG1 DP SG1 speeds in (p.u) c) Responses of the SVs reactive powers in (p.u)
respectively, the static droop of SV and SG, which are equal to d) responses of the SV voltages in (p.u).
1/5% .
Due to the load increase, the local terminal voltage Vabc (a) (b)
decreased by about 0.6 % from nominal (as in Fig.3.d and (a) (b)
Fig.4.d). In response, the reactive power output of SV1 and
SG1increased by about 0.195 p.u (see Fig.3.c and Fig.4.c),
depending on the droop control value
VSG1 0.992 0.998
QSG1 0.195 p.u . (14)
Dq SG1 0.03
4
The SCR at POI is evaluated as follows: TABLE 1. CCTS , CURRENT DEFAULT ICC, ETH, ZTH AND THE SCRS AT POI
FOR DIFFERENT RATIOS OF SV INTEGRATION WITH A LINE LENGTH L=400KM.
Base Case: the system in Fig. 5 is in normal operating
condition, and governed by: Length 400km
VL Eth Z th . IL (15)
Cases CCT SCR
(s) Icc (p.u) Eth (p.u) Zth
from which IL and VL have been measured.
(p.u)
Contingency Case: a three phase short-circuit fault is applied 4SG 2.1 2.9 0.95 0.32 1.09
at the POI bus, and the fault current I cc is measured for 3SG-1SV 1.75s 2.12 0.95 0.44 0.80
different cases as summarized in Tables 1, 2, 3 and 4. The fault 2SV-2SG 0.9s 1.65 0.949 0.57 0.64
current is given by:
3SV-1SG 0.75s 1.3 0.948 0.72 0.49
E 4SV 0.6s 1.27 0.948 0.74 0.48
I cc th (16)
Zth
Then, the Thevenin voltage is computed as:
TABLE 2. CCTS , CURRENT DEFAULT ICC, ETH, ZTH AND THE SCRS AT POI
V .I FOR DIFFERENT RATIO OF SV INTEGRATION WITH A LINE LENGTH L=300KM.
Eth L cc (17)
I cc IL
Length 300km
Finally, the SCRs in Tables 1,2,3,4, are given by:
Cases CCT (s) SCR
Icc (p.u) Eth (p.u) Zth
Sac ,POI
SCRPOI (18) (p.u)
Pd ,POI 4SG 2.5 3.53 0.965 0.27 1.52
3SG-1SV 1.85 2.82 0.965 0.34 1.08
where Sac,POI is the short-circuit capacity and Pd,POI is the load
2SV-2SG 1 1.72 0.964 0.56 0.84
power at POI. As indicated in equation (9), Sac,POI is calculated
by: 3SV-1SG 0.8 1.55 0.964 0.62 0.59
4SV 0.75 1.52 0.964 0.63 0.58
Sac, POI Eth . Icc (19)
The load power Pd, POI is equal to 500MW which may also be
considered as the RES power to be injected at the POI. TABLE 3. CCTS , CURRENT DEFAULT ICC, ETH, ZTH AND THE SCRS AT POI
FOR DIFFERENT RATIO OF SV INTEGRATION WITH A LINE LENGTH L=200KM.
C. the impact of the SVs integrations on the CCT and the SCR
Length 200km
First, it can be observed from Tables 1-4 (column 2) that Cases CCT (s) SCR
the best CCT is for case 4SG. The CCT decreased with
Icc (p.u) Eth Zth
increasing ratios of SV integration, and with the decoupling (p.u) (p.u)
between generators by increasing their tie lines. Fig.6.a 4SG 2.8 4.6 0.97 0.21 1.77
illustrates the worst CCTs for the case with full SV integration 3SG-1SV 2.3 3.7 0.97 0.26 1.42
and no SG. 2SV-2SG 1.1 2.4 0.969 0.4 0.92
Second, it is confirmed that the strength of the grid at POI is 3SV-1SG 0.9 2 0.969 0.48 0.77
well reflected by its SCR and its fault current Icc. In fact, the Icc 4SV 0.8 1.90 0.969 0.51 0.73
5
Table 5. The CCTs for different inertia constant.
(a) (b) CCT(s)
Constant 0.5 2 4 10 15 20
Inertia H (s)
3SG-1SV 1.5 2.1 2.4 2.5 2.8 unstable
2SV-2SG 0.95 1 1.2 1.8 2 unstable
3SV-1SG 0.75 0.8 1 1.15 1.5 unstable
4SV 0.22 0.45 0.98 1.1 1.4 unstable
Figure 6. The CCTs and the SCRs of different configurations and line length.
