Impact of VSC Control Strategies and Incorporation of Synchronous Condensers On Distance Protection Under Unbalanced Faults

This articleDownloaded
has been accepted

fromforhttp://iranpaper.ir
publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TIE.2018.2835389, IEEE
http://www.itrans24.com/landing1.html
Transactions on Industrial Electronics
IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS
Impact of VSC Control Strategies and

Incorporation of Synchronous Condensers on
Distance Protection under Unbalanced Faults
Jundi Jia, Student Member, IEEE, Guangya Yang, Senior Member, IEEE,
Arne Hejde Nielsen, Senior Member, IEEE, and Peter Rønne-Hansen
Abstract—The short circuit response of a voltage source system affecting its control variables and hence its output [2].
converter (VSC) under grid unbalanced faults mainly de- If the control system is not designed properly, undesirable
pends on the design of its control system. Due to the performances such as output voltage and current distortions,
limited semiconductor overload capability, the short circuit
current contributed by a VSC should be restricted within DC-link voltage oscillations and output power oscillations can
the limit for each phase. This might bring up challenges to be observed. This may even result in an undesirable trip of
the protection system of a converter-dominated power sys- the converter. As transmission system operators (TSOs) have
tem. This paper derives a generic converter peak current imposed strict requirements on converter-based sources such
limitation method for three different VSC control strategies. as fault-ride-through (FRT) and voltage support capability, a
The impact of the control strategies and the combined
impact of a VSC with a synchronous condenser on distance variety of control strategies based on symmetrical components
protection are evaluated using a commercial relay through have been proposed to improve VSC performances under
hardware-in-the-loop (HIL) tests. Based on the test results, unbalanced conditions. In [3]–[7], the control strategies are
we propose to avoid using constant reactive power control developed based on the objectives of achieving balanced cur-
strategy. It poses an adverse impact on the reliability and rent injection, minimization of DC voltage ripples or nullifying
speed of distance protection regardless of the presence of
SC at the point of common coupling (PCC), while constant oscillations in either active or reactive powers. Generally, the
active power and balanced current control strategies favor above control strategies can be regarded as different special
the performances of distance protection. cases of [8], where flexible scalars are introduced to form
Index Terms—Converter, distance protection, short cir- the current references to flexibly control the oscillations in
cuit current, synchronous condensers, unbalanced faults. the active and reactive powers. In [9], the relative relationship
between the positive- and negative-sequence powers can be
flexibly adjusted. Based on [9], the studies in [10]–[12]
I. I NTRODUCTION regulate the grid phase voltages complying to predefined
boundaries and [13] focuses on the effectiveness of unbalanced
S a concern of the worldwide climate change and grow-
A ing demands for electricity, the integration of renewable
energy into power systems has gained increasing attention. For
voltage compensation. With converter current restricted in each
phase, [14] and [15] aim to maximally use the power capability
of the converter under unbalanced faults. Therefore, the short
example, Denmark aims to achieve 100% renewable energy
circuit response of a VSC can be significantly different from
supply by 2050, eliminating the dependency on fossil fuels [1].
each other under unbalanced faults and which control strategy
This has led VSC-based sources (e.g., Type-IV wind power
is more suitable is still under open discussion.
plants, photovoltaic power plants, HVDC transmissions) up to
Distance protection is widely utilized in high-voltage trans-
several hundred megawatts to be connected to the high-voltage
mission networks and a variety of studies has been conducted
transmission network.
regarding the impact of VSCs on distance protection. The
However, the control system of a VSC is sensitive to grid
speed of distance relays subject to balanced faults is evaluated
disturbances such as unbalanced faults. The negative-sequence
by simulations in [16], where different fault types and pene-
voltage appearing at the PCC will propagate in the VSC
tration levels of renewable generation are examined. In [17],
distance relays may refuse to trip when there is not enough
Manuscript received July 22, 2017; revised November 10, 2017, fault current under balanced faults and communication-aided
January 19, 2018 and April 8, 2018; accepted May 1, 2018. This work
was supported by the Danish ForskEL Project “Synchronous Condenser protection is suggested to overcome this problem. According
Application in Low Inertia Power Systems (SCAPP)” under Grant 12196, to [18], the control action of VSCs may cause an underreach-
administrated by Energinet. ing problem for the backup distance protection located on
Jundi Jia, Guangya Yang and Arne Hejde Nielsen are with the Center
for Electric Power and Energy, Department of Electrical Engineering, adjacent lines. However, the studies mentioned above neither
Technical University of Denmark, Kgs. Lyngby, Denmark (e-mail: jun- discuss unbalanced faults nor test on a real distance relay. Even
[email protected]; [email protected]; [email protected]). though unbalanced faults are of interests in [19]–[24], none
Peter Rønne-Hansen is with Siemens A/S, Ballerup, Denmark (e-mail:
[email protected]). of them has considered the impact of different VSC control
strategies incorporating a converter current limit in each phase.
0278-0046 (c) 2018 IEEE. Translations and content mining are permitted for academic research only. Personal use is also permitted, but republication/redistribution requires IEEE permission. See
http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
has been accepted
With conventional power plants gradually replaced by tive powers at the PCC can be expressed by (1) and (2) using
converter-based generations, a future power system may ex- the instantaneous power theory [30]:
perience significant drops on the system short circuit strength.
P+ P−
This could raise problems such as voltage instability, un- z }| { z }| {
− − − −
desirable dynamic behaviors of converters and malfunctions p = v + · i+ + +
P + v · iP + v · iP + v · iP (1)
of protection systems. Since the short circuit response of a | {z } | {z }
P P
e
synchronous condenser (SC) resembles that of a synchronous
Q+ Q−
generator, SCs may serve as an alternative to improve the z }| { z }| {
+ − − − −
system short circuit strength and thus the application of SCs q = v⊥ · i+ + +
Q + v⊥ · iQ + v⊥ · iQ + v⊥ · iQ (2)
has gained increasing attention in recent years [25]–[29]. The | {z } | {z }
Q Q
refurbishment of conventional power plants to SCs have been
e
proposed in [27], [28] to address dynamic voltage control T T

