PSSE Offshore Wind HVDC

Offshore Wind Power in the North Sea
Grid Integration of 1000 MW Offshore Wind Power into the

Norwegian Power System
Knut Magnus Sommerfelt
Master of Science in Energy and Environment

Submission date: June 2008
Supervisor: Terje Gjengedal, ELKRAFT
Co-supervisor: Espen Hagstrøm, Statkraft
Norwegian University of Science and Technology

Department of Electrical Power Engineering
Problem Description
The last decade attention has really been drawn towards CO2 emissions and climate changes.
Norway has an ambitious goal of no net CO2 emission in 2050, and great effort has to be put into
reaching this target. Due to the great wind resources along the coast wind power could be an
important contribution. In order to install large amounts of wind power in the future, the turbines
should be built offshore. Statkraft has a vision of 1000 MW offshore wind power within 2012. The
distance to areas with depths suitable for today’s technology makes AC cable connection difficult,
indicating that HVDC should be used.
In this thesis grid integration of a 1000 MW offshore wind farm in the North Sea shall be
investigated. Simulations shall be done with HVDC Light in order to investigate the transient
stability on the onshore power system. Possible connection points in the grid have to be found.
Simulations with HVDC Light should be compared to a solution with AC cable connection in order
to investigate the impact of different transmission technologies on the onshore power system.
Assignment given: 15. January 2008

Supervisor: Terje Gjengedal, ELKRAFT
Preface
This thesis has been written at the Department of Electric Power Engineering at the
Norwegian University of Science and Technology in cooperation with Statkraft. The work is a
continuation of a project written during the autumn 2007, which dealt with transmission
system for offshore wind power.
In this thesis I have been introduced to PSS/E for the first time, which has been a valuable
experience. Because of problems with compiling of the user models in PSS/E, it took a long
time before I could start with the work in the thesis. Additionally, I did not receive the HVDC
Light model from ABB until April.
The HVDC Light model from NTNU could only provide a Fortran compiler for the newest
PSS/E version 31. The compiler used for this version is an Intel Fortran compiler. The object
files for the user models provided from Statnett and ABB were compiled with the Compaq
Visual Fortran compiler which is used for earlier versions of PSS/E. In order to establish a
dynamic library DLL file for PSS/E with the user models, the Compaq Visual Fortran
compiler is needed. Eventually, an earlier version of PSS/E was installed with a Compaq
Visual Fortran compiler belonging to Statkraft.
I would like to thank Leif Warland at Sintef Energy Research for help with getting started
with PSS/E, and Albert Leirbukt from ABB for help with the HVDC Light model. I would
also thank my supervisor at Statkraft Development and NTNU, Terje Gjengedal.
Finally, a large thanks to Charlotte and Kasper for all the support!
Knut Magnus Sommerfelt

