Power Conversion Group
School of Electrical and Electronic Engineering
The University of Manchester, UK
DeepWind’2015
Trondheim
4-6th February 2015
Antony Beddard
Detailed Modelling of MMC-HVDC
Links
Power Conversion Group
School of Electrical and Electronic Engineering
The University of Manchester, UK
DeepWind’2015
Trondheim
4-6th February 2015
• Demand for VSC-HVDC systems is growing worldwide.
• Modular Multi-level Converters (MMC) is the VSC topology
of choice.
• Focus is on the DC components
VSC-HVDC
XT=15%
Yg/D
MMC2
Vs2(abc)
220kV 370kV
D/Yg
MMC1
Is1(abc)
PCC1XT=15%
370kV 410kV
Vs1(abc)
Idc1
Is2(abc)
AC voltage magnitude and
frequency control
1000MW
Windfarm1
Active and reactive power
control
Rbrak
Vdc2
PCC2
Vn
SCR=3.5
Zn
XT=15%
Yg/D
MMC3
Vs3(abc)
220kV 370kV
D/Yg
MMC4
Is4(abc)
PCC4XT=15%
370kV 410kV
Vs4(abc)
Idc4
Is3(abc)
AC voltage magnitude and
frequency control
1000MW
Windfarm2
Active and reactive power
control
Rbrak
Vdc3
125km DC cable
PCC3
Vn
SCR=3.5
Zn
130km DC cable
200km DC cableIdc2
Idc3
Power Conversion Group
School of Electrical and Electronic Engineering
The University of Manchester, UK
DeepWind’2015
Trondheim
4-6th February 2015
MMC
SM1
Arm
Single IGBT
Sub-module
SM2
SMn
SM1
SM2
SMn
SM1
SM2
SMn
SM1
SM2
SMn
SM1
SM2
SMn
SM1
SM2
SMn
Larm
+Vd/2
-Vd/2
Va
0V
Iua
Ila
Vua
Vla
• Types of MMC
• Half-bridge
• Full-bridge
• Alternate arm converter
• Selecting MMC Parameters
• Number of voltage levels
• SM capacitance
• Arm reactance
Power Conversion Group
School of Electrical and Electronic Engineering
The University of Manchester, UK
DeepWind’2015
Trondheim
4-6th February 2015
MMC Modelling Techniques
• Type 1 – Full physics based model
• Type 2 – Full detailed model
• Type 3 – Traditional detailed model (TDM)
• Type 3.5 – Accelerated model (AM)
• Type 4 – Detailed equivalent model (DEM)
• Type 5/6 – Average value model (AVM)
• Type 7 – Phasor domain model
• Type 8 – Power flow model
SM1
Arm
Single IGBT
Sub-module
SM2
SMn
SM1
SM2
SMn
SM1
SM2
SMn
SM1
SM2
SMn
SM1
SM2
SMn
SM1
SM2
SMn
Larm
+Vd/2
-Vd/2
Va
0V
Iua
Ila
Vua
Vla
Power Conversion Group
School of Electrical and Electronic Engineering
The University of Manchester, UK
DeepWind’2015
Trondheim
4-6th February 2015
Comparison of TDM(3), AM(3.5) and DEM(4)
DC line-to-line Fault
DC Fault
Signal DEM error (%) AM error (%)
Id 0.41 2.29
Va 0.22 1.12
Iua 0.51 1.83
Vc 0.07 0.07
TDM DEM AM
4.45 4.6 4.75 4.45 4.6 4.75 4.45 4.6 4.75
Time (s) Time (s) Time (s)
2
Id(k
A)
Va
(kV
)
300
-300
Iua
(kA
)
4
-8
Vc
(kV
)
21
18
-18
1
10
100
1000
10000
16 31 61
Sim
ula
tio
n d
ura
tio
n (
s)Number of MMC levels
TDM
DEM
AM
Power Conversion Group
School of Electrical and Electronic Engineering
The University of Manchester, UK
DeepWind’2015
Trondheim
4-6th February 2015
HVDC Cable Modelling
Types of HVDC Cable Model:-
• Lumped Parameter Model
• Bergeron Model
• Frequency Dependent Mode Model (FDMM)
• Frequency Dependent Phase Model (FDPM)
Layer Material
Radial Thickness
(mm)
Resistivity
(Ω/m)
Relative
Permittivity
Relative
Permeability
Conductor Stranded Copper 24.9 2.2x10-8* 1 1
Conductor screen Semi-conductive polymer 1 - - -
Insulation XLPE 18 - 2.5 1
Insulator screen Semi-conductive polymer 1 - - -
Sheath Lead 3 2.2x10-7 1 1
Inner Jacket Polyethylene 5 - 2.3 1
Armour Steel 5 1.8x10-7 1 10
Outer cover Polypropylene 4 - 1.5 1
Sea-return Sea water/air - 1 - -
*Copper resistivity is typically given as 1.68*10-8Ω/m. It has been increased for the cable model in PSCAD due to the stranded
nature of the cable which cannot be taken into account directly in PSCAD.
Power Conversion Group
School of Electrical and Electronic Engineering
The University of Manchester, UK
DeepWind’2015
Trondheim
4-6th February 2015
0
100
200
300
400
500
600
700
2.490 2.494 2.498 2.502 2.506 2.510
Vo
ltag
e (
kV)
Time (s)
Vd2-CEPIM
Vd2-Bergeron
Vd2-FDMM
Vd2-FDPM
0
2
4
6
8
10
12
14
16
2.490 2.494 2.498 2.502 2.506 2.510
Cu
rre
nt
(kA
)
Time (s)
Idc2-CEPIM
Idc2-Bergeron
Idc2-FDMM
Idc2-FDPM
Comparison of Cable Models
• The choice of cable model can have a
significant impact on the simulation
results.
• Computational efficiency is
approximately the same for the
travelling wave models.
• The CEPIM was found to be the least
computationally efficient model.
• FDPM is therefore the default model of
choice for typical VSC-HVDC studies
in this work.
Power Conversion Group
School of Electrical and Electronic Engineering
The University of Manchester, UK
DeepWind’2015
Trondheim
4-6th February 2015
XT=15%
Yg/D
MMC2
Vs2(abc)
220kV 370kV
D/Yg
MMC1
Is1(abc)
PCC1XT=15%
370kV 410kV
Vs1(abc)
Idc
Is2(abc)
AC voltage magnitude and
frequency control
1000MW
Windfarm
Active and reactive power
control
DC link voltage control and
AC voltage magnitude control
AC GridRbreak
Vd2
165km DC cablePCC2
Dynamic Braking System
• Typically employed for HVDC windfarm connections
• Normally located onshore
• Common models:
• Voltage dependent current source
• Power electronic switch with resistor
• Control – Two level switching, PWM etc.
Power Conversion Group
School of Electrical and Electronic Engineering
The University of Manchester, UK
DeepWind’2015
Trondheim
4-6th February 2015
HVDC Circuit Breaker
• Required for large HVDC grids
• Hybrid DC breakers are currently the preferred topology
• Modelling options – Cigre WG B4-57 technical brochure
SA
Ib
Is
Ls
Disconnector
Auxiliary DC Breaker
Main DC Breaker
Current
Limiting
Reactor
Residual
DC
Current
Breaker
Hybrid DC Breaker
BRKLs
SA
I Ib
Ic
Is
CL
Mechanical DC Breaker Solid-state Breaker
Power Conversion Group
School of Electrical and Electronic Engineering
The University of Manchester, UK
DeepWind’2015
Trondheim
4-6th February 2015
Control
• MMC Control
• Modulation - Nearest level control (NLC), selective harmonic elimination etc.
• Capacitor balancing controller (CBC)
• Circulating current suppressing controller (CCSC),
• Outer controllers similar to traditional VSCs. i.e. not specific to valve topology
Onshore
outer
controller
Current
controller
P*/V*dc/freq*
I* V*Q*/V*ac
Inner VSC
control
FSVSC
V
MTDC
controller
P*/V*dcOffshore
outer
controller
V*ac
freq*
Power Conversion Group
School of Electrical and Electronic Engineering
The University of Manchester, UK
DeepWind’2015
Trondheim
4-6th February 2015
MTDC Control Strategies
Control Method MMC1 control mode MMC4 control mode Comments
Centralised DC slack bus DC voltage & AC voltage
magnitude
Active power & reactive
power
P*=500MW
Voltage margin control DC voltage & AC voltage
magnitude
Voltage margin & reactive
power
Vd-High=620kV,
Vd-Low=580kV
Droop control Standard droop & AC
voltage magnitude
Standard droop & reactive
power
Droop gain =- 0.1
XT=15%
Yg/D
MMC2
Vs2(abc)
220kV 370kV
D/Yg
MMC1
Is1(abc)
PCC1XT=15%
370kV 410kV
Vs1(abc)
Idc1
Is2(abc)
AC voltage magnitude and
frequency control
1000MW
Windfarm1
Active and reactive power
control
Rbrak
Vdc2
PCC2
Vn
SCR=3.5
Zn
XT=15%
Yg/D
MMC3
Vs3(abc)
220kV 370kV
D/Yg
MMC4
Is4(abc)
PCC4XT=15%
370kV 410kV
Vs4(abc)
Idc4
Is3(abc)
AC voltage magnitude and
frequency control
1000MW
Windfarm2
Active and reactive power
control
Rbrak
Vdc3
125km DC cable
PCC3
Vn
SCR=3.5
Zn
130km DC cable
200km DC cableIdc2
Idc3
Power Conversion Group
School of Electrical and Electronic Engineering
The University of Manchester, UK
DeepWind’2015
Trondheim
4-6th February 2015
Example Simulation Results – MMC disconnection
Centralised DC slack bus Margin Control Droop control
Power Conversion Group
School of Electrical and Electronic Engineering
The University of Manchester, UK
DeepWind’2015
Trondheim
4-6th February 2015
Example Simulation Results – MMC1 AC fault
-2
-1.5
-1
-0.5
0
0.5
1
1.5
2
2.5
2.24 2.26 2.28 2.30 2.32 2.34 2.36 2.38 2.40 2.42 2.44Curr
ent
(kA
)
Time (s)
Droop
Margin
Slack
Power Conversion Group
School of Electrical and Electronic Engineering
The University of Manchester, UK
DeepWind’2015
Trondheim
4-6th February 2015
Thank you!
Power Conversion Group
School of Electrical and Electronic Engineering
The University of Manchester, UK
DeepWind’2015
Trondheim
4-6th February 2015
• Slide 1 – Picture courtesy of: Danish Energy Authority (left) CleanTechnica
(right)
• Slide 6 - Picture courtesy of ABB
References
Top Related