where v = [va vb vc ] and i = [ia ib ic ] represent the
issues and improve system short circuit ratios. However,
instantaneous voltage and current vectors at the PCC; the
conventional power plants may not always serve as the best
operator “·” denotes the arithmetic dot product of vectors;
locations for SCs and newly-installed SCs can be anticipated.
the superscripts “+” and “−” represent the positive- and
The studies in [28], [29] have presented methods on the
negative-sequence components; the subscript “⊥” refers to an
optimal allocation of SCs minimizing the total cost. The +
orthogonal version of the original vector (e.g., v+ leads v⊥
results have suggested that there is a need of installing new ◦ − − ◦
by 90 with the same amplitude; v leads v⊥ by 90 with the
SCs for more VSCs. Therefore, it is necessary to examine
same amplitude); P and Q are constant power terms resulting
the cooperation of VSC and SC from protection perspective
from the interaction between the voltage and current in the
considering different VSC control strategies, which has not
same sequence. They consist of positive-sequence power terms
been investigated in the previous studies.
(P + , Q+ ) and negative-sequence power terms (P − , Q− ). Pe
This paper investigates the impact of different VSC control
and Q e are oscillating power terms due to the interaction
strategies and the impact of incorporating synchronous con-
between the voltage and current in different sequences. With
densers on distance protection under unbalanced faults through
(1) and (2), the current references can be constructed in
HIL tests. Three representative control strategies from the
different ways, giving the short circuit current from a VSC
literature, namely constant active power, balanced current and
diverse characteristics.
constant reactive power control, are examined systematically.
In [8], a flexible control over the relative amplitudes of oscil-
In order to limit the converter current in each phase, a generic
lating active and reactive powers is presented. Two adjustable
converter peak current limitation method is derived for the ex-
scalars kp and kq are introduced in current references:
amined control strategies. The tests are designed based on the
variations of the sources of short circuit current, VSC control P ref
strategies, SC capacities, fault locations and types. According iref
P = 2 2 (v
+
+ kp v − ) (3)
|v+ | + kp |v− |
to the tests results from the commercial relay, we propose
Qref
to avoid using constant reactive power control strategy, while iref
Q = 2
+
2 (v⊥
−
+ kq v ⊥ ) (4)
constant active power and balanced current control strategies |v+ | + kq |v− |
favor the performances of distance protection.
where P ref and Qref are active and reactive power refer-
The rest of this paper is organized as follows. Section II
ences; iref
P and iref
Q are active and reactive current references
outlines VSC control strategies under unbalanced faults. A
respectively. With (3) and (4) substituted into (1) and (2), the
generic converter peak current limitation method is derived
oscillating power terms can be expressed as:
for the examined control strategies. Synchronous condensers
are shortly introduced in Section III. Section IV presents a (1 + kp )P ref + − (1 − kq )Qref + −
systematic HIL evaluation on the impact of different VSC Pe = v v + v⊥ v (5)
Dp Dq
control strategies and incorporation of synchronous condensers ref ref
on distance protection. In Section V, HIL tests are conducted Qe = (1 + kq )Q v+ v− + (1 − kp )P −
v + v⊥ (6)
⊥ ⊥
on the western Danish power system for further verifications. Dq Dp
2 2 2 2
Finally, conclusions are provided in Section VI. Dp = v+ + kp v− , Dq = v+ + kq v− (7)
In this paper, three most representative scenarios are con-

II. VOLTAGE S OURCE C ONVERTER C ONTROL sidered: 1) Constant active power control: the oscillations in
active power are nullified with kp = −1 and kq = 1; the short
A. Control Strategies
circuit current contains negative-sequence component; 2) Bal-
Typically, the control system of a grid-connected VSC anced current control: the short circuit current contains only
system consists of a slower outer controller and a faster inner positive-sequence component with kp = kq = 0. Oscillations
current controller. The inner current controller is responsible to with the same amplitude are registered in active and reactive
track the current references generated by the outer controller. powers; 3) Constant reactive power control: oscillation-free
For a three-phase three-wire VSC, zero-sequence components reactive power is achieved with kp = 1 and kq = −1; the
are not present. Therefore, the instantaneous active and reac- short circuit current contains negative-sequence component.
has been accepted
Voltage (p.u.)
Voltage (p.u.)
Voltage (p.u.)
Current (p.u.)
Current (p.u.)
Current (p.u.)
Power (p.u.)
Power (p.u.)
Power (p.u.)
Time (s) Time (s) Time (s)

(a) (b) (c)
Fig. 1. Short circuit response of a VSC (with 1.2 p.u. converter current limit) to an A–g fault on wye-winding side: (a) Constant active power control. (b)
Balanced current control. (c) Constant reactive power control. (Measurements are taken from the delta-winding side.)
B. Converter Current Limit solution of this quadratic equation can be obtained by:
p
When the PCC experiences a voltage dip, the current refer- −B + (B 2 − 4AC)
P = (14)
ences can become dangerously high if the active and reactive 2A
A = y 2 m2 cos2 γ + n2 sin2 γ

power references remain unchanged. This may damage power (15)
electronic devices and result in an undesirable disconnection of B = −2Qxy sin γ cos γ(mr − ns) (16)
the converter from the grid. Therefore, the active and reactive
C = Q2 x2 r2 sin2 γ + s2 cos2 γ − x2 y 2 Iˆ2

(17)
power references should be reduced properly to safely guard
m = v+ + kp v− , n = v+ − kp v−

the current within its limit in each individual phase. (18)
+ − + −
In [14], a peak current limitation method is presented by r = v + kq v , s = v − kq v (19)
fully utilizing the power capacity of the converter. It is capable 2 2 2 2
x = v+ + kp v− , y = v+ + kq v− (20)
of injecting active and reactive powers simultaneously with
the current in each phase restricted. However, the derived As long as the values of Q and Iˆ are given, the maximum
equations cannot be directly used if current references are active power Pa that is permitted without exceeding current
chosen as (3) and (4). According to the current ellipse theory limit in phase-A can be determined. With the value of γ
presented in [9], the relationship among the values of phase- changed to (|φ+ | − |φ− |)/2 + π/3 and (|φ+ | − |φ− |)/2 − π/3
A peak current I, ˆ average active power and average reactive respectively, the maximum active power Pb and Pc can also
power can be expressed by (8)–(11): be derived for phase-B and phase-C [9] by using (14)–(20).
Therefore, the maximum permitted active power P limit can
Iˆ2 = (IpL cos γ − IqL sin γ)2 + (IqS cos γ + IpS sin γ)2 (8) be decided by:
P+ P− P+ P− P limit = min {Pa , Pb , Pc } (21)
IpL = + + − , IpS = + − − (9)
|v | |v | |v | |v |
However, the equations above assume that the current limit
Q+ Q− Q+ Q− is not reached by only injecting reactive power. In order to
IqL = + + − , IqS = + − − (10)
v v v v impose a limit on reactive power, a similar procedure of
⊥ ⊥ ⊥ ⊥
|φ+ | − |φ− | deriving P limit is performed, where Q is regarded as an
γ= (11) unknown variable instead of P and the value of P is set to
2
zero. Then the limit for reactive power can be expressed by:
where IpL and IpS are the values of the long and short axes of
the active current ellipse; IqL and IqS are that of the reactive Qlimit = min {Qa , Qb , Qc } (22)
current ellipse; φ+ and φ− are the phase angles of positive- In order to illustrate the effectiveness of the above equations,
and negative-sequence voltages respectively. With (3) and (4) the presented method with Iˆ = 1.2 p.u. is implemented in the
taken as current references, (9) and (10) can be rewritten as: control system of a VSC. The VSC is connected to the delta
winding of the interface transformer. The wye winding, whose
(|v+ | + kp |v− |) (|v+ | − kp |v− |)
IpL = P neutral is solidly grounded, is connected to the grid. With
2 2 , IpS =P 2 2 (12)
|v+ | + kp |v− | |v+ | + kp |v− | an A–g fault applied at wye-winding side of the transformer,
(|v+ | + kq |v− |) (|v+ | − kq |v− |) the corresponding three-phase voltage, three-phase current and
IqL = Q 2 , IqS =Q (13)
2
|v+ | + kq |v− |
2
|v+ | + kq |v− |
2 output powers measured at delta-winding side are given in Fig.
1 as an example, where reactive power injection is prioritized.
By substituting (11)–(13) into (8), a quadratic equation can As expected, the fault current supplied by the converter is
be derived if P is considered as an unknown variable. The restricted within 1.2 p.u. in each phase after initial transients.
has been accepted
21/400 kV
Hardware Relay
DC
Physical Voltage Protection Physical
Logics
Amplifer
Grid GTDI GTAO
SG or SC
Fault
Fig. 2. Single-line diagram of the system for showing SG and SC response.

xxx
Control Grid V&I

Models
Protection Virtual Virtual Extract V&I

Signals GTDI GTAO signals
RTDS Simulation
Fig. 5. Hardware-in-the-loop test platform.
TABLE I
T EST SYSTEM PARAMETERS
VSC1 Rating 500 MW Line 1–2 156 87◦ Ω