Oslo 04.06.2008
iii
iv
Abstract
This master’s thesis is written at the Department of Electric Power Engineering at the
Norwegian University of Science and Technology. The work has been carried out at Statkraft
Development in Oslo. The thesis deals with transmission systems and grid integration of
offshore wind power, and is a continuation of the project written during the autumn 2007.
Norway has great potential for offshore wind power, but the depths just outside the coast
probably make floating wind turbines necessary. In order to use today’s technology for
offshore wind turbines with foundations on the seabed, water depths cannot be much larger
than 60 meters. It is possible to install the wind turbines at such depths, but the distance to
these areas make AC cable transmission difficult because of the reactive power production in
the cables. VSC HVDC is a technology well suited for offshore wind power, and HVDC
Light is now commercially available for rating up to 1174 MW. Theory for VSC HVDC in
systems in general is given first in the thesis. The case for this thesis is grid integration of a
1000 MW offshore wind farm into the Norwegian power system.
PSS/E was used in simulations of grid integration of the offshore wind power. Two possible
connection points in the south of Norway were found based on load flow simulations; Feda
and Kårstø. Only a load flow situation with peak load and production in the isolated
Norwegian power system was provided. Different load flow situations for the two connection
points were established in order to investigate the dynamic response at the connection points
for situations with lower load and production.
A case with two sets of 100 km AC cables was used for the dynamic simulations as well as a
HVDC Light link with a 600 km cable. SVCs were added at the connection point for the case
with AC cable connection in order to fulfil the requirement for capacitive and inductive power
factor at the connection point to the grid for wind power. No such compensation of reactive
power is necessary for HVDC Light, as the converter can adjust the power factor. The voltage
at the connection points is 300 kV. Dynamic simulations were done based on the fault ride
through requirement from the Norwegian TSO Statnett for power plants connected to voltage
level higher than 200 kV.
Different disturbances were done in the power system onshore close to the two connection
points. The simulations done with AC cables and SVCs for reactive power compensation
showed that the power system was not able to return to a stable operating point in all the
simulations. With HVDC Light on the other side, simulations showed that the voltage at the
connection points recovered to the pre fault value in all the simulations. The voltage recovery
was within the voltage profile defined in the fault ride through requirement, and the wind
turbines had to stay connected.
The wind farm was modelled as one equivalent generator offshore, and a standard PSS/E
induction generator model was used. For the case with HVDC Light, the voltage offshore was
practically unaffected by the disturbances onshore. The energy produced during the fault was
stored as rotational kinetic energy in the wind turbine in order to avoid the DC voltage to
increase drastically. This is an approximation done in this thesis. Wind projects planned with
HVDC Light will have a DC chopper. A fault onshore will not affect the wind farm, as the
power produced during the fault is dissipated in the DC resistance.
v
vi
Contents
1 INTRODUCTION .......................................................................................................................................... 1
2 AC SYSTEMS ................................................................................................................................................. 2
2.1 OVERHEAD LINES ..................................................................................................................................... 2
2.2 CABLES .................................................................................................................................................... 2
2.3 SURGE IMPEDANCE LOADING .................................................................................................................. 3
2.4 STRENGTH OF AC SYSTEMS ..................................................................................................................... 3
3 BACKGROUND FOR HVDC....................................................................................................................... 5
3.1 HVDC IN GENERAL ................................................................................................................................. 5
3.2 HVDC LIGHT........................................................................................................................................... 6
4 VSC HVDC TECHNOLOGY ....................................................................................................................... 7
4.1 SINGLE PHASE VSC ................................................................................................................................. 7
4.1.1 Switching ............................................................................................................................................ 8
4.2 THREE-PHASE VSC................................................................................................................................ 10
4.2.1 Switching .......................................................................................................................................... 10
4.3 POWER FLOW IN VOLTAGE SOURCE CONVERTERS................................................................................ 11
4.4 CONTROL MODES ................................................................................................................................... 13
4.4.1 Normal operation ............................................................................................................................. 13
4.4.2 Offshore wind power ........................................................................................................................ 14
5 PSS/E MODELS ........................................................................................................................................... 15
5.1 THE NORWEGIAN MODEL....................................................................................................................... 15
5.2 WIND TURBINE MODEL .......................................................................................................................... 16
5.3 THE HVDC LIGHT MODEL..................................................................................................................... 17
5.3.1 Modules ............................................................................................................................................ 17
5.3.2 Power flow representation ............................................................................................................... 17
5.3.3 Losses ............................................................................................................................................... 18
5.3.4 Dynamic model................................................................................................................................. 19
5.3.5 Offshore wind power application .................................................................................................... 21
5.3.6 Time step........................................................................................................................................... 22
6 LOAD FLOW CASES.................................................................................................................................. 23
6.1 CONNECTION POINT FEDA ..................................................................................................................... 23
6.1.1 Feda with HVDC from the Netherlands and Denmark ................................................................... 26
6.2 CONNECTION POINT KÅRSTØ ................................................................................................................. 28
6.3 BACKGROUND FOR LOAD FLOW SITUATIONS ......................................................................................... 31
6.4 ESTABLISHING OF THE LOAD FLOW CASES ............................................................................................ 32
6.4.1 Case 1 ............................................................................................................................................... 33
6.4.2 Case 2 ............................................................................................................................................... 33
6.4.3 Case 3 ............................................................................................................................................... 34
6.4.4 Case 4 ............................................................................................................................................... 34
6.4.5 Case 5 ............................................................................................................................................... 34
6.5 ADDING THE OFFSHORE WIND POWER ................................................................................................... 34
6.5.1 AC cable connection ........................................................................................................................ 35
6.5.2 HVDC Light connection................................................................................................................... 36
vii
7 DYNAMIC SIMULATIONS....................................................................................................................... 37
7.1 GRID CODE ............................................................................................................................................. 37
7.2 CONVERSION OF LOADS ......................................................................................................................... 38
7.3 DEFINITION OF DISTURBANCES .............................................................................................................. 38
7.3.1 Description of Event 1...................................................................................................................... 39
7.4 SIMULATION RESULTS............................................................................................................................ 40
7.5 CASE 1, AC CABLE................................................................................................................................. 41
7.5.1 Event 1, bus fault at Øie................................................................................................................... 41
7.5.2 Event 2, line fault and tripping of line............................................................................................. 44
7.5.3 Event 3, three-phase short-circuit at Feda...................................................................................... 45
7.6 CASE 1, HVDC LIGHT ........................................................................................................................... 46
7.7 CASE 2, AC CABLE................................................................................................................................. 51
7.8 CASE 2, HVDC LIGHT ........................................................................................................................... 53
7.9 CASE 3, AC CABLE................................................................................................................................. 54
7.10 CASE 3, HVDC LIGHT ........................................................................................................................... 56
7.10.1 Event 1, bus fault at Øie .................................................................................................................. 56
7.10.2 Event 2, line fault and tripping of line ............................................................................................ 56
7.11 CASE 4, AC CABLE................................................................................................................................. 56
7.11.1 Event 1, bus fault at Sauda .............................................................................................................. 56
7.11.3 Event 3, three-phase short-circuit at Kårstø................................................................................... 58
7.12 CASE 4, HVDC LIGHT ........................................................................................................................... 59
7.13 CASE 5, AC CABLE................................................................................................................................. 60
7.14 CASE 5, HVDC LIGHT ........................................................................................................................... 62
8 IMPROVEMENTS OF DYNAMIC BEHAVIOR.................................................................................... 64
8.1 CHANGING OF THE POWER FACTOR ....................................................................................................... 64
8.2 ADDING SYNCHRONOUS CONDENSERS .................................................................................................. 67
9 DISCUSSION ................................................................................................................................................ 71
10 CONCLUSION ............................................................................................................................................. 73
11 FURTHER WORK....................................................................................................................................... 75
12 REFERENCES.............................................................................................................................................. 76
viii
APPENDIX A: LIMIT CHECKING REPORTS ..................................................................................................I
APPENDIX B: SHORT-CIRCUIT POWER CALCULATIONS .................................................................... XI
CASE 1 .................................................................................................................................................................. XI
CASE 2 .................................................................................................................................................................. XI
CASE 3 .................................................................................................................................................................XII
CASE 4 .................................................................................................................................................................XII
CASE 5 .................................................................................................................................................................XII
APPENDIX C: ESTABLISHING OF LOAD FLOW CASES.......................................................................XIII
CASE 2 ................................................................................................................................................................XIII
CASE 3 ............................................................................................................................................................... XIV
CASE 5 .................................................................................................................................................................XV
APPENDIX D: DYNAMIC SIMULATION RESULTS ................................................................................. XVI
CASE 1, AC CABLE ........................................................................................................................................... XVII
CASE 1, HVDC LIGHT ...................................................................................................................................... XXII
CASE 2, AC CABLE ......................................................................................................................................... XXXI
CASE 2, HVDC LIGHT ................................................................................................................................ XXXVII
CASE 3, AC CABLE ..........................................................................................................................................XLVI
CASE 3, HVDC LIGHT ...........................................................................................................................................LI
CASE 4, AC CABLE .............................................................................................................................................. LX
CASE 4, HVDC LIGHT .....................................................................................................................................LXIV
CASE 5, AC CABLE ....................................................................................................................................... LXXIII
CASE 5, HVDC LIGHT ..................................................................................................................................LXXVI
ix
List of Tables
TABLE 2.1: TYPICAL PARAMETERS FOR OVERHEAD TRANSMISSION LINES [12] ......................................................... 2
TABLE 2.2: TYPICAL PARAMETERS FOR CABLES (PILC) [12] .................................................................................... 2
TABLE 5.1: MODULES IN HVDC LIGHT [1] .............................................................................................................. 17
TABLE 5.2: LOSSES IN HVDC LIGHT [1] .................................................................................................................. 18
TABLE 6.1: LOAD FLOW SITUATIONS ........................................................................................................................ 32
TABLE 7.1: LOAD CONVERSION................................................................................................................................. 38
TABLE 7.2: SIMULATIONS RESULTS........................................................................................................................... 40
x
List of Figures
FIGURE 3.1: MAXIMAL TRANSMITTED POWER THROUGH AN AC-CABLE [4].............................................................. 5
FIGURE 4.1: BLOCK DIAGRAM OF A SINGLE-PHASE INVERTER [17] ............................................................................ 7
FIGURE 4.2: OUTPUT WAVEFORMS OF THE SINGLE-PHASE SWITCH MODE INVERTER [17] ......................................... 7
FIGURE 4.3: ONE-LEG SWITCH MODE INVERTER [17].................................................................................................. 8
FIGURE 4.4: PULSE-WIDTH MODULATION OF A SINGLE-PHASE CONVERTER [17] ....................................................... 9
FIGURE 4.5: THREE-PHASE SWITCH-MODE CONVERTER [17].................................................................................... 10
FIGURE 4.6: PULSE-WIDTH MODULATION OF A THREE-PHASE CONVERTER [17] ...................................................... 11
FIGURE 4.7: SYSTEM TOPOLOGY OF A VSC HVDC SYSTEM [5]............................................................................... 12
FIGURE 5.1: PQ DIAGRAM OF A HVDC LIGHT CONVERTER [1]................................................................................ 18
FIGURE 5.2: HVDC LIGHT MODEL IN PSS/E [2]....................................................................................................... 20
FIGURE 6.1: OVERVIEW OF THE 300 KV POWER GRID IN THE SOUTH OF NORWAY ................................................... 24
FIGURE 6.2: LOAD FLOW RESULT AT FEDA, ORIGINAL LOAD FLOW .......................................................................... 25
FIGURE 6.3: LOAD FLOW RESULT WITH 1000 OFFSHORE WIND, CONNECTION POINT FEDA...................................... 26
FIGURE 6.4: LOAD FLOW RESULT AT FEDA WITH FULL IMPORT OF HVDC .............................................................. 27
FIGURE 6.5: OVERVIEW OF THE 300 KV GRID IN THE SOUTHWEST OF NORWAY ...................................................... 28
FIGURE 6.6: LOAD FLOW RESULT AT KÅRSTØ, ORIGINAL LOAD FLOW ..................................................................... 29
FIGURE 6.7: LOAD FLOW RESULT WITH 1000 MW OFFSHORE WIND, CONNECTION POINT KÅRSTØ ......................... 30
FIGURE 6.8: EQUIVALENT OFFSHORE WIND FARM CONNECTED WITH AC CABLES ................................................... 36
FIGURE 6.9: EQUIVALENT OFFSHORE WIND FARM CONNECTED WITH AC CABLES ................................................... 36
FIGURE 7.1: FAULT RIDE THROUGH REQUIREMENT FOR POWER PLANTS [19]........................................................... 37
FIGURE 8.1: LOAD FLOW AT KÅRSTØ (CASE 5), REDUCED COS(ĭ)........................................................................... 65
FIGURE 8.2: LOAD FLOW AT KÅRSTØ (CASE 5), SYNCHRONOUS CONDENSERS ........................................................ 67
xi
1 Introduction
The last decade attention has really been drawn towards CO2 emissions and climate changes.
Norway has an ambitious goal of no net CO2 emission in 2050. Emission quotas can be
bought in other countries, but a large contribution has to be done nationally. Even though the
Norwegian power production mainly consists of hydro power, great effort has to be put into
reaching the target. New power production has to be installed in order to be self supplied with
power in a year with average precipitation, but the time where large hydro power plants were
built in Norway has passed. Due to the great wind resources along the coast, wind power is a
promising alternative. In order to install large amounts of wind power, the turbines should be
built offshore in the future; the wind resources are even greater offshore, and the wind
turbines will not occupy the coastline. With offshore wind power in the North Sea, oil and gas
installations can be supplied with offshore wind power instead of using gas turbines with low
efficiency.
The technology for offshore wind turbines is available, but only for turbines founded to the
seabed for depths up to about 60 meters. Several concepts for floating wind turbines are under
development, but they are not yet commercially available. Outside the Norwegian coast
depths quickly reach several hundred meters. This implies that floating turbines are necessary
for wind farms relatively close to the shore. However, there are areas in the North Sea south
of Norway with depths suitable for today’s offshore wind turbine technology. The distance is
several hundred kilometers, making AC cable connection difficult, or even impossible,
because of the large amount of reactive power produced by the cables. As a result of this
HVDC probably has to be used. Traditional HVDC based on thyristors is not well suited for
offshore wind power applications, among other things because of the need for short-circuit
capacity offshore. HVDC Light on the other hand, is a technology well suited for offshore
wind power. The technology is based on Voltage Source Converters with IGBT valves, and
now has capacity up to 1174 MW.
This thesis is a continuation of a project written during the autumn 2007, which was an
introduction to different transmission technologies for offshore wind power. VSC HVDC
technology is further described in this thesis. The purpose of this thesis is to study the impact
on the dynamic behavior of a 1000 MW offshore wind farm connected to the Norwegian grid
with different transmission technologies. Possible connection points in the south of Norway
for the offshore wind power shall be selected. PSS/E will be used for all power system
simulations. The model for the Norwegian power system is provided by Statnett, and the
model for HVDC Light developed by ABB will be used in simulations. Simulations with AC
cables will also be done in addition to HVDC Light, in order to compare the results.
With 1000 MW offshore wind power connected to the high voltage grid, the transmission
technology will have great influence on the system stability. Transient stability will be
investigated for the two different transmission technologies; AC cable and HVDC Light. The
wind farm is not modeled in detail, and the focus of the thesis will be power system impact
onshore. Some assumptions will have to be made because of limitations regarding the control
system for offshore wind power applications in the current HVDC Light model in PSS/E.
1
AC systems
2 AC systems
This chapter gives an overview over some aspects of AC power systems that are relevant
regarding connection of offshore wind power. The relationship between impedances in
transmission lines and cables is given. This can be used to explain the need for HVDC for
transmission in long cables. A brief introduction to AC system strength is also presented. The
chapter is partly a brief summary of the project written during the autumn 2007.
2.1 Overhead lines
Transmission lines can be described electrically with series impedance and shunt admittance
[12]. The series impedance consists of resistance R due to the resistivity in the conductor, and
reactance XL due to the magnetic field around the conductor. The shunt admittance consists of
conductance G due to leakage current between the phases and ground, and susceptance BC due
to the electric field between the conductors. The effect from the shunt conductance in
transmission lines is small, and usually neglected. Typical parameters for overhead lines are
given in Table 2.1 for different voltages.
Table 2.1: Typical parameters for overhead transmission lines [12]
2.2 Cables
Cables can be described electrically with the same elements as transmission lines. However
the series impedance and shunt admittance of power cables have different values from
transmission lines due to the construction of the cables. The shunt susceptance in cables is
much higher than in transmission lines because the conductors are closer to each other, and
the conductors are surrounded by metallic screens. Typical parameters for direct-buried
paper-insulated lead-covered cables (PILC) are given in Table 2.2 for different voltages.
Table 2.2: Typical parameters for cables (PILC) [12]
2
AC systems
For cables with high voltage, the three conductors are in separate cables. Hence, the
susceptance will decrease compared to a three-core cable because of increased distance
between the conductors.
2.3 Surge Impedance Loading
Transmission lines and cables can be described by their characteristic impedance ZC