In addition, thanks to its variable virtual inertia, SV has
a) CCTs. b) SCRs. improved the transient stability of the tested network. However,
the SV has some limitations highlighted along this work which
In addition, as it can be seen from Fig. 6.b and Table 4 (column consisted of:
3 and 5), with a line length L=50km, higher fault current values -Low participation in the fault current leading to low SCR and
are depicted, because of the reduction of the Thevenin consequently a reduction of the network strength;
equivalent impedance -Maximum value of the inertia constant beyond which the
system loses its stability;
Z th seen from the POI bus which is considered strong. A higher -The increase of the integration ratio of SVs power has a
fault current value reflects a strong response of the generation negative impact on network performances (by reducing the
in a PES, which produces, at the fault location, a higher current SCR and CCT) as opposed to SGs.
and a voltage drop. Nonetheless, in Table 1 with a line length
REFERENCES
L=400km, the fault current levels Icc decreased when the POI
[1] H. Urdal, r. Ierna, j. Zhu, c. Ivanov, a. Dahresobh and d. Rostom, “System
bus is considered weak, since the SCR is smaller than 2. Similar strength considerations in a converter dominated power system,” IET, renew.
observations for the CCT can be made based on data in Tables Power gen., vol. 9, no. 1, pp.10-17, jan., 2015.
1-4 and Fig.6.a, where the CCT tends to overestimate the [2] A. Gavrilovic, “AC/DC system strength as indicated by short-circuit
ratios”, International Conference on AC and DC Power Transmission”,
system strength in terms of transient stability. The lower CCT September 17-20, 1991, pp.27–32.
is observed with weaker POI bus; the higher CCT indicates [3] C. Jia, J. Wen & X. Bie, “Study of The Effect of AC System Strength on
strong POI bus. Thus, strong buses in a PES indicated by large the HVDC Startup Characteristics”, International Conference on Sustainable
SCR and CCT values feature higher power quantity from SGs Power Generation and Supply, September 8-9, 2012.
[4] D. Wu, G. Li, M. Javadi, “Assessing Impact of Renewable Energy
and less from SVs. The higher values of fault current at theses Integration on System Strength Using Site-Dependent Short-circuit Ratio”
buses tend to maintain stable operation during disturbances. IEEE Transactions on Sustainable Energy , Vol.9, July 2018, 1072 – 1080.
[5] Y. Zhang, S. H. Huang, J. Schmall, J. Conto, J. Billo and E. Rehman,
D. The effect of the SV inertia constant on the CCT “Evaluating system strength for large-scale wind plant integration”, IEEE
Power and Energy Society General Meeting, July 2014, pp. 1-5.
One important advantage of SVs over SGs is the [6] S. H. Huang, J. Schmall, J. Conto, Y. Zhang, Y. Li and J. Billo, “Voltage
possibility to vary their parameters, like the inertia constant, to stability of large-scale wind plants integrated in weak networks: an ERCOT
case study”, IEEE Power and Energy Society General Meeting, July 2015, pp.
improve power system stability. In this section, the inertia 1-5.
constant H of cases with SV integration is varied to enhance the [7] Agada Ihuoma Nkechi, Abdul Motin Howlader and Atsushi Yona
transient stability in term of CCT. The SG inertia constant H is “Integration of Photovoltaic Energy to the Grid, Using the Virtual Synchronous
fixed to 4s, while the SV inertia constant is increased up to 20s. Generator Control Technique”, Journal of Energy and Power Engineering 12
(2018) 329-339.
The results summarized in Table 5, confirm the improvement [8] Q.-C. Zhong, and G.Weiss, "Synchronverters: Inverters that mimic
of transient stability with increased inertia, limited by a synchronous generators," IEEE Trans. Ind. Electron., Apr. 2011, vol. 58, no. 4,
threshold value, beyond which the system becomes instable. pp. 1259–1267.
For the system strength, the inertia constant has no significant [9] R. Aouini, B. Marinescu, K. Ben Kilani, M. Elleuch, Synchronverter-Based
Emulation and Control of HVDC Transmission, IEEE Trans. on Power
effect on the SCR, since it doesn’t affect the fault current Icc. Systems, No. 1, 278-286, January 2016.
This is another limitation of SV performance. [10] R. Aouini, I. Nevzi, K. Ben-Kilani, M. Elleuch, “Exploitation of
Synchronverter Control to Improve the Integration of Renewable Sources to the
CONCLUSION Grid control, J. Electrical Systems 12-3 (2016): 515-528.
[11] P-L. Nguyen, Q-C. Zhong, F. Blaabjerg, and J-M. Guerrero, "
This paper dealt with a comparative analysis of a SV/SG in Synchronverter-based Operation of STATCOM to Mimic Synchronous
four-areas network. The performance criteria used are CCT and Condensers", 2012 7th IEEE Conference on Industrial Electronics and
SCR. With appropriate controls, it is shown that SVs are able to Applications (ICIEA).
recover many performances of SGs in terms of frequency and [12] P. Kundur, "Power system stability and control", Mc Graw-HillInc, 1994.
[13] M. Callavik, A. Blomberg, J. Häfner, B. Jacobson, The Hybrid HVDC
voltage droop control. Breaker—An Innovation Breakthrough Enabling Reliable HVDC Grids,
Technical Paper, ABB Grid Systems, 2012 (November).