VSC2 Rating 500 MW Line 2–3 & 2–4 146 87◦ Ω
SG Rating 600 MW Line 3–5 & 4–5 186 86◦ Ω
SC1 Rating 160 Mvar Line 6–2 96 86◦ Ω
SC2 Rating 50–250 Mvar VT 400 kV/100 V
VSC DC voltage 350 kV CT 1000 A/1 A
(a) (b)
Fig. 3. Short circuit response of synchronous sources to an A–g fault on
wye-winding side: (a) Synchronous generator. (b) Synchronous condenser.
(Measurements are taken from the delta-winding side)
during the transients. This is another benefit of using SC.
400/13 kV By comparing Fig. 3 to Fig. 1, it can be observed that the
150/400 kV SC1
short circuit current from synchronous sources in phase-A and
Bus 5
phase-C is more than three times higher than that of the VSC.
VSC1
In Fig. 3, the currents in phase-A and phase-C are almost
Examined Bus Bus 3 Bus4 out of phase by 180◦ . However, the current phases in Fig.
Distance
1 are mainly affected by the VSC control strategy. Among
Bus 2
150/400 kV Relay the three examined control strategies, constant active power
M control yields short circuit current most similar to that of Fig. 3
VSC2 Bus 1 in terms of phases. The short circuit current in balanced current
13/400 kV
SC2 400/21 kV control is completely symmetrical. For constant reactive power
Bus 6 SG control, the currents in phase-A and phase-C are closer to be
Fig. 4. Single-line diagram of the 400 kV test system
in phase rather than out of phase. Therefore, there might exist
certain impacts on distance protection under unbalanced faults
when replacing SG with VSC and applying an SC at the PCC.
III. SC R ESPONSES UNDER U NBALANCED FAULTS
IV. C ASE S TUDIES
A synchronous condenser (SC) is a synchronous machine
without a prime mover. During the start-up phase, the rotor of A. System Overview
the SC can be driven to the synchronous speed by a speed- The investigation is firstly tested on a 400 kV system shown
controlled motor [31]. Once the synchronous speed is reached, in Fig. 4. A synchronous generator (SG) is connected to bus 1
the field is supplied and then the SC can be connected to the (slack bus). Two VSCs (VSC1 and VSC2) are connected to bus
grid. Since there is no mechanical torque provided to its shaft, 5 and 6 respectively. Balanced current control is deployed for
SC cannot provide sustaining active power. Depending on the VSC1 while the control strategy of VSC2 can be freely shifted
excitation level performed by the automatic voltage regulator, among constant active power, balanced current and constant
an SC can either generate or absorb reactive power in the reactive power control. Under fault conditions, reactive power
steady state. One benefit of applying SC is that it can help injection takes the first priority and its reference is calculated
to increase the system short circuit strength. It can provide using Qref = |v+ | ·IQ . The value of IQ is obtained from [32]
short circuit current with a similar magnitude as a synchronous and it can be expressed as IQ = −2.5 |v+ | + 2.25 (0 ≤ IQ ≤
generator (SG) does to dynamically support the grid voltage. 1). To enhance the fault-ride-through operation of the VSCs,
In order to illustrate the effect clearly, a simple system shown both VSCs are equipped with a dynamic braking resistor on
in Fig. 2 is simulated with SG or SC as an example. Both the DC side to drain the excess electrostatic energy according
SG and SC share the same parameters. In steady state, they to [33]. Synchronous condensers SC1 and SC2 are equipped
are controlled to exchange zero reactive power with the grid at the PCC of these two converters. The capacity of SC2 can
and SG deliveries 1 p.u. active power. Figure 3 compares their be varied from 50 Mvar to 250 Mvar with a 50 Mvar step. A
response to an A–g fault on the high voltage side. distance relay is used to protect line 6–2 at the terminal near
As shown in Fig. 3, both SG and SC provide a considerable bus 6. Further details of the system are listed in Table I. In
amount of reactive power to the grid during the fault. Since steady state, VSC1 and VSC2 deliver 325 MW and 500 MW
SC is a rotating machine, it also provides inertia for the active power respectively with unity power factor while SG
system resulting in its active power exchange with the grid delivers 175 MW active power.
has been accepted
Z L12* Z L12*
100 100 100 4.0
Case 1 Case 1 Case 1 4.0
Case 2-P Case 2-P Case 2-P
80 Case 2-I 80 Case 2-I 80 Case 2-I 3.0 3.0
Response Time (ms)
Response Time (ms)
Response Time (ms)

Case 2-Q Case 2-Q Case 2-Q
X/Ohm(secondary)
X/Ohm(secondary)
60 60 60 2.0 2.0
1.0 1.0
40 40 40
0.0 0.0
20 20 20
-1.0 -1.0
0 0 0
25% 50% 75% 25% 50% 75% 25% 50% 75% -2.0
-2.0
Fault Location of Line 6-2 Fault Location of Line 6-2 Fault Location of Line 6-2 -2 -1 0 1 2 3 4
-2.0 -1.0 0.0 1.0 2.0 3.0 4.0
(a) (b) (c) R/Ohm(secondary) R/Ohm(secondary)
(a) (b)
Fig. 6. Average response time for Case 1 and Case 2: (a) A–g fault. (b) A–B Z L12* Z L12*
4.0 4.0
fault. (c) A–B–g fault. (100 ms in (b) represents refuse-to-trip failure)
3.0 3.0
X/Ohm(secondary)
X/Ohm(secondary)
2.0 2.0
1.0 1.0
In this paper, the investigation is conducted through HIL 0.0 0.0
tests using a commercial distance relay in real time. As -1.0 -1.0
illustrated in Fig. 5, the grid models are simulated in Real Time -2.0
-2.0 -1.0 0.0 1.0 2.0 3.0 4.0
-2.0
-2.0 -1.0 0.0 1.0 2.0 3.0 4.0
Digital Simulator (RTDS). The three-phase voltage and current R/Ohm(secondary) R/Ohm(secondary)
(c) (d)
signals needed by the distance relay are firstly extracted from
Fig. 7. Impedance plane of AB element for A–B fault at 50% of line 6–2: (a)
simulations using a Gigabit Transceiver Analogue Output Card Case1. (b) Case2–P. (c) Case2–I. (d) Case2–Q.
(GTAO). Since the maximum output voltage of its terminal is
limited by +/-10 V, the signals measured from the secondary Z/Ohm
K2:X L12* K3:X L12* K4:X L12* K5:X L12*
Case1
sides of the CT and VT in simulations have to be scaled down 10 Case2-P
Reactance
Case2-I
properly by adjusting the output scaling factor of the virtual 5
Case2-Q
GTAO component in RTDS. Then, an amplifier scales up the 0
signals measured from the GTAO so that amplified signals -5