according to Equation (2.1) [12].
R jZ L L§ R ·
ZC ¨1 j ¸ (2.1)
jZC C© 2Z L ¹
R is the series resistance, L is the series inductance and C is the shunt capacitance. The shunt
conductance G is normally ignored for transmission lines and cables. From the characteristic
impedance, surge impedance load SIL can be expressed according to Equation (2.2).
V02
SIL (2.2)
ZC
V0 is the rated voltage of the line or cable and ZC is the characteristic impedance of the line or
cable. If a line is loaded at a level below SIL, the line will produce reactive power. The effect
from the shunt capacitance is bigger than the effect from the series inductance. On the other
side, if a line is loaded at a level above SIL, the line will consume reactive power due to the
increased current in the line. The effect from the shunt capacitance is dependent of the voltage,
and hence approximately constant.
The SIL of a cable is much higher than for a transmission line for the same voltage [12]. This
is due to the high capacitance, as can be seen by comparing Table 2.1 and Table 2.2. A typical
power cable will not be able to transfer power equal to SIL, because of the thermal limit.
In the case with an offshore wind farm connected to the grid with AC cables, the reactive
power produced by the cable will be highest when the cable is connected, but the active power
transfer from the wind farm is small. The higher the active power, the more reactive power is
consumed in the series inductance. However, at full power AC cables will still produce large
amounts of reactive power.
2.4 Strength of AC systems
A normal parameter for measuring the strength at a busbar in an AC network is short-circuit

power SC MVA [9]. The definition of short-circuit power is given in Equation (2.3).
Vac2
SC MVA I fault Vac (2.3)
Z th
Ifault is the fault current for a three phase fault with zero impedance at the busbar, Vac is the
nominal line voltage at the point [kV] and Zth is Thevenin impedance seen from the point [ȍ].
3
AC systems
Even though the short-circuit impedance is equal to zero, the fault current will be limited by
the Thevenin impedance, which corresponds to the diagonal element in the node impedance
matrix for the system for the given busbar [7]. Physically, this means that the fault current has
to return through impedances from ground other places in the power system.
In fault studies, a generator is represented as Norton equivalent, with a current source in

parallel to the generator impedance connected to earth. The closer a generator is located to a
busbar, the more will it contribute to reducing the Thevenin impedance seen from the busbar.
More lines connected to the busbar will also contribute to reducing the Thevenin impedance;
hence also increase the short-circuit power as described in Equation (2.3).
In order to connect a HVDC link based on thyristor technology to an AC system, two

requirements have to be fulfilled [9]. The short-circuit power has to be large enough
compared to the power from the HVDC link. This is called short-circuit ratio, and is defined
according to Equation (2.4).
SC MVA Vac2
SCR (2.4)
PHVDC Z th PHVDC
SC MVA is the short-circuit power in the connection point [MVA], PHVDC is the rating of the
HVDC link [MW], Vac is the nominal line AC voltage in the connection point [kV] and Zth is
the Thevenin impedance seen from the connection point [ȍ]. A rule of thumb is defined in [9],
and recommends a SCR larger than 3 for traditional HVDC.
In addition to a sufficient large SCR, the mechanical inertia of the AC system has to be large
enough in order to maintain the required voltage and frequency. An effective inertia constant
Hdc is defined in [9] is defined according to Equation (2.5).
total rotational inertia of the AC system [MW s]

H dc (2.5)
MW rating of the AC link
The inertia constant is recommended to be at least 2 - 3 s in order to have satisfactory

operation of the HVDC link.
For offshore wind power, grid connection with AC cables will also need a certain level of
short-circuit power at the connection point compared to the power from the wind farm [23].
This will be shown in the dynamic simulations presented in chapter 7.
4
Background for HVDC
3 Background for HVDC

This chapter gives a brief introduction to HVDC in general, as a summery of the project
written during the autumn 2007.
3.1 HVDC in general
The first commercial HVDC transmission system was built in Sweden in 1954 [12]. The
technology was based on mercury arc valves. The development of thyristors made HVDC
systems more attractive. In 1972 the first thyristor based HVDC system provided a back-to-
back connection between the asynchronous areas of Quebec and New Brunswick. HVDC
systems based on converters with thyristors are often referred to as traditional HVDC.
There are three main reasons for using HVDC power transmission systems. For large
quantities of power over long distances, AC transmission is not economical due to the need
for compensation of reactive power [3]. Connection of two asynchronous areas is impossible
with AC. A back-to-back HVDC connection or a HVDC connection with cable or
transmission lines makes power transfer between the areas possible. Further on is power
transfer with AC underground or subsea cables is difficult due to the high capacitance of
cables, as described in 2.2. The distance beyond which it is impractical to use AC cables
depends on the transmitted power, voltage level, cable construction and compensation for
reactive power. Figure 3.1 illustrates the maximal transmitted power in AC cables for
different voltage levels and reactive power compensation strategies for a typical AC cable.
Figure 3.1: Maximal transmitted power through an AC-cable [4]
Traditional HVDC systems are not described in detail in this thesis. However, references to
traditional HVDC systems will occur in order to emphasize the difference for Voltage Source
Converter (VSC) HVDC systems.
5
Background for HVDC
3.2 HVDC Light
Unlike traditional HVDC systems, the converters in VSC HVDC systems are based on IGBT
transistors. HVDC Light is the name of ABB’s VSC HVDC technology, and has been
available since 1997. In this thesis, HVDC Light is used in all dynamic HVDC simulations.
The reason for this is that ABB offers a model in PSS/E representing HVDC Light. Siemens
has a similar technology to HVDC Light called HVDC PLUS. This technology is still not
installed, but commercially available.
Details regarding the VSC HVDC technology will be given in this thesis. In chapter 4, the
switching in general Voltage Source Converters is described, as well as power flow and
control modes. VSC HVDC systems are referred to in general. This is because the detailed
information about the switching of HVDC Light is restricted. However, the basic construction
of a HVDC Light system is similar to the technology presented in chapter 4. In this paragraph,
practical information relevant for offshore wind power is presented. It has to be emphasized
that costs of the technology is not presented, because ABB doesn’t give out this information.
The HVDC Light technology is developed for applications both onshore and offshore. For
offshore applications, the high-voltage equipment is installed inside a container with
ventilation system designed to protect it from the salt and humid air [1]. Costs of installing
equipment offshore are higher than onshore. Because of this, the layout for offshore use is
made more compact than for onshore use. An example presented in [1] is a 40 MW station.
For a land station with this power rating, the dimensions are 40 x 18 x 8,5 meters, and for
offshore applications the module is 16 x 10 x 15 meters. An example of a 700 MW land
station is also given. The dimensions of the station onshore are 60 x 30 x 20 meters. With the
same ratio between onshore and offshore dimensions, the dimensions of the 700 MW offshore
module would be 24 x 17 x 35 meters. This is however a very rough assumption, but gives an
image of the dimensions.
The HVDC Light submarine cables are insulated with XLPE (Cross-linked polyethylene).
ABB is the first cable producer to use this insulation in DC cables. Cables insulated with
XLPE are lighter and more mechanically robust [1]. The result is that the cables can be laid
on greater depths than paper insulated cables. Repairing of the cables is also easier than for
paper insulated cables.
6
VSC HVDC technology
4 VSC HVDC technology

This chapter gives an introduction to the technology used in VSC HVDC systems. The
information is mainly based on [17] and [1], and explains the basic theory of switch-mode
converters, such as the converters in HVDC Light.
The thyristors in a traditional HVDC converter need a three phase voltage on the AC-side to
be able to commutate [17]. A normal terminology for this type of converter is LCC (Line
Commutated Converter). The converter in HVDC Light is based on IGBT (Insulated Gate
Bipolar Transistors). Unlike thyristors, the transistors in have turn off capability. This results
in a better control of the converter output. The input to switch-mode converters is DC-voltage;
the terminology for this kind of converters is therefore Voltage Source Converters (VSC).
4.1 Single phase VSC
Figure 4.1 shows a block diagram for a single phase switch-mode inverter. The voltage Vd is
the DC-voltage and the voltage v0 is the AC-voltage. The power flows from the DC-side to
the AC-side, and assuming the inverter supplies an inductive AC motor load, the AC-current
i0 will lag the AC-voltage v0. This is shown in Figure 4.2.
Figure 4.1: Block diagram of a single-phase inverter [17]
Figure 4.2: Output waveforms of the single-phase switch mode inverter [17]
In period 1, both the current i0 and the voltage v0 are positive. The power flows from the DC-
side to the AC-side. Due to the lagging of the current, in period 2 the voltage is negative while
the current is still positive. The power flows from the AC-side to the DC-side, and the
converter is in a rectifier mode of operation. In period 3 both the current and voltage are
7
VSC HVDC technology
negative, but the power again flows from the DC-side to the AC-side. Finally, in period 4 the
current is negative and the voltage is positive. The power flows from the AC-side to the DC-
side. From this, it can be seen that the converter has to be able to work in all four quadrants of
the i0 - v0 plane.
4.1.1 Switching
The output AC-voltage in a switch-mode inverter is obtained by switching the transistors in a
certain way. First, the switching of the single phase inverter illustrated in Figure 4.1 will be
presented. Figure 4.3 shows a half bridge single-phase switch-mode converter. The voltage Vd
is assumed to be a constant DC-voltage, vAN is the output AC-voltage and i0 is the AC-current.
Figure 4.3: One-leg switch mode inverter [17]
In order to control the switching, a signal needs to be given to the transistors. This is achieved
by using a Pulse-Width-Modulated (PWM) switching scheme. A control signal vcontrol is
compared with a triangular signal vtri in order to decide which of the transistors TA+ and TA-
that should be conducting. This is shown in Equation (4.1) and (4.2), and in Figure 4.4.
1
vcontrol ! vtri TA+ is on v A0 Vd (4.1)
2
1
vcontrol vtri TA- is on v A0 Vd (4.2)
2
The control signal vcontrol is a sinusoidal signal with the frequency f equal to the desired output
voltage frequency of the converter. The triangular signal has constant amplitude Vtri and
frequency fs. As can be seen from the lower part of Figure 4.4, the voltage vA0 is not a
sinusoidal wave, and the fundamental component (va0)1 has to be filtered out. Since the
voltage is of a very high frequency, the filters can be small. However, it is inevitable that
harmonics of a higher order is produced [17].
8
VSC HVDC technology
Figure 4.4: Pulse-width modulation of a single-phase converter [17]
As long as the amplitude of Vcontrol is smaller then the amplitude of Vtri, the amplitude of the
fundamental component of the output voltage (va0)1 is proportional with the amplitude of the
control signal vcontrol. This is called linear modulation. The factor ma in Equation (4.3) is
called modulation ratio.
vcontrol , peak Vd Vd
(v A0 )1, peak ma (4.3)
vtri , peak 2 2
For ma larger then 1, the output voltage (va0)1 is no longer proportional with the amplitude of
Vcontrol. This is called overmodulation, and will not be dealt with in this thesis.
9
VSC HVDC technology
4.2 Three-phase VSC
A three-phase VSC converter consists of three legs. Each leg is identical to the leg in the
converter in Figure 4.3. Figure 4.5 shows the configuration of a three-phase switch-mode
converter.
4.2.1 Switching
There is a separate control signal for each of the three legs in the three-phase converter. The
control signals are sinusoidal waves as for the one-leg converter, 120° shifted. This is shown
in Figure 4.6.
Figure 4.5: Three-phase switch-mode converter [17]
As for the one-leg switch mode converter, the fundamental component of the output AC-
voltage is dependent only of the DC-voltage Vd and the switching control signals to the
transistors as long as ma<1. The peak value of the fundamental component of the output AC-
voltage for leg A with respect to the point N from Figure 4.5 can be expressed according to
Equation (4.4).
vcontrol , peak Vd Vd
(v AN )1, peak ma (4.4)
vtri , peak 2 2
The fundamental component of the line-to-line rms voltage can be expressed according to
Equation (4.5).
3 3 vcontrol , peak
(vLL )1 (v AN )1 (vBN )1 (v AN )1, peak Vd | 0.612maVd (4.5)
2 2 2 vtri , peak
10
VSC HVDC technology
Figure 4.6: Pulse-width modulation of a three-phase converter [17]
4.3 Power flow in Voltage Source Converters
A VSC HVDC system has the advantage compared to a traditional LCC HVDC system that
the VSC converter can provide separate control of active and reactive power. This is done by
controlling the amplitude and phase angle of the produced AC-voltage, which is obtained by
changing the control signal vcontrol in Figure 4.6.
Consider a system topology as shown in Figure 4.7.
11
VSC HVDC technology
Figure 4.7: System topology of a VSC HVDC system [5]
The VSC converter is connected to an AC-system with voltage Vs. Both sides of the DC-link
have the same configuration. The resistance R and inductance L represent the impedance of
the commuting reactor, which filters out the fundamental component of the AC-voltage
produced by the converter as described in 4.1.1 [5]. The capacitor on the DC-side supports the
DC-voltage, as well as reducing harmonics on the DC-side. On the AC-side there is a shunt
filter, the purpose of this is to filter out harmonics created by the converter in the AC-system
[5].
Assume the power flowing from the AC-system into the converter, corresponding to rectifier
mode of operation for the converter. The voltage Vc lags Vs with the angle į. By neglecting
the resistance R, the power transferred to the converter from the AC-system can be derived as
in Equation (4.6) [5].

ªV VC º ªV cos(G ) jVS sin(G ) VC º
SC PC jQC VC I
VC « S » VC « S » (4.6)
¬ jZ L ¼ ¬ jZ L ¼
With the resistance neglected, the active power from the AC-system equals the received
active power at converter. The expression for the active power is found from Equation (4.6) in
Equation (4.7) by dividing the imaginary part of the numerator by jȦL, and multiply it with
VC.
ª V sin(G ) º VCVS
PS PC VC « S » sin(G ) (4.7)
¬ ZL ¼ ZL
The reactive power QC consumed by the converter is found from Equation (4.6) in Equation
(4.8), by dividing the real part of the numerator by jȦL, and multiply it with VC.
ªV cos(G ) VC º VCVS cos(G ) VC2

QC VC « S » (4.8)
¬ ZL ¼ ZL
The expression for the power delivered by the AC-system is given in Equation (4.9).
12
VSC HVDC technology

ªV VC º ªV VC cos(G ) jVC sin(G ) º
SS PS jQS VS I
VS « S » VS « S » (4.9)
¬ jZ L ¼ ¬ jZ L ¼
The reactive power drawn from the AC-system is expressed in Equation (4.10).
VS2 VSVC cos(G )

QS (4.10)
ZL
In traditional HVDC systems with LCC converters, the direction of the active power flow is
changed by changing the firing pulse to the thyristors so that the DC-voltage changes polarity,
while the direction of the DC-current remains the same [17]. To change the direction of the
active power flow in VSC converters the direction of the DC-current is changed, while the
DC-voltage polarity is the same. In this case, the rectifier becomes a inverter, and vice-versa.
Considering the system in Figure 4.7 and Equation (4.7), the power flow direction would
change if the voltage Vc was made to lead Vs.
In a traditional HVDC system, the converts consume reactive power independently of active
power flow direction. In VSC HVDC systems, the reactive power can be controlled at each
converter according to (4.8).
As can be seen from the equations in this section, the VSC converter has two controllable
variables in order to control the active and reactive power at its terminals. By applying a
phase shift į to the sinusoidal control signal Vcontrol from Figure 4.4, the output voltage will
get the same phase shift. By increasing the modulation ratio ma, the amplitude on the output
AC-voltage will increase, according to Equation (4.5).
4.4 Control modes
Different control strategies can be used in VSC HVDC. The chosen control strategies are
based on which purpose the VSC HVDC system is used for. The control strategy of an
offshore wind farm differs from the control strategy of power transfer between two strong AC
systems [2].
4.4.1 Normal operation

As described in 4.3, the active and reactive power at both converters can be controlled by
changing phase angle of the output AC-voltage į and AC voltage amplitude modulation ratio
ma. Each converter can either be set to regulate the AC voltage by adjusting the reactive
power at the connection busbar, or set to operate at constant reactive power. In addition to the
above mentioned parameters, the DC-voltage has to be controlled by one of the converters
[11]. The other converter controls the active power, based on an active power set point. In
normal operation of a HVDC VSC link, the active power from the rectifier must equal the
active power at the inverter plus the losses in the cable. Assume the rectifier has constant DC-
voltage control and the inverter has constant active power control. If the inverter increases the
voltage angle į due to a change on the active power set point, the active power from the
inverter to the AC-system will increase according to Equation (4.7). With no change in the
power transfer at the rectifier, the capacitance on the DC-side will discharge and the DC-
voltage will decrease. The rectifier has to increase the voltage angle to transfer more power to
13
VSC HVDC technology
the DC-link in order to charge the capacitances and increase the DC-voltage. Capacitances are
associated both with the DC cable and the converters [2]. The general Equation (4.11)
explains the mechanism, where vc is the voltage over the DC side capacitance, C is the total
DC capacitance and i is the current flowing through the capacitance due to imbalance of
active power between the two converters.
t
1
C ³0
vc i dt (4.11)
In a traditional HVDC system, a communication link between the two converter stations is
needed in order to control the power flow. The voltage is measured in one end, and the DC-
voltage is controlled in the other end so that the voltage difference divided by the DC-line or
cable resistance equals the set point of the DC-current [17]. A VSC HVDC system does not
require any communication between the two converters. The converters communicate through
the measured DC voltage at each end.
4.4.2 Offshore wind power

If a VSC HVDC system is used to connect offshore wind power to the grid onshore, a
different control strategy should be used than for normal operation described in 4.4.1 [2], [11].
Consider the system in Figure 4.7 assuming the bus with voltage Vs is the point of common
coupling for the offshore wind farm. The power flows into the converter, corresponding to
rectifier mode of operation. The power from the wind Pwind is a function of the cube of the
wind speed, according to Equation (4.12) [16].
1
Pwind U AU 3 (4.12)
2
ȡ [kg/m3] is the density of the air, A [m2] is the swept area of the turbine and U [m/s] is the
wind speed. The offshore converter has to absorb all the power produced by the wind turbine
generators, instead of transferring power according to an active power set point. For normal
induction generator wind turbines, the frequency at the point of common coupling tends to
increase for increasing wind speed [11]. The increase in frequency is detected by the rectifier,
and the phase angle is increased with respect to the voltage at the point of common coupling
in order transfer more power into the DC side. This action will cause the frequency to return
to the reference value. Because of the imbalance of active power between the rectifier and the
inverter, there will be a current flowing into the capacitor on the DC-side causing an increase
in DC-voltage according to the general Equation (4.11). The inverter has to increase the
voltage angle į of the output voltage with respect to the terminal AC-voltage in the connected
AC-system. The power out of the inverter will increase, and the capacitor will discharge
causing the DC-voltage to return to the reference value.
If there is no wind, or the offshore wind farm is disconnected for some reason, the onshore
converter can still function as a STATCOM [10]. Reactive power can be regulated within the
PQ diagram in Figure 5.1.
14
PSS/E models
5 PSS/E models
This chapter describes the models used in the simulations. There are several programs
available to perform stationary and dynamical analysis of power systems. In this thesis, the
options were Simpow and PSS/E. The decision was made to use PSS/E. PSS/E is owned by
Siemens [18].
First of all, Statnett, the TSO in Norway, uses PSS/E for analysis on the main grid in Norway.
They have developed a detailed model of the Norwegian power system. The model is
restricted, but available to companies and universities with confidential agreements. Secondly,
companies develop user models for power system equipment in PSS/E. ABB has developed a
model for HVDC Light in PSS/E, which is commercial available for companies with
confidential agreements.
PSS/E has a large library of models, including a model for VSC HVDC. This model is an
early version of the HVDC Light model to ABB. The dynamic response of a VSC HVDC
system is highly influenced by the control system of the converters. According to ABB [14],
the control system of a HVDC VSC is not realistic represented in this model. The dynamic
response is therefore not valid for a HVDC Light system that is on the market today. Based on
this, the decision was made to use PSS/E with the user model for HVDC Light provided by
ABB.
5.1 The Norwegian model
The PSS/E model provided by Statnett for the power system in Norway consists of 2200
buses, with voltage ranging from 3 to 400 kV. Statnett uses a model that includes the other
Nordic countries, as well as user models of the HVDC converters connected to Denmark and
the Netherlands. Unfortunately, this complete model is restricted, and could not be given out.
The isolated Norwegian model is divided into 17 different geographical areas.
There is only one official model for the Norwegian power system, and this is a peak load
situation [6]. The instant load consumption in Norway is 22931 MW, which is very high.
During the last five years, the highest hourly average load consumption was 21915 MW [20].
Since there are no connections to other countries in the model, all the power is produced in
Norway. Almost all the generators are connected, producing a total of 22640 MW. During the
last five years, the highest hourly average production was 24636 MW. The surplus of the
power in this situation had to be exported.
Limit checking reports for voltages and branches for voltage level from 130 kV to 400 kV can
be found in Appendix A.1 and A.2 for this load flow situation. Some of the voltages are
outside the range 0,9 – 1,1 pu, and some of the branches are loaded more than 100 % of the
rating. It is assumed that the load flow situation is realistic, and modifications on the system
are not done in order to get all branches within the rating and voltages from 130 to 400 kV
within the interval of 0,9 – 1,1 pu.
Statnett is not able to give out any other load flow situations than the peak load situation. This
implies that the users of the model have to scale down the load and production in the system
in order to establish load flow situations with lighter load.
15
PSS/E models
Before dynamic simulations with the model could be done, a modification had to be made.
When running a dynamic simulation without any disturbances on the grid, the system was
unstable. The initial conditions of several variables for the governor at Svartisen were out of
range after initializing the dynamic simulation. Statnett recommended to replace the governor
model NORGOV at Svartisen with the standard PSS/E model HYGOV [6]. Without any
documentation of the user models, the advices from Statnett were followed. The two models
are identical, except that NORGOV reduces the water flow in the regulated turbine if a
monitored line is over loaded. In the case of the load flow situation in the model given from
Statnett, a line in mid Norway was overloaded, and the governor response led to unstable
behavior of the entire system.
5.2 Wind turbine model
In this study a model of conventional induction generator is used to model the wind turbines.
The PSS/E model is called CIMTR3. The arguments for this are explained in this section and
in 5.3.5. Dynamic data for the wind turbine is taken from the equivalent wind turbine
generator at Smøla.
The trend of wind turbines today is utilization of some kind of frequency converter. The rotor
in induction generator turbines is often fed with a converter; the wind turbine is then called
Double Fed Induction Generator (DFIG). Another normal type of wind turbine used today is
full frequency converter and synchronous generator. Both these kinds of wind turbines are
more sophisticated than induction generator turbines directly connected to the grid. They are
able to operate at variable speed in order to maximize the energy capture from the wind [16].
Additionally, they have capability to regulate the power factor at the connection point.
Induction generator wind turbines operate at rotational speed given by the grid frequency.
They also consume reactive power [8].
An interesting aspect regarding offshore wind power connected with HVDC Light is to use as
cheap wind turbines as possible. Both the aspects with rotational speed and reactive power
can to some extent be solved with conventional induction generators connected with HVDC
Light. The offshore converter can adjust the frequency of the entire grid offshore in order to
adjust the speed of the wind turbines. Additionally, the converter can supply the reactive
power needed for the wind turbines, as long as the converter operates within the PQ diagram
in Figure 5.1.
In this study it is assumed that the wind turbines are compensated for reactive power
corresponding to cos(ĳ) = 0,99 at the busbar where the wind turbines are connected offshore.
This power factor corresponds to the wind farm at Smøla in the PSS/E model. In the
simulations, one equivalent machine is used to model the entire wind farm [21]. This is also
done at Smøla, and is a normal simplification for dynamic power system simulations.
16
PSS/E models
5.3 The HVDC Light model
The user model of HVDC Light is developed and provided by ABB. This paragraph will
explain the use of the model in PSS/E. A user manual is provided, with sufficient information
to set up and use the model [2]. However, the description of the control system is not
available, as the user model files were provided as compiled files.
5.3.1 Modules
HVDC Light is a module based technology, with power rating from 98,7 MW to 1174,6 MW
[1]. To increase the power of the converters, modules are placed in series and parallel in order
to increase the current and voltage. The current and voltage rating of the different modules are
shown in Table 5.1.
Table 5.1: Modules in HVDC Light [1]
5.3.2 Power flow representation

In PSS/E, the HVDC Light system has to be represented in the load flow before dynamical
simulations can be done. Standard generator elements from PSS/E are used in order to
establish the load flow. The rectifier is modeled as a PU bus connected to a generator and
shunt. Because the power is going from the AC to the DC side, the specified power to the
generator has to be negative. The generator impedance Zsource in PSS/E specifies the
commuting reactance L in Figure 4.7. The resistive part R is equal to zero. The shunt AC
filter in Figure 4.7 is represented by the reactive power generation of the AC filter capacitors
in PSS/E. Additionally, a converter transformer has to be added. One side is connected to the
Point of Common Coupling (PCC), and the other side is connected to the PU bus with the
generator and AC filter. The voltage at the converter busbar is 408 kV for the largest HVDC
Light converter.
The inverter is described in the load flow with the same units as the rectifier, but the power
from the generator has to be positive, because the power is going from the DC side into the
AC side. The reactive power limits to be specified in the power flow for the generator
equivalents depend on the active power. The HVDC Light converter has a capability curve
according to Figure 5.1. It is up to the user to select the right reactive power limits for the
converter.
17
PSS/E models
Figure 5.1: PQ diagram of a HVDC Light converter [1]
5.3.3 Losses
The DC cable is not modeled in the load flow. In dynamic simulations, a user model
communicates between the two converters, and includes the cable resistance. The power
combination has to be defined by the user between the rectifier and inverter so that the power
going into the rectifier is larger than the power going out from the inverter. Equation (5.1)
describes the power loss in the whole DC system.
Ploss Prectifier Pinverter (5.1)
The power loss in the DC system consists of converter loss, and cable loss.
From Table 5.2 the converter losses at full power can be calculated by comparing the sending
power with the receiving power when the system is has a back-to-back configuration. In this
case, the inverter and the rectifier are at the same location.
Table 5.2: Losses in HVDC Light [1]
18
PSS/E models
For the largest converter named M9, the power loss in percentage for each converter at full
power can be described according to Equation (5.2).
Psending Preceiving 1147.9 1110.1

Converter loss, per converter | 1.65% (5.2)
2 Psending 2 1147.9
In percentage, the converter losses for the modules M7 to M9 are identical.
The converter MW losses can be assumed to be independent of the power through the HVDC
Light system, since a large part of the losses are caused by the switching of the IGBT’s [14].
The cable loss can be calculated by finding the DC current, given a certain cable dimension
and length. In [1], an overview of resistance for different cables is presented. In this thesis,
data for the largest cable is used, with a copper conductor area of 3000 mm2. The resistance is
0,0059 ȍ per km. With a 600 km distance to the offshore wind farm, the total resistance of the
cable will be 7,08 ȍ. For 1000 MW produced offshore wind power, the received power is
found by using Equation (5.3) to (5.5). The DC current in Equation (5.3) is found by dividing
the sending power on DC side Psending,DC for the rectifier by the DC voltage multiplied by 2.
Converter loss for one converter is subtracted from the produced wind power in order to find
the sending DC power. The DC voltage has to be multiplied by 2 because the given DC
voltage is pole to ground.
(1147.9 1110.1) 6
Psending , DC 1000 106 10
I dc 2 | 1635 A (5.3)
2 Vdc 2 300 103
Ploss ,cable Rdc I dc2 7.08 16352 | 19 MW (5.4)
For 1000 MW offshore wind power, the received power is found in Equation (5.5).
Preceived Pwind Ploss ,converters Ploss ,cable 1000 37.8 19 943.2 MW (5.5)
Ploss,converters are the total converter loss for both the converters. In the simulations received
power of 940 MW is assumed.
5.3.4 Dynamic model

In order to represent the dynamics behavior of the HVDC Light system, ABB has developed
two user models called CHVDCL and DC_HL2 [2]. The user model CHVDCL represents the
dynamics of the converters, and calculates the current injection to be applied by the standard
generator models used to model the converters in the load flow.
The model DC_HL2 represents the dynamic behavior of the DC system. Both the cable and
the converters have capacitances with corresponding time constants T according to Equation
(5.6).
19
PSS/E models
1 C U 2 2
C U dN
2 2
dN
T (5.6)
PN PN
T is the charging time of the capacitance [s], C is the capacitance pole to ground [F], UdN is
the nominal DC voltage, pole to neutral [kV], and PN is the nominal active power [MW]. The
model is based on simple integrator blocks according to Equation (5.7) [2].
1
Integrator block= (5.7)
sT
In the dynamic description file of the model, a default value of a 100 km cable is given. The
user manual doesn’t mention that the time constant should be changed for longer cables, but it
is assumed that the time constant is 6 times larger for a 600 km cable. According to Equation
(5.6), the time constant is proportional with the capacitance of the cable, which is proportional
with the length of the cable. The time constant for the converters remains unchanged for
increased cable length.
The DC voltage at the converter regulating the DC voltage is calculated as the integration of
the net injected current via the integrator. The net current is the difference in active power
injection by the converters, minus losses, divided by the DC voltage [2]. For the other
converter, the DC voltage is calculated by subtracting or adding the voltage drop due to the
resistivity in the system.
Figure 5.2 shows the structure of the load flow and dynamic model of HVDC Light.
Figure 5.2: HVDC Light model in PSS/E [2]
For each of the modules described in 5.3.1, a default dynamic description file is provided. The
control mode for the model as described in 4.4 is given as integer constants in the dynamic
description file.
20
PSS/E models
5.3.5 Offshore wind power application

The characteristics of HVDC Light make the technology well suited for transmitting power
from offshore wind farms [1]. One main advantage is the controllability of the converters. A
HVDC Light system can supply power to a passive network at a desired frequency. There are
no requirements regarding short-circuit power at the connected AC system [1]. However,
there are some limitations in the control system in the current version of the HVDC Light
model in PSS/E regarding offshore wind power applications and transient studies. According
to ABB, this will be implemented in the next version of the model [2]. In this paragraph, these
limitations are described, and approximations done in the dynamic simulations are explained.
It has to be emphasized that the limitations only apply to the current HVDC Light model in
PSS/E, and has nothing to do with the actual control system of HVDC Light.
The trend for wind turbines is that they are able to operate at variable speed in order to
maximize the power output from the wind speed by keeping the tip speed ratio constant [16].
DFIG turbines or full frequency converter turbines with synchronous generators are the most
common types for variable speed wind turbines. The consequence of these technologies is that
the local AC system offshore can not be considered having rotating masses that can be used to
store power temporarily. The frequency converter on the turbine side defines the frequency on
the turbine side; hence also the synchronous speed of the turbine. The frequency converter on
the grid side adopts the grid frequency. There is a decoupling between the grid frequency and
the rotational speed of the generators. On the other side, with normal induction generators
directly connected to the grid, the synchronous speed is directly given from the frequency of
the grid. The rotational speed of the turbines is close, but not equal to the synchronous speed,
as inductions generators operate with a slip when power is produced. Slip is the difference
between the rotational speed and synchronous speed, caused by mechanical torque at the
turbine [8]. For normal operation, the slip is in the range of about 1 %. The rotational speed of
the wind turbines will be about 1 % larger than the synchronous speed.
If the onshore voltage Uac is reduced due to a fault, the power transmission of the HVDC
Light has to be reduced. The maximum power transferred from the onshore converter to the
AC system is limited, according to Equation (5.8).
S 3 U ac I (5.8)
With HVDC Light, the AC voltage offshore is kept almost constant for a fault onshore. This
will be shown in the simulations in chapter 7. The wind turbines can continue to produce
power as before the fault onshore occurred. However, because of the limitation in the power
transfer for the onshore converter during the fault, the wind power fed into the DC system
offshore has to be reduced in order to avoid the DC voltage to increase. This can be done in
two ways; either increase the wind turbine speed, or burn the excess produced power in a DC
chopper during the fault [14]. DC choppers are planned for future offshore wind power
applications with HVDC Light. With a DC chopper, the offshore wind farm will not see the
fault onshore, and can continue to produce the same amount of power. This is not modeled in
the current model of HVDC Light. In the simulations done in this thesis, the rotational speed
is increased in order to store the energy offshore during an onshore fault instead of feeding the
produced power into the DC system. Because of the decoupling of grid frequency and
rotational speed for frequency converter wind turbines, normal induction generator wind
turbines are used in the simulations in this study.
21
PSS/E models
If the offshore converter is in passive control mode, it is set to transfer all the power produced
offshore; it has the function of a swing bus offshore. This is described in 4.4.2. The AC
amplitude and frequency is constant at the defined value. For a fault onshore, the control
system for passive mode does not change the amplitude of the AC voltage or the frequency
offshore. The wind turbines produce power at the same level as before the fault, causing the
DC voltage to increase drastically during the fault onshore.
As an approximation to avoid this, the control system can be changed to normal operation
during a fault onshore. This method is used in the simulations. The offshore converter is in
Udc control mode, while the onshore converter is in active power control mode. When the
fault occurs onshore, the DC voltage will increase immediately because of the power
imbalance between the two converters. The power fed into the DC system from the offshore
converter is quickly reduced as a response to this in order to reduce the DