Study on Transient Overvoltages in the

Converter Station of HVDC-MMC links

H. Saad, P. Rault, S. Dennetière, K. Dudas

Abstract – Transient overvoltages in converter station Several articles and research work have been performed on
equipment are difficult to predict using analytical tools, therefore insulation coordination (and fault behavior) of HVDC-LCC
it is conducted by means of EMT simulations. To get the worst link as in [2] and [3]. However, there is only few articles
case values, several HVDC set point configurations (as regarding insulation coordination on HVDC-MMC link
active/reactive power set points) and fault locations inside the dealing with internal converter station faults. In [1], an
converter station must be simulated. Parametric studies using
overview on the overvoltages in MMC station is presented and
EMT-type software is conducted, in this paper, to simulate this
high number of scenarios. A generic HVDC-MMC link and the
in [5]-[4] studies on transient overvoltages and the impact on
impact of arm inductance location are considered. Transient the dc cable is performed. In this paper, parametric studies
overvoltages at each electrical nodes in the converter station are using EMTP-RV software [8] is conducted to simulate this
provided and analyzed. These set of results and studies provide high number of scenarios and to identify the worst case
insights for researchers and engineers who are involved in scenario.
insulation coordination of HVDC-MMC link. Circuit configuration of converter station can vary
depending on project specification and manufacturers. The
Keywords: Insulation coordination, EMTP, HVDC impact of the arm reactor location on equipment stresses is
transmission, MMC, VSC, station faults, switching overvoltage. also studied. A generic HVDC-MMC link based on [7] and on
the Cigré DC grid benchmark [6] is considered. Overvoltage at
I. INTRODUCTION each electrical node in the converter station are presented.

T he inclusion of High Voltage Direct Current (HVDC)