0.000 0.010 0.020 0.030 0.040 0.050 t/s
equal to the real values obtained from the CT and VT. The Z/Ohm
K2:R L12* K3:R L12* K4:R L12* K5:R L12*
Case1
protection signal generated by the relay is sent back to RTDS 10 Case2-P
Resistance Case2-I
using a Gigabit Transceiver Digital Input Card (GTDI). When 5 Case2-Q
there is no current through its terminal, the digital input read 0
-5
by the RTDS processor card will be a logic “0”. In order to 0.000 0.010 0.020 0.030 0.040 0.050 t/s
change the value to logic “1”, an external 5 V signal is needed Fig. 8. Measured impedance values for A–B fault at 50% of line 6–2.
to drive the current into the GTDI terminal. Since the distance
relay uses a potential-free switch, a 5 V DC voltage source
is connected in series with the switch and GTDI terminals. provided by VSC2. The control strategy of VSC2 under
Therefore, a logic “1” will be registered if the switch closes, unbalanced faults are shifted among constant active power
forming a closed loop for the HIL test setup. In the following (Case2–P), balanced current (Case2–I) and constant reactive
case studies, the zone-1 performances of the relay using the power (Case2–Q) control. The converter current limit is set
classic method are evaluated. This study does not involve relay to 1 p.u. in each phase for both VSC1 and VSC2. For each
coordination and communication. The relay is set to protect type and location of the faults specified above, the tests are
90% of line 6–2 with the quadrilateral characteristic. Solid A– repeated 10 times as Case 1.
g, A–B, and A–B–g faults are simulated at 25%, 50% and 75% The average response time for Case 1 and Case 2 is
of line 6–2 respectively. The zero time instant corresponds to summarized in Fig. 6 for different fault types and locations.
the instant when the fault is initiated. The response time of Compared with Case 1, the response time in Case 2 generally
each test is defined as the time elapsed from the fault being increases, especially for A–B faults. This indicates that the
initiated until the protection signal being registered in RTDS. sensitivity of distance protection might be deteriorated as a
result of low short circuit current level in converter-dominated
B. Case 1: Synchronous generator solely power systems. Among the three examined control strategies,
constant active power and balanced current control give sim-
In this case, a more traditional power system is simulated by
ilar performances in terms of response time. However, with
replacing VSC2 with a 500 MVA synchronous generator and
constant reactive power control used for VSC2, the sensitivity
disconnecting SC2. Therefore, the short circuit current seen
of the relay is impacted significantly for A–g faults and the
by the distance relay is only contributed by the synchronous
relay even fails to trip under A–B faults. As an example, Fig. 7
generator. For each fault type and location, the HIL tests are
presents the impedance plane given by the relay when an A–B
repeated 10 times. All the faults are initiated when the phase-A
fault occurs at 50% of line 6–2 for Case 1 and 2. By comparing
voltage of bus 6 crosses zero from the negative to the positive
Fig. 7(b)–(c) to Fig. 7(a), more transients are observed before
in order to ensure each test has the same pre-fault conditions.
the locus stabilizes at the indicated fault location inside the
zone. However, in Fig. 7(d), the impedance locus exhibits
unfavorable characteristics. Even though the locus enters zone-
C. Case 2: Voltage source converter solely 1, it does not indicate a fault location clearly. Corresponding to
This case tests the system in Fig. 4 with SC2 disconnected each scenario in Fig. 7, the measured impedance versus time
so that the short circuit current seen by the relay is solely is further plotted in Fig. 8. The curve for Case2-Q moves in
has been accepted
while for constant reactive power control strategy, there are:

−
 +   
ref va⊥ + va⊥ sin(ωt)
P
iref
P = 2 2
 v + + v −  = M7 
b⊥ b⊥ 0 
+ −
|v | + |v | v + + v −
c⊥ c⊥ −sin(ωt)
(28)
−
 +   
Fig. 9. Inputs (primary-side values) to distance relay for A–B fault at 50% of va⊥ − va⊥ cos(ωt)
Qref
line 6–2. (Case2-Q) iref
Q = 2 2
 v + − v −  = M8 
b⊥ b⊥ 0 
|v+ | − |v− | v + − v − −cos(ωt)
c⊥ c⊥
(29)
and out from the effective zone during the fault while the
curves for the other three scenarios stay inside the effective where M5 -M8 represent the magnitudes of the currents. As the
zone stably during the fault. currents are restricted in magnitudes and the phase information
Corresponding to Case2-Q in Fig. 7(d), Fig. 9 presents the is of interest here, M5 -M8 are not given explicitly for brevity.
three-phase voltage and current inputs to the relay. The three- If the currents expressed by (26)-(29) are referred back to the
phase voltage (from the wye-winding side of the converter wye-winding side, the current in phase-A and phase-B will be
transformer) under fault conditions can be approximately out of phase (similiar to that of synchronous generators during
represented by vy . If it is referred to the delta-winding side, A–B faults) with (26) and (27). However, for constant reactive
the voltage becomes v∆ : power control with (28) and (29), the current in phase-A and
    phase-B will be in phase with the same magnitude, which
M1 sin(ωt) M3 sin(ωt) is significantly different from that of a synchronous source.
vy =  M1 sin(ωt)  , v∆ =  0  (23) Typically, the apparent impedance given by the AB element
−M2 sin(ωt) −M3 sin(ωt) of a distance relay is calculated by [34]:
√
where M1 , M2 and M3 = (M1 + M2 )/( 3N ) represent the V̇A − V̇B
ZAB = (30)
magnitudes of the corresponding voltages. N refers to the I˙A − I˙B
turns ratio of the transformer from the wye-winding side to the
delta-winding side. By decomposing v∆ into its symmetrical where V̇A , V̇B and I˙A , I˙B are the voltage and current inputs
components, there are: expressed in phasors. Therefore, currents being identical in
phase and amplitude (Fig. 9) may give a zero value in the
  − 
M4 sin(ωt + π6 ) M4 sin(ωt − π6 )
 + 
denominator of (30). This makes the distance relay unable to

va va
v +  =  M4 sin(ωt − π )  , v −  =  M4 sin(ωt + π )  calculate the impedance reliably, which explains why discon-
b 6 b 6
vc+ M4 sin(ωt + 5π 6 ) vc
−
M4 sin(ωt − 5π tinuous locus and curves are present for Case2-Q with constant
6 )
reactive power control strategy. Such discontinuous features
(24)
in AB elements can also be observed under A–B–g faults
where M4 = (M1 + M2 )/(3N ) represents the magnitudes of with constant reactive power control. Since the operation of
the corresponding voltages. Then, the orthogonal versions of distance relay under A–B–g faults also relies on AG and BG
the positive- and negative-sequence voltages are given by: elements, the relay can operate successfully. However, the use
 +   −
of constant reactive power control still adversely impacts the
M4 sin(ωt − π3 ) M4 sin(ωt + π3 )
    
va⊥ va⊥ reliability of the distance relay.
 v +  =  M4 sin(ωt − π)  ,  v −  =  M4 sin(ωt + π) 
b⊥ b⊥
+ −
vc⊥ M4 sin(ωt + π3 ) vc⊥ M4 sin(ωt − π3 )
(25) D. Case 3: Incorporation of synchronous condensers
In order to evaluate the combined effect of VSC and SC on
With current references (3) and (4) deployed, the phases of the distance relay when an SC is equipped at the PCC of a
iref
P and iref
Q are solely decided by the phases of the voltages VSC, the system in Fig. 4 is simulated with SC2 connected.
+ −
(v , v− , v⊥
+
and v⊥ ) and the values of the flexible scalars (kp Therefore, the fault current seen by the relay is jointly provided
and kq ). Therefore, by substituting (24) and (25) into (3) and by VSC2 and SC2. In addition, the capacity of SC2 is varied
(4), the current references expressed in abc-frame for constant among 50, 100, 150, 200 and 250 Mvar as different scenarios.
active power control strategy can be represented by: To ensure Case 2 and Case 3 has the same pre-fault conditions,
the excitation of SC2 is adjusted in each scenario so that Case
va − va−
 +   
ref cos(ωt)
P 3 has the same power flow results as Case 2 prior to the fault.
iref
P = 2
 +
2 vb − vb
−
= M5 −2cos(ωt)
|v+ | − |v− | v + − v − The same HIL tests are repeated under different VSC2 control
c c cos(ωt)
strategies and SC2 capacities.
(26)
 + −
   Firstly, the DC-link voltages of VSC2 under different sce-
va + va sin(ωt)
Qref narios are shown in Fig. 10. With the dynamic braking resistor,
iref
Q = 2 2
v + + v −  = M6 −2sin(ωt)
b b the DC-link voltages are restricted within 1.1 p.u. of the
|v+ | + |v− | v + + v − sin(ωt)
c c nominal value. According to Fig. 10(a), constant active power
(27) control gives nearly zero oscillation in the DC-link voltage,
has been accepted
100 100 100

Constant P Constant P Constant P
Balanced I Balanced I Balanced I
80 80 80
Constant Q Constant Q Constant Q
SG SG SG
Time (ms)
Time (ms)
Time (ms)
60 60 60
40 40 40
20 20 20
0 0 0
0 50 100 150 200 250 0 50 100 150 200 250 0 50 100 150 200 250
SC2 capacity (Mvar) SC2 capacity (Mvar) SC2 capacity (Mvar)
(a) (b) (a) (b) (c)
100 100 100
Fig. 10. Comparison on the VSC2 DC-link voltages for A–B fault at 50% of Constant P Constant P Constant P
line 6–2: (a) Case 2. (b) Case 3 with constant active power control in VSC2. 80
Constant Q
SG
80
Constant Q
SG
80
Constant Q
SG
Time (ms)
Time (ms)
Time (ms)
(The number in the legend represents the capacity of SC2) 60 60 60
40 40 40
Z L12* Z L12*
20 20 20
4.0 4.0
3.0 0 0 0
3.0 0 50 100 150 200 250 0 50 100 150 200 250 0 50 100 150 200 250
X/Ohm(secondary)
X/Ohm(secondary)

2.0 2.0
(d) (e) (f)
1.0 1.0
100 100 100
0.0 0.0 Constant P Constant P Constant P
80 80 80
-1.0 Constant Q Constant Q Constant Q
-1.0 SG SG SG
Time (ms)
Time (ms)
Time (ms)
60 60 60
-2.0 -2.0
-2 -1 0 1 2 3 4 5 -3 -2 -1 0 1 2 3 4 5 40 40 40
R/Ohm(secondary) R/Ohm(secondary)
20 20 20
(a) (b)
0 0 0
Fig. 11. Impedance plane of AB element for A–B fault at 50% of line 6–2 0 50 100 150 200
SC2 capacity (Mvar)
250 0 50 100 150 200
SC2 capacity (Mvar)
250 0 50 100 150 200
SC2 capacity (Mvar)
250
in Case 3 with constant reactive power control: (a) SC2: 100 Mvar. (b) SC2: (g) (h) (i)
200 Mvar.
Fig. 12. Average response time for Case 1, 2 and 3: (a) A–g fault at 25%.
(b) A–g fault at 50%. (c) A–g fault at 75%. (d) A–B fault at 25%. (e) A–B
fault at 50%. (f) A–B fault at 75%. (g) A–B–g fault at 25%. (h) A–B–g fault
while constant reactive power control yields the largest oscil- at 50%. (i) A–B–g fault at 75%. (100 ms in (d)–(f) represents refuse-to-trip
failure)
lations. On the other hand, as shown in Fig. 10(b), different
SC2 capacities have almost no notable effect on the DC-link
voltages if the control strategy of VSC2 is fixed.
With the help of SC2, the measured impedance locus is
generally improved with fewer transients in Case 3. As an
illustration, with constant reactive power control deployed,
Fig. 11 presents the impedance plane under A–B faults at
50% of line 6–2 when a 100 Mvar or a 200 Mvar SC2 is
(a) (b)
applied. In contrast to Fig. 7(d), the locus in Fig. 11 stabilizes
Fig. 13. Comparison on the fault current in phase-A with different SC2
at the indicated fault location inside the zone and moves out capacity for A–B fault at 50% of line 6–2: (a) Constant active power control.
of the zone after the fault is cleared by the relay. The average (b) Constant reactive power control. (The number in the legend represents the
response time in Case 3 is summarized in Fig. 12 with different capacity of SC2)
SC2 capacities, where the results from Case 1 (black dashed
line) and Case 2 (points corresponding to 0 Mvar) are also
included for the sake of comparison. of SC2. For constant active power control in Fig. 13(a), the
With the capacity of SC2 increased from 0 to 250 Mvar, the amplitude of the current is gradually boosted by increasing
speed of the distance relay is improved generally and getting SC2 capacity. However, for constant reactive power control
closer to that of Case 1. However, with a 50 Mvar SC2, the in Fig. 13(b), compared to the scenario without SC2 (IA(0)),
relay still fails to operate when constant reactive power control the amplitude of the current is even reduced when a 50 Mvar
is deployed. On the one hand, the capacity of a 50 Mvar SC2 SC2 is used (IA(50)). This effect makes the combined short
is relatively small compared to a 500 MVA VSC2 and hence circuit current such small that it is even not enough to activate
the short circuit current from SC2 is not significant in terms of the distance relay for impedance calculation. Even though
amplitudes compared with that from VSC2. On the other hand, the relay can operate successfully by further increasing the
as discussed above, the phases of the short circuit current from capacity of SC2, the total available short circuit current in Fig.
VSC2 under unbalanced faults can differ from that of SC2 13(b) is less than that in Fig. 13(a) with a SC2 of the same
significantly. As shown in Fig. 1–2 and derived in (23)-(29), capacity. This suggests that constant reactive power control
constant active power control yields the most similar short impairs the advantages of synchronous condensers due to the
circuit current to that of synchronous sources, while constant phase differences between the short circuit current from the
reactive power control gives the largest deviation in terms of VSC and the SC.
phases. Therefore, due to the differences in the current phases, As shown in Fig. 6(a) and Fig. 6(c), the relay performances
the application of SC2 at the PCC may even further reduce the under grounded faults are not impacted so much if constant
short circuit current level when constant reactive power control active power or balanced current control is used. Since the
is used. In order to illustrate this problem, Fig. 13 compares the wye-winding side of VSC2 interface transformer is solidly
combined short circuit currents of VSC2 and SC2 in phase-A grounded, it provides a path for the zero-sequence current to
for an A–B fault at 50% of line 6–2 with different capacities flow under grounded faults. As a result, the current limita-
has been accepted
SG:
100 100 100 LCC-HVDC SC
NJVB3
Constant P Constant P Constant P
Balanced I Balanced I Balanced I Sweden
80 80 80
Constant Q Constant Q Constant Q
SG SG SG Norway LCC-HVDC
Time (ms)
Time (ms)
Time (ms)
NVV VHA
60 60 60
FER
40 40 40
Wind Farm:
Norway Type III
20 20 20 TJE
VSC-HVDC
0 0 0 Wind Farm:
0 50 100 150 200 250 0 50 100 150 200 250 0 50 100 150 200 250 Type IV SG:
IDU SC
SC2 capacity (Mvar) SC2 capacity (Mvar) SC2 capacity (Mvar) SSVB3
TRI
(a) (b) (c) Wind Farm:
KAE
ASR MAL
Type III
100 100 100
Constant P Constant P Constant P EDR L3 LAG
FVO
Balanced I Balanced I Balanced I Wind Farm: KIN
80 80 80 SG: FYVB7
Constant Q Constant Q Constant Q Type IV
SG SG SG L2 REV
Time (ms)
Time (ms)
Time (ms)
60 60 60 SG: SC
ESVB3 SVS
L1
40 40 40 FGD
Denmark:
SG: Sjælland
20 20 20 SCEDR KAS LCC-HVDC
SKVB3
Holland
0 0 0 Examined
0 50 100 150 200 250 0 50 100 150 200 250 0 50 100 150 200 250 VSC-HVDC area Germany
(d) (e) (f) Fig. 16. Single-line diagram of a simplified western Danish power system
Fig. 14. Average response time for Case 4: (a) A–g fault at 25%. (b) A–g
fault at 50%. (c) A–g fault at 75%. (d) A–B–g fault at 25%. (e) A–B–g fault TABLE II
at 50%. (f) A–B–g fault at 75%. D IFFERENT TOPOLOGIES OF THE EXAMINED AREA PRIOR TO FAULTS
Component
Z L1E* Z L2E* Z L1E* Z L2E*
4.0 4.0 Topology L1 L2 L3 SCEDR
3.0 3.0
T1 Y N N N
T2 Y N N Y
X/Ohm(secondary)
X/Ohm(secondary)
2.0 2.0
T3 Y N Y N
1.0 1.0
T4 Y N Y Y
0.0 0.0 T5 Y Y Y N
-1.0 -1.0
-2.0 -2.0
-2.0 -1.0 0.0 1.0 2.0 3.0 4.0 -2.0 -1.0 0.0 1.0 2.0 3.0 4.0
(a) (b)
4.0
Z L12*
4.0
Z L12* (Fig. 15(b) and (d)). By comparing Fig. 15(a) to Fig. 15(b), it
3.0 3.0 can be observed that the locus exhibits more transients when
X/Ohm(secondary)
zero-sequence current is limited.

X/Ohm(secondary)
2.0 2.0
1.0 1.0
According to the test results in Fig. 12 and Fig. 14, the
0.0 0.0
improvement in the relay speed correlates with the VSC2
-1.0 -1.0
-2.0 -2.0
control strategy and the SC2 capacity. For the test system in
-2.0 -1.0 0.0 1.0 2.0 3.0
R/Ohm(secondary)
4.0 -2.0 -1.0 0.0 1.0 2.0 3.0
R/Ohm(secondary)
4.0
this paper, a SC2 of 100 Mvar at bus 6 is a good choice for
(c) (d) constant active power control since the improvement in speed
Fig. 15. Impedance plane for A–B–g fault at 25% of line 6–2 in Case 4 with starts to saturate with higher SC2 capacities. Similarly, a SC2
balanced current control: (a) AG and BG elements, no SC2. (b) AG and BG
elements, with 50 Mvar SC2. (c) AB element, no SC2. (d) AB element, with
of 150 Mvar is an optimal choice for balanced current control.
50 Mvar SC2. Besides requiring a higher capacity of SC2, balanced current
control does not perform as well as constant active power
control for phase-phase faults (Fig. 12(e)–(f)). In the worst
tion of VSC has less impact on the relay performances for case, constant reactive power control requires a synchronous
grounded faults if zero-sequence current with high amplitudes condenser of at least 200 Mvar to achieve similar performances
is present. In order to investigate the relay performances when as the other two control strategies.
there is a lack of zero-sequence current, a 300 Ω grounding
resistance is added to the grounding branch of the VSC2 V. T ESTS ON W ESTERN DANISH POWER SYSTEM
interface transformer. The same tests in Case 2 and Case 3 In this section, the HIL tests are further conducted on a
are repeated for A–g and A–B–g faults. The average response larger power system with more integrated converters. Figure
time of the relay is summarized in Fig. 14, which is denoted 16 presents the single-line diagram of the simplified western
as Case 4. By comparing Fig. 14(a)–(c) to Fig. 12(a)–(c) or Danish power system (DK1), where each 400 kV bus is
comparing Fig. 14(d)–(f) to Fig. 12(g)–(i), the response time assigned a three-letter name. The system is developed based
increases in Case 4 when there is no SC2 connected (0 Mvar). on the present DK1 in [28], [35], but with new planed lines to
Once SC2 is connected to the system, the response time is represent a future scenario. In order to generate a future case,
brought back for constant active and balanced current control. three conventional SGs (ESVB3, NJVB3 and SKVB3) are
This is because the step-up transformer of SC2 provides a path assumed to be phased out and disconnected from the system.
for the zero-sequence current that helps the relay operation. In the following studies, the tests are performed regarding
However, constant reactive power control still yields the worst the distance relay located on L1 near bus EDR. The relay
performances because its use reduces the combined short is set to protect 90% of L1. Five different topologies of the
circuit current from VSC and the rest of the system. Figure 15 examined area are considered, which are summarized in Table
presents the impedance plane of AG, BG and AB elements for II. For example, topology 2 (T2) means L1 and SCEDR are
an A–B–g fault at 25% of line 6–2 when there is no SC2 (Fig. present while L2 and L3 are disconnected from the system.
15(a) and (c)) and when there is a 50 Mvar SC2 connected Prior to a fault on L1, all five HVDC links and three wind
has been accepted
100 100 100

T1
T2
T4
T5
T1
T2
T4
T5
T1
T2
T4
T5
compared to the other two control strategies. The test results
80 80 80
T3
on DK1 are similar to those from Section IV, which further
Response Time (ms)
Response Time (ms)
Response Time (ms)

T3 T3
60 60 60
verifies that the reliability of distance protection is correlated
40 40 40
with the choice of VSC control strategies.
20 20 20
0 0 0
Constant P Balanced I
(a)
Constant Q Constant P Balanced I
(b)
VI. C ONCLUSIONS
Constant Q Constant P Balanced I
(c)
Constant Q
Z L12* Z L12*
Fig. 17. Average response time of the distance relay under different pre-fault With the converter current limit considered, this work in- 4.0 4.0
3.0 3.0
topologies for an A–B fault at: (a) 25% of L1. (b) 50% of L1. (c) 75% of vestigates the impact of VSC control strategies and the in-
X/Ohm(secondary)
X/Ohm(secondary)
2.0
L1. (100 ms represents refuse-to-trip failure) 2.0
corporation of synchronous condensers on distance protection
1.0 1.0
4.0
Z1 Z L12*
4.0
Z1 Z L12*
0.0
under unbalanced 0.0faults through HIL tests. The test results
3.0 3.0 -1.0 have shown that the
-1.0 reliability and speed of distance protection
can be adversely -2.0 affected due to a lower short circuit current

X/Ohm(secondary)
X/Ohm(secondary)
2.0 2.0 -2.0

-2.0 -1.0 0.0 1.0 2.0 3.0 4.0 -2.0 -1.0 0.0 1.0 2.0 3.0 4.0
1.0 1.0 level and non-conventionalR/Ohm(secondary)

R/Ohm(secondary)
short circuit current characteristics,
Z L12*
0.0 0.0 4.0 especially when the 4.0 short circuit current is mainly contributed
Z L12*
-1.0 -1.0 3.0 by VSCs. The application 3.0 of a synchronous condenser at the
X/Ohm(secondary)
X/Ohm(secondary)
-2.0 -2.0
-2.0 -1.0 0.0 1.0 2.0 3.0 4.0 -2.0 -1.0 0.0 1.0 2.0 3.0 4.0
2.0
PCC helps retain better 2.0
relay performances under unbalanced
1.0 1.0
Z1
(a)
Z L12* Z1
(b)Z L12*
faults, but this depends on the deployed VSC control strategies.
0.0 0.0
4.0 4.0
-1.0
Based on the test
-1.0
results, we propose to avoid the use of
3.0 3.0
constant reactive power control together with distance relay.
X/Ohm(secondary)
X/Ohm(secondary)
-2.0 -2.0
2.0 2.0
-2.0 -1.0 0.0 1.0 2.0 3.0 4.0
1.0 1.0
It can cause
-2.0 -1.0 0.0 1.0 2.0 3.0 4.0
R/Ohm(secondary) the relay to be unable
R/Ohm(secondary) to calculate the impedance
0.0 0.0
accurately or can even lower the combined short circuit current
-1.0 -1.0 when there is a synchronous condenser connected at the PCC,
-2.0 -2.0
-2.0 -1.0 0.0 1.0 2.0 3.0 4.0
thus affecting the reliability of the relay. When the short
-2.0 -1.0 0.0 1.0 2.0 3.0 4.0
R/Ohm(secondary) R/Ohm(secondary) circuit current is solely provided by a VSC, both constant
(c) (d)
active power and balanced current control can be considered
Fig. 18. Impedance plane for A–B fault at 75% of L1 in DK1: (a) T1, constant
reactive power control. (b) T1, balanced current control. (c) T2, balanced
for practice. When a synchronous condenser is connected at
current control. (d) T5, balanced current control. the PCC of a VSC, constant active power control can be
considered for the VSC as it requires a smaller synchronous
condenser than the other two control strategies to retain the
farms are in operation. It is assumed that the control strategy relay performances.
of the VSC-HVDC station at bus EDR can be changed among
constant active power, balanced current and constant reactive R EFERENCES
power control. The converter current limit is set to 1 p.u. [1] “Energy strategy 2050,” The Danish Ministry of Climate and Energy,
and the reactive power reference is generated using the same Copenhagen, Denmark, Feb. 2011.
assumption as described in Section IV. The whole system is [2] F. Blaabjerg, R. Teodorescu, M. Liserre, and A. V. Timbus, “Overview
of control and grid synchronization for distributed power generation
modelled with details in RTDS. systems,” IEEE Trans. Ind. Electron., vol. 53, no. 5, pp. 1398–1409,
For an A–B fault at different locations of L1, Fig. 17 Oct. 2006.
summarizes the average response time of the relay under [3] H. Song and K. Nam, “Dual current control scheme for PWM converter
under unbalanced input voltage conditions,” IEEE Trans. Ind. Electron.,
different VSC control strategies with various pre-fault system vol. 45, no. 5, pp. 953–959, Oct. 1998.
topologies of the examined area. Regarding T1 where the short [4] A. Yazdani and R. Iravani, “A unified dynamic model and control for
circuit current seen by the relay is mainly provided by the the voltage-sourced converter under unbalanced grid conditions,” IEEE
Trans. Power Del., vol. 21, no. 3, pp. 1620–1629, Jul. 2006.
VSC station, the relay has the slowest response and fails to [5] S. Alepuz, S. Busquets-Monge, J. Bordonau, J. A. Martı́nez-Velasco,
trip when constant reactive power control is deployed. Figure C. A. Silva, J. Pontt, and J. Rodrı́guez, “Control strategies based on
18(a)–(b) presents the impedance plane for A–B faults at 75% symmetrical components for grid-connected converters under voltage
dips,” IEEE Trans. Ind. Electron., vol. 56, no. 6, pp. 2162–2173, Jun.
of L1 when constant reactive power and balanced current 2009.
control are used. In Fig. 18(a), the locus still has discontinuous [6] K. Ma, W. Chen, M. Liserre, and F. Blaabjerg, “Power controllability
features similar to Fig. 7(d), leading to a refuse-to-trip failure. of a three-phase converter with an unbalanced AC source,” IEEE Trans.
Power Electron., vol. 30, no. 3, pp. 1591–1604, Mar. 2015.
Once again, constant reactive power yields the worst relay [7] R. Kabiri, D. G. Holmes, and B. P. McGrath, “Control of active and
performances. On the other hand, in T2–T5 where there are reactive power pipple to mitigate unbalanced grid voltages,” IEEE Trans.
additional short circuit current contribution from SCDER, L2 Ind. Appl., vol. 52, no. 2, pp. 1660–1668, Mar.–Apr. 2016.
[8] F. Wang, J. L. Duarte, and M. A. M. Hendrix, “Pliant active and reactive
or L3, the measured impedance locus has fewer transients power control for grid-interactive converters under unbalanced voltage
(by comparing Fig. 18(c)–(d) to Fig. 18(b)) and the response dips,” IEEE Trans. Power Electron., vol. 26, no. 5, pp. 1511–1521, May
time of the relay is improved. When the short circuit current 2011.
[9] R. Teodorescu, M. Liserre, and P. Rodrı́guez, Grid converters for
is jointly provided by the VSC and synchronous sources photovoltaic and wind power systems. Chichester, UK: John Wiley
(T2–T5), constant active power control yields slightly better & Sons, Ltd., 2011.
performances than balanced current control in terms of the [10] J. Miret, A. Camacho, M. Castilla, L. G. D. Vicuña, and J. Matas,
“Control scheme with voltage support capability for distributed genera-
relay response time. However, constant reactive power control tion inverters under voltage sags,” IEEE Trans. Power Electron., vol. 28,
still leads to the longest response time of the distance relay no. 11, pp. 5252–5262, Nov. 2013.
has been accepted
[11] A. Camacho, M. Castilla, J. Miret, R. Guzman, and A. Borrell, “Reactive power-transmission/FACTS/Synchronous%20Condenser/Synchronous

power control for distributed generation power plants to comply with Condenser.pdf
voltage limits during grid faults,” IEEE Trans. Power Electron., vol. 29, [32] Technical Regulation 3.2.5 for Wind Power Plants With a Power Output
no. 11, pp. 6224–6234, Nov. 2014. Above 11 kW, Energinet, Fredericia, Denmark, Jun. 2015.
[12] A. Milicua, G. Abad, and M. Á. R. Vidal, “Online reference limitation [33] A. A. van der Meer, M. Ndreko, M. Gibescu, and M. A. M. M. van der
method of shunt-connected converters to the grid to avoid exceeding Meijden, “The effect of frt behavior of vsc-hvdc-connected offshore
voltage and current limits under unbalanced operation part I: theory,” wind power plants on ac/dc system dynamics,” IEEE Trans. Power Del.,
IEEE Trans. Energy Convers., vol. 30, no. 3, pp. 852–863, Sep. 2015. vol. 31, no. 2, pp. 878–887, Apr. 2016.
[13] F. Nejabatkhah, Y. Li, and B. Wu, “Control strategies of three-phase [34] S. H. Horowitz and A. G. Phadke, Power system relaying, third edition.
distributed generation inverters for grid unbalanced,” IEEE Trans. Power Chichester, UK: John Wiley & Sons, Ltd., 2006.
Electron., vol. 31, no. 7, pp. 5228–5241, Jul. 2016. [35] H. T. Nguyen, G. Yang, A. H. Nielsen, and P. H. Jensen, “Frequency
[14] A. Camacho, M. Castilla, J. Miret, A. Borrell, and L. G. D. Vicuña, stability improvement of low inertia systems using synchronous con-
“Active and reactive power strategies with peak current limitation for densers,” in Proc. 2016 IEEE Int. Conf. on Smart Grid Commun., pp.
distributed generation inverters during unbalanced grid faults,” IEEE 650–655, Sydney, NSW, Australia, Nov. 2016.
Trans. Ind. Electron., vol. 62, no. 3, pp. 1515–1525, Mar. 2015.
[15] J. L. Sosa, M. Castilla, J. Miret, J. Matas, and Y. A. Al-Turki, “Control Jundi Jia received the B.Sc. degree in electrical
strategy to maximize the power capability of pv three-phase inverters engineering and automation from Beijing
during voltage sags,” IEEE Trans. Power Electron., vol. 31, no. 4, pp. Jiaotong University, China, in 2012, and the
3314–3323, Apr. 2016. M.Sc degree in electric power engineering from
[16] R. Li, C. Booth, A. Dyśko, A. Roscoe, J. Zhu, and H. Urdal, “Protection KTH Royal Institute of Technology, Stockholm,
challenges in future converter dominated power systems,” in Proc. 6th Sweden, in 2014.
Protection, Automation and Control World Conf., pp. 1–6, Glasgow, UK, In 2015, he was a strategy consulting analyst
Jun./Jul. 2015. on resource utilities in Accenture, Beijing, China.
[17] S. M. Holder, L. Hang, and B. K. Johnson, “Investigation of transmission Currently, he is pursuing Ph.D. degree in the
line protection performance in an electric grid with electronically cou- Center for Electric Power and Energy, Technical
pled generation,” in Proc. North Amer. Power Symp., pp. 1–6, Manhatan, University of Denmark, Kongens Lyngby,
KS, USA, Sep. 2013. Denmark. In 2017, he was a visiting researcher in National Renewable
[18] L. He, C. Liu, A. Pitto, and D. Cirio, “Distance protection of AC grid Energy Laboratory (NREL), Colorado, USA. His research interests
with HVDC-connected offshore wind generators,” IEEE Trans. Power include power system dynamics and protection, power electronics and
Del., vol. 29, no. 2, pp. 493–501, Apr. 2014. integration of renewable energy.
[19] A. Roy and B. K. Johnson, “Transmission side protection performance
with Type-IV wind turbine system integration,” in Proc. North Amer.
Power Symp., pp. 1–6, Pullman, Washington, USA, Sep. 2014. Guangya Yang received the B.E., M.E., and
[20] A. Hooshyar, M. A. Azzouz, and E. F. El-Saadany, “Distance protection Ph.D. degrees all in the field of electric power
of lines emanating from full-scale converter-interfaced renewable energy system, in 2002, 2005, and 2008, respectively.
power plants part I: problem statement,” IEEE Trans. Power Del., Since 2009, he has been with the Technical
vol. 30, no. 4, pp. 1770–1780, Aug. 2015. University of Denmark, Kongens Lyngby,
[21] M. M. Alam, H. Leite, J. Liang, and A. D. S. Carvalho, “Effects of VSC Denmark, as a Postdoctoral Researcher, and
based HVDC system on distance protection of transmission lines,” Int. he is currently an Associate Professor with
J. of Elect. Power and Energy Syst., vol. 92, pp. 245–260, Nov. 2017. the Center for Electric Power and Energy,
[22] I. Erlich, T. Neumann, F. Shewarega, P. Schegner, and J. Meyer, “Wind Department of Electrical Engineering, Technical
turbine negative sequence current control and its effect on power system University of Denmark. Since 2009, he has
protection,” in Proc. IEEE PES General Meeting, pp. 1–5, Denver, CO, been leading several industrial collaborative
USA, Jul. 2013. projects in Denmark in the field of monitoring, operation and protection
[23] B. Chen, A. Shrestha, F. A. Ituzaro, and N. Fischer, “Addressing of renewable energy systems. His research interests include renewable
protection challenges associated with Type 3 and Type 4 wind turbine energy integration, smart grids, and cyber-physical energy systems.
generators,” in Proc. 68th Annu. Conf. for Protective Relay Engineers,
pp. 335–344, College Station, TX, USA, Mar. 2015.
[24] M. M. Alam, H. Leite, N. Silva, and A. D. S. Carvalho, “Performance
evaluation of distance protection of transmission lines connected with Arne Hejde Nielsen is an Associate Professor
vsc-hvdc system using closed-loop test in rtds,” Electric Power Systems at the Centre for Electric Power and Energy,
Research, vol. 152, pp. 168 – 183, Nov. 2017. Department of Electrical Engineering, Technical
[25] M. Nedd, Q. Hong, K. Bell, C. Booth, and P. Mohapatra, “Application University of Denmark, Kongens Lyngby,
of synchronous compensators in the gb transmission network to address Denmark. He has 30 years experience in
protection challenges from increasing renewable generation,” in Proc. electric power engineering; the first years
Cigre Study Committee B5 Colloq., pp. 1–6, Auckland, New Zealand, were from ASEA AB, Central Research and
Sep. 2016. Development Department, Sweden, with focus
[26] S. Kynev, G. Pilz, and H. Schmitt, “Comparison of modern statcom and on measurement technology and motor design
synchronous condenser for power transmission systems,” in Proc. IEEE and control. Over the past decade, his focus has
Elect. Power and Energy Conf., pp. 1–6, Ottawa, Canada, Oct. 2016. been on electric power systems, especially on
[27] N. Masood, R. Yan, T. K. Saha, and S. Bartlett, “Post-retirement the implementation of renewable energy sources in the power system.
utilisation of synchronous generators to enhance security performances
in a wind dominated power system,” IET Gener., Transm. Distrib.,
vol. 10, no. 13, pp. 3314–3321, Oct. 2016. Peter Rønne-Hansen received the M.Sc.
[28] Z. H. Rather, Z. Chen, P. Thøgersen, and P. Lund, “Dynamic reactive degree in electrical engineering technologies
power compensation of large-scale wind integrated power system,” IEEE from Technical University of Denmark, Kongens
Trans. Power Syst., vol. 30, no. 5, pp. 2516–2526, Sep. 2015. Lyngby, Denmark in 1991. He had various
[29] E. Marrazi, G. Yang, and P. Weinreich-Jensen, “Allocation of syn- working experience as an electrical engineer
chronous condensers for restoration of system short-circuit power,” J. in different companies, such as ELSAM
of Modern Power Syst. and Clean Energy, vol. 6, no. 1, pp. 17–26, Jan. Engineering, Vattenfall A/S, Denmark, and N1
2018. A/S, Denmark. Currently, he is a protection
[30] H. Akagi, E. H. Watanabe, and M. Aredes, Instantaneous power theory engineer in Siemens A/S, Ballerup, Denmark.
and applications to power conditioning. Hoboken, NJ, USA: John His interests includes power system protection
Wiley & Sons, Inc., 2006. and renewable energy integration.
[31] The stable way–synchronous condenser solutions. (2017, Jul.).
[Online]. Available: https://www.energy.siemens.com/hq/pool/hq/

Impact of VSC Control Strategies and Incorporation of Synchronous Condensers On Distance Protection Under Unbalanced Faults

Uploaded by

Copyright:

Available Formats

Impact of VSC Control Strategies and Incorporation of Synchronous Condensers On Distance Protection Under Unbalanced Faults

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Impact of VSC Control Strategies and Incorporation of Synchronous Condensers On Distance Protection Under Unbalanced Faults

Uploaded by

Copyright:

Available Formats

This articleDownloaded

has been accepted

Impact of VSC Control Strategies and

proposed in [27], [28] to address dynamic voltage control T T

In this paper, three most representative scenarios are con-

Time (s) Time (s) Time (s)

Fig. 2. Single-line diagram of the system for showing SG and SC response.

Control Grid V&I

Protection Virtual Virtual Extract V&I

Fig. 5. Hardware-in-the-loop test platform.

VSC1 Rating 500 MW Line 1–2 156 87◦ Ω

Response Time (ms)

Response Time (ms)

tests using a commercial distance relay in real time. As -1.0 -1.0

GTAO component in RTDS. Then, an amplifier scales up the 0

signals measured from the GTAO so that amplified signals -5

there is no current through its terminal, the digital input read 0

while for constant reactive power control strategy, there are:

100 100 100

SC2 capacity (Mvar) SC2 capacity (Mvar) SC2 capacity (Mvar)

zero-sequence current is limited.

100 100 100

Response Time (ms)

Response Time (ms)

can be adversely -2.0 affected due to a lower short circuit current

2.0 2.0 -2.0

1.0 1.0 level and non-conventionalR/Ohm(secondary)

[11] A. Camacho, M. Castilla, J. Miret, R. Guzman, and A. Borrell, “Reactive power-transmission/FACTS/Synchronous%20Condenser/Synchronous

You might also like