transmission link in ac grids is expanding rapidly. The
use of voltage source converters (VSCs) based on Modular
These set of results and studies are useful for researchers and
engineers who are involved in insulation coordination of
HVDC-MMC station. XLPE technology is used more and
Multilevel Converter (MMC) topology is becoming more more for HVDC cables projects. The main advantages of
attractive mainly due to their higher performances and lower XLPE cables compared with Mass Impregnated (MI) and OF
cost. (oil-filled) cables are their cost and their environment impact.
During the design phase of a HVDC project, insulation Nevertheless they are more sensitive to voltage transients and
coordination studies is a crucial point for the lifetime of the especially polarity reversal [5]. This paper also contributes to a
system and, therefore, should be addressed carefully. In better assessment of transients that can stress DC cables.
HVDC-MMC link system, if the equipment withstand voltage The paper is organized as follow: Section II introduces the
is chosen to be too high, manufacture of converter equipment VSC-HVDC generic model used in this study. Section III
can be difficult and will increase the total cost of the system. If describes the parametric test setup considered for running fault
the withstand voltage of equipment is chosen to be too low, the transient studies. Section IV displays and analyses the maximal
failure probability because of malfunction and faults will voltage stress on different components, while providing time
consequentially increase, causing unavailability of the HVDC domain results for relevant situations. Section V analyses the
link in return. behaviour of the worst case faults. Finally, Section VI resumes
Therefore, one of the main objective of the insulation the overvoltage study considering DC surge arresters to protect
coordination studies is to establish the maximum steady-state, the XLPE cable.
temporary and transient voltage levels to which the various
components of the system will be exposed [3]. Internal faults II. HVDC SETUP
in the converter station must be evaluated to determine these The generic monopolar HVDC point-to-point link based on
maximum overvoltages at each equipment. These maximum the Cigré brochure B4-57 [6] and discribed in [7] is
overvoltages are difficult to predict using analytical tools, considered (Figure 1). The ac grids are presented as equivalent
therefore it is conducted by means of EMT simulations. sources with a short-circuit level. The transmission capacity of
However, to get the worst case value, several HVDC set point the link is 1,000 MW. The dc cable is rated ± 320 kV with
configurations (as active/reactive power set points) and fault 200 km lentgh and is modeled using a wideband line model
locations inside the converter station must be simulated. [9]. A MMC 201-level (200 SMs/arm) is considered with a
time step of 50 µs. The control strategy considers an
H. Saad, P. Rault and S. Dennetière are with Réseau de Transport active/reactive power flow control on MMC-1 and a dc
d'Electricité (RTE), Paris, France (e-mail: hani.saad@rte-france.com). voltage/reactive power control on MMC-2. Control system
details are reported in [6]. In this paper, the protection system
Paper submitted to the International Conference on Power Systems
Transients (IPST2017) in Seoul, Republic of Korea June 26-29, 2017 has been further developped to trip the link when fault occurs.
The considered protections are: AC undervoltages, overcurrent
on each arm, DC overcurrents and overvoltage on DC simulated is 512. DC pole-to-pole fault and AC faults on the
terminals. primary side are not considered since it is expected that such
After fault occurence, the protection system send the trip faults will not lead to higher overvoltages on converter
order, i.e. the converter is blocked and the ac circuit breaker is equippiment. Metalic faults (i.e. with no impedance) are
opened. In order to account for delays between protection considered because they intend to generate the worst
system and power circuit equipement, artificial delays are transients.
added between the order reception and the action: 200 µs for For each configuration, absolute maximum peak
blocking the MMC and 40 ms for opening the AC circuit overvoltage values are measured at each electrical node of the
breaker (BRK1 and BRK2 in Figure 1) . converter station as depicted in Figure 3. The following
voltages are: AC primary ( v prim abc ), AC secondary ( vsec abc ),
MMC-201L MMC-201L
L =0.15%
400/320 kV arm
1GW
640kV
Larm =0.15%
400/320 kV
arm-to-ground ( vugℓ ), arm pole-to-pole ( vuℓ abc ) and DC
abc
C=40 kJ/MVA C=40 kJ/MVA
1060 MVA 1060 MVA
BRK1 Ltrf = 18% Ltrf = 18% BRK2 pole-to-ground ( Vdc− and Vdc+ ).
+
1 2 200km 2 1
+
DC cable
S1 S2
+

+
MMC1 MMC2

Figure 1: MMC-HVDC transmission system F2 F3 F4 ..

..
..
.. ..
..
F7
Accurate MMC model must be used, because internal fault
leads to fast dynamic transients. Non-linear IGBT/diodes
model are used in the converter model to account for switching F5
surge when MMC blocks [7] and [10].
F6
III. PAREMETRIC STUDY SETUP F1
To identify the worst case scenario that leads to the
maximum transient overvoltage on each converter station
F8
equipment, a wide range of scenarios is simulated. Parametric
studies are conducted using EMTP-RV software. Maximum ..
..
..
..
..
..

active/reactive power transit and solid faults are considered for

all scenarios. Parametric studies considers: HVDC
active/reactive power transit directions, internal fault types,
Figure 2: Internal fault locations
internal fault instant occurring on the AC point on wave and
AC grid short-circuit level (SCL). Table 1 summarizes these
parameter variations and the number of configurations. In
Figure 2, MMC topology and fault locations are depicted. vua ..
..
..
..
..
..

Vdc+
Table 1: Setup configuration for parametric study
Parameter Number of configurations
v primabc vsecabc vug
8 configurations see Figure 2: a
F1 - Phase-to-ground fault
F2 - Three phase-to-ground fault
F3 - Two phase-to-ground fault
vℓg
a
Fault type F4 - Phase-to-phase fault
F5 - Positive arm-to-ground fault
F6 - Negative arm-to-ground fault vℓ a .. .. ..
Vdc−
.. .. ..
F7 - Positive DC pole-to-ground fault
F8 - Negative DC pole-to-ground fault
8 configurations : Fault instant with equal
distrubtion at each 2,5 ms Figure 3: Overvoltage measuring locations
Fault instant VacphA
IV. OVERVOLTAGE RESULTS
Time
instant In this section, the EMT parametric studies results are
Transit of active power 2 configurations : ±1000 MW presented and analyzed.
Transit of reactive power 2 configurations : ±300 MVar A. Maximum overvoltages without surge arresters
2 configurations for S1/S2
Short circuit level First step, the 512 configurations are simulated in time
SCLmax = 50 GVA and SCLmin = 3 GVA
domain with no surge arresters in order to get the maximum
From Table 1, the total number of configurations to be transient overvoltages of the converter station.
 F1 F2 F3 F4 F5 F6 F7 F8
550

max peak voltage (kV)

1) DC pole-to-ground overvoltage 500
For the simulated test cases, the maximum peak overvoltage 450

measured at Vdc− and Vdc+ are plotted in Figure 4. The x axis 400
350
represents the 512 simulated number and the fault types.
300
Whereas, the y axis presents the absolute maximum switching
250
overvoltage peaks registered for each simulated configuration. 0 64 128 192 256 320 384 448 512
Simulations
It is observed that F7 and F8 (i.e. positive and negative DC
pole-to-ground faults) represent the worst case scenario that Figure 6: Overvoltage on AC primary side
lead to the highest overvoltage values. Nevertheless, F5/F6 500
400 SCLmin
lead also to high overvoltages. The maximum switching 300

voltage (kV)
200
overvoltage registered is equal to 685 kV. Generally speaking, 100
0
the configuration of the link, fault instant and SCL have a -100
small impact on this overvoltage. We notice that; high SCL -200
-300 SCLmax
tends to increases the overvoltage in the considered test case. -400
-500
Figure 5 shows the time-domain results of Vdc+ due to F8 0.48 0.5 0.52 0.54 0.56 0.58 0.6
times (s)
including the impact of SCL and active power transit. It can be
Figure 7: v prim b due to F1
noticed, that the configuration -1000 MW/ -300 MVar/
SCLmax leads to the highest overvoltage peak for the 3) AC secondary overvoltage
considered test case. Figure 8 shows the maximum overvoltage on vsec abc of

F1 F2 F3 F4 F5 F6 F7 F8
each configuration. The phase-to-ground fault (F1) seems to be
700
max peak voltage (kV)

the fault type which generates the highest overvoltage on

600 vsec abc around 675 kV. A strong SCL and with a capacitive
500 reactive power transit tend to increase the overvoltage value.
Note that fault types F5 to F8 generate also high overvoltage
400
close to thus found during F1. The impact of the SCL on the
300 AC secondary voltage phace C waveforms is depicted in
0 64 128 192 256 320 384 448 512
Simulations
Figure 9.
Figure 4: Maximum overvoltages on Vdc− (blue) and Vdc+ 700
F1 F2 F3 F4 F5 F6 F7 F8
max peak voltage (kV)

(green) 600
680 500
620
voltage (kV)

400
560
-1000 MW/
500 -300 MVar/ +1000 MW/ 300
-1000 MW/ SCLmax -300 MVar/
440 SCLmax 200
-300 MVar/
380 0 64 128 192 256 320 384 448 512
SCLmin
320
Simulations
0.53 0.535 0.54 0.545 0.55 0.555 0.56 Figure 8: Maximum overvoltage on vsec abc
times (s)

Figure 5: Time domain waveforms of Vdc+ due to F8 800 +1000 MW/-300 MVar/
600
+1000 MW/
-300 MVar/SCLmin
SCLmax
voltage (kV)

400
200
2) AC primary overvoltage 0
Figure 6 shows the maximum overvoltage registered for -200
each simulation for the primary three phases v prim abc . The -400
-600
maximum value is equal to 532 kV. Typical configuration that 0.48 0.5 0.52 0.54 0.56 0.58 0.6
leads to this overvoltage is F1 fault type. Nevertheless, for the times (s)
simulated test cases, the maximum overvoltage on the primary Figure 9: vsec c due to F1
AC side is lower than the overvoltage that can be generated 4) Arm-to-ground overvoltage
from the AC network. The impact of the SCL strength is
The maximum overvoltage results of vugℓ (i.e. six arm-to-
illustrated in the time domain waveforms (Figure 7). One can abc
notice, that the maximum peak overvoltage occurs during the ground measurements) are provided in Figure 10. The
AC breaker opening instant (around 50ms after fault maximum peak reaches 814 kV during F1 fault. Strong SCL
occurrences) and not during fault instant. The fault current is and with an inductive reactive power transit tends to increase
inductive and leads to a switching overvoltage when it is the overvoltage value as depicted in Figure 11.
interrupted by BRK1.
 F1 F2 F3 F4 F5 F6 F7 F8
voltages are plotted in Figure 14, Figure 15 and Figure 16
max peak voltage (kV)

800
respectively.
600
F1 F2 F3 F4 F5 F6 F7 F8
800

max peak voltage (kV)

400 700

600
200
0 64 128 192 256 320 384 448 512 500
Simulations
400
Figure 10: Maximum overvoltages on vugℓ
abc 300
0 64 128 192 256 320 384 448 512
800 +1000 MW/-300 MVar/SCLmin Simulations
600
Figure 14: Vdc− Vdc+ overvoltage - Arm reactor at DC side
voltage (kV)

400
200
0
-200 F1 F2 F3 F4 F5 F6 F7 F8
800

max peak voltage (kV)

-400
-600
+1000 MW/-300 MVar/SCLmax
-800 600
0.48 0.5 0.52 0.54 0.56 0.58 0.6
times (s)
400
Figure 11: vℓg waveforms due to F1
c
5) Arm pole-to-pole overvoltage 200
0 64 128 192 256 320 384 448 512
The maximum overvoltage results of vuℓ abc (i.e. six arm Simulations

pole-to-pole measurements) are provided in Figure 12. The Figure 15: vsec abc overvoltage - Arm reactor at DC side
maximum peak reaches 1036 kV during F5/F6 and F1 fault. In F1 F2 F3 F4 F5 F6 F7 F8
max peak voltage (kV)

1000
this test case, strong SCL and with a positive active power
transit tends to increase the overvoltage value as depicted in 900
Figure 13. 800

F1 F2 F3 F4 F5 F6 F7 F8 700
max peak voltage (kV)

1000
600
0 64 128 192 256 320 384 448 512
900 Simulations
800 Figure 16: vuℓ abc overvoltage - Arm reactor at DC side
700

600 Based on these results, it can be noticed that the arm reactor
0 64 128 192 256 320 384 448 512
location has an impact on the overvoltage values and also on
Simulations
Figure 12: Maximum overvoltages on vuℓ abc the fault type that lead to the worst case scenario. For Vdc− and
Vdc+ overvoltages, comparison between Figure 4 and Figure 14
1000 +1000 MW/SCLmax
800
-1000 MW/ reveals that when the inductance arm is placed on the dc side
voltage (kV)

SCLmax
(instead of the ac side), the worst case are related with F5/F6
600
(rather than F7/F8). In addition, the maximum peak
400
+1000 MW/
overvoltage is increased and reaches around 770 kV. For the
200
SCLmin ac secondary overvoltages (Figure 15), the F5/F6 fault type
0
0.5 0.52 0.54 0.56 0.58 0.6
becomes the worst case (instead of the F1 fault type seen in
times (s) Figure 8). Also the maximum peak overvoltage equal to
Figure 13: vℓ c waveforms due to F1 730 kV is increase when arm reactor is installed on the DC
side. For vuℓ abc overvoltages, arm reactor location does not
B. Impact of arm inductance location have an impact on the maximum overvoltage value (1037 kV)
Converter station can have different circuit configuration and, as can be expected, when arm reactor is on dc side, only
depending on manufacturers and/or project specifications. In F1 faults lead to such high overvoltage.
this section, the impact of arm inductance location is
evaluated. The arm inductance that is located on the ac side
C. Fault transient analyses
terminal in Figure 3 is now placed on the dc side terminal of
each arm. Faults F5 and F6 are kept between the valves and In this section the general behavior of the worst fault cases
arm inductance. Same parametric studies, as depicted in Table is analyzed. Based on previous results (section II.A.), it is
1, are simulated. Maximum overvoltage measurements of the noticed that the most critical faults are the F1 and F5 to F8.
DC pole-to-ground, AC secondary and arm pole-to-pole The general behavior of the DC pole-to-ground faults (i.e. F7
and F8) is reported in [5] and will not be further explained due 700
to space limit. Moreover, the F5 and F6 faults are symmetrical, 600

voltage (kV)
500
therefore only F5 is investigated because the behavior is
400
identical. 300
In this section, the setup configuration of the link is - 200
1000 MW/-300 MVar/ SCLmax and fault instant occurs at the 100
0
maximum negative peak of vseca (see Figure 17). 0.51 0.52 0.53 0.54 0.55 0.56 0.57 0.58 0.59 0.6
times (s)
1) Phase-to-ground fault (F1)
For the considered configuration, F1 fault instant occurs at Figure 19: vua (blue) and vℓ a (green) during F1
0.532 sec. Overcurrent protection blocks the converter 600 µs
after the fault instant and the AC breaker open at around 2) Positive arm-to-ground fault (F5)
0.58 sec. As shown in Figure 17, a solid ground fault on phase Converter behavior due to F5 is close to F1 since both
A at the secondary side of the wye-delta transformer creates an faults are rather electrically similar (Figure 2).
overvoltage on phase B and C because the delta side reference Similar to previous case, the vsecb and vsecc following the
to the ground has a high impedance. During fault instant, the
fault, significantly increase as shown in Figure 20, but not as
faulty phase voltage is zero, therefore, vsecb and vsecc become
much as previous since the fault is not directly on the
equal to the phase to phase voltage of delta winding side (until secondary winding side.
the AC breaker opens). Overvoltages on DC voltages and arm voltages are depicted
The DC voltages and arm voltages of phase A are depicted in Figure 21 and Figure 22 respectivly. Different behavior is
in Figure 18 and Figure 19 respectivly. Before converter noted between F1 and F5 faults during the time between the
blocking, the voltage raise is limited by the arm inductor. It fault and blocking instant (around t=0.532 s). An equivalent
limits the inrush current coming from the SMs capacitors. circuit and a zoomed waveform, during this period intereval, is
Once the converter blocks, since the healthy phase to ground presented in Figure 23, Figure 24 and Figure 25. A voltage
voltages become higher than the DC voltage, the lower spike of around 640 kV is observed on Vdc+ . This overvoltage
freewheeling diodes of the each SM conduct (on the positive
corresponds to the prefault arm voltage of vua equal to 640kV
arm, when vsecb and vsecc are higher than Vdc+ , and vice versa).
(Figure 25) where one pole is abrutly clamped to zero when
Therefore, Vdc+ and Vdc− becomes almost equal to the maximum fault occurs. Unlike, internal bus fault in this case there is no
and minimum value of vsecb and vsecc . Hence the positive and arm inductance between the fault and the arms (except the
IGBTs and diodes stray inductor) to limit the current raise.
negative DC voltages oscillate at the frequency of the AC
The current spike magnitude depends on the submodule
system.
capacitor values, the power electonic devices parasitic
600 elements and the DC cable characteristics, therefore its value
400
rely on the data accuracy. Nevertheless, switching overvoltage
voltage (kV)

200
0
values of this phenomenon will not change drasticaly.
-200 600
-400 400
voltage (kV)

Vsec a
-600 200
0.51 0.52 0.53 0.54 0.55 0.56 0.57 0.58 0.59 0.6 0
times (s) -200
Figure 17: vsecabc during F1 -400
-600
600 0.51 0.52 0.53 0.54 0.55 0.56 0.57 0.58 0.59 0.6
400 times (s)
voltage (kV)

200 Figure 20: vsecabc during F5

0
-200 600 Zoom
-400 400
voltage (kV)

-600 200
0.51 0.52 0.53 0.54 0.55 0.56 0.57 0.58 0.59 0.6 0
times (s) -200
Figure 18: Vdc+ (blue) and −
Vdc (green) during F1 -400
-600
0.51 0.52 0.53 0.54 0.55 0.56 0.57 0.58 0.59 0.6
times (s)

Figure 21: Vdc+ (blue) and Vdc− (green) during F5

 700 DC pole-to-ground surge arresters was chosen to meet this
Zoom
600 requirement.
voltage (kV)

500
Similar to previous sections, the same parametric study
400
300
setup (section III. ) has been simulated including the surge
200 arresters, the MMC circuit and faults in Figure 2. The results
100
of Vdc− and Vdc+ including the surge arresters are presented in
0
0.51 0.52 0.53 0.54 0.55 0.56 0.57 0.58 0.59 0.6 Figure 26. The total energy absorption of the surge arresters
times (s)
installed at the DC terminal are presented in Figure 27.
Figure 22: vua (blue) and vℓ a (green) during F5
F1 F2 F3 F4 F5 F6 F7 F8
600

max peak voltage (kV)

Vdc+ 500

vua ..
..
..
..
..
..
400

300
0 64 128 192 256 320 384 448 512
F5 Simulations

Figure 26: Vdc− (blue) and Vdc+ (green) overvoltage

Figure 23: Equivalent circuit between fault and blocking
instant (around t=0.532 s) of F5 fault including surge arresters
F1 F2 F3 F4 F5 F6 F7 F8
15
600
energie (MJ)
voltage (kV)

10
400

200 Fault instant Blocking instant 5

0
0.531 0.5315 0.532 0.5325 0.533 0.5335 0.534 0
0 64 128 192 256 320 384 448 512
times (s)
Simulations
Figure 24: Zoom on Vdc+ during F5 Figure 27: Energy absorptions of DC surge arresters
700
600
From Figure 26, we can notice that the maximum residual
voltage (kV)

500
400
voltage is related to F5/F6 faults. As described in the
300 subsection IV. C. , this overvoltage depicted during F5/F6
Blocking instant
200 Fault instant faults are expectable (when the arm inductance is at ac side -
100 Figure 2) and are short (tens to hundreds of us) with high
0
0.531 0.5315 0.532 0.5325 0.533 0.5335 0.534 current value. Unlike F7/F8, the residual overvoltage induced
times (s) by F5/F6 are only limited at 584 kV by the surge arresters
Figure 25: Zoom on vua (blue) and vℓ a (green) during F5 since the inrush current produced from the capacitor’s arm is
high. Nevertheless, in case of F5/F6 faults, the energy
absorbed by the DC pole-to-ground surge arresters is rather
V. INFLUENCE OF DC SURGE ARRESTERS
low (around 2 MJ) with respect to F7/F8 faults which reaches
Based on the maximum switching overvoltage at the AC 12 MJ.
primary, AC secondary and arm to ground nodes, the Time domain results of Vdc+ due to F8 faults are illustrated
Switching Impulse Protective Level (SIPL) can be selected
accordingly to meet the insulation withstand level requirement. in Figure 28. The maximum peaks are now limited to 522 kV.
However, transient overvoltages at DC poles (Figure 5) do These overvoltage values will change according to the design
not meet XLPE DC cable requirements. This overvoltage is and characteristics of the surge arresters.
composed of switching and temporary overvoltages that does
not respect insulation coordination of the DC cable and can 500
voltage (kV)

damage the XLPE insulation and cable junctions [5]. Several 440
solutions exist to limit such overvoltage. A simple solution
380
consists in the installation of several surge arresters at DC
terminals. They must have a high energy rating since they are 320
not only designed to limit switching overvoltages but also 0.52 0.54 0.56 0.58 0.6 0.62
temporary overvoltages [5]. Based on the type test times (s)

recommended in [11], a typical value that cable can withstand Figure 28: Vdc+ including surge arresters due to F8
is in the range of 1.8 pu. In this paper, the characteristics of
 VI. CONCLUSIONS
This article has presented a methodology to study transient
overvoltages for HVDC-MMC link. A parametric studies (with
512 simulated configurations) using a generic HVDC-MMC
link has been used. Transient overvoltage results due to several
internal faults have been presented. It can be concluded that
the main internal faults that lead to the highest overvoltages
are the DC pole-to-ground, one phase-to-ground and arm-to-
ground faults. Short-circuit level, active/reactive power transit
have an impact on these overvoltage values.
The impact of the arm inductance location has been also
investigated. The worst case scenario and overvoltage value
are affected by the arm reactor location.
Finally, the inclusion and impact of the surge arrester
design at DC terminal has been presented. It has been shown
that for this specific example the higher residual voltage at dc
terminal is related to F5/F6 faults (instead of F7/F8 faults).
However, the highest energy absorption are related to F7/F8
faults.

VII. REFERENCES
[1] Gu, Yi Lei, et al. "Study of Overvoltage Protection and Insulation
Coordination for MMC based HVDC." Applied Mechanics and
Materials. Vol. 347. Trans Tech Publications, 2013.
[2] Elahi, H., et al. "Insulation coordination process for HVDC converter
stations: preliminary and final designs." IEEE Transactions on Power
Delivery 4.2 (1989): 1037-1048
[3] IEC, TS. "60071-5: 2002 Insulation coordination-part 5: procedures for
high voltage direct current (HVDC) converter stations." (2014).
[4] F. Badrkhani Ajaei and R. Iravani, "Cable Surge Arrester Operation Due
to Transient Overvoltages Under DC-Side Faults in the MMC–HVDC
Link," in IEEE Transactions on Power Delivery, vol. 31, no. 3, pp.
1213-1222, June 2016.
[5] S. Dennetière, H. Saad, A. Naud, P. Hondaa, “Transients on DC cables
connected to VSC converters” Jicable'15 - Versailles 21-25 June, 2015
[6] Cigré, W.G. B4-57." Guide for the Development of Models for HVDC
Converters in a HVDC Grid“ Cigré Technical Brochure 604, Dec. 2014.
[7] H. Saad, S. Dennetière, Mahseredjian, et al, "Modular Multilevel
Converter Models for Electromagnetic Transients," IEEE Transactions
on Power Delivery, vol. 29, no. 3, pp. 1481-1489, June 2014.
[8] J. Mahseredjian, S. Dennetière, L. Dubé, B. Khodabakhchian and L.
Gérin-Lajoie: “On a New Approach for the Simulation of Transients in
Power Systems”. Electric Power Systems Research, vol. 77, issue 11,
September 2007, pp. 1514-1520.
[9] A. Morched, B. Gustavsen, and M. Tartibi, “A Universal Model for
Accurate Calculation of Electromagnetic Transients on Overhead Lines
and Underground Cables”, IEEE Trans. on Power Delivery, vol. 14, no.
3, pp. 1032–1038, Jul. 1999.
[10] H. Jin, "Behavior-mode simulation of power electronic circuits," Power
Electronics, IEEE Transactions on, vol. 12, no. 3, pp.443-452, May
1997.
[11] Cigré, W.G. B1-32."Recommendations for testing DC extruded cable
systems for power transmission at a rated voltage up to 500 kV” Cigré
Technical Brochure 496, April. 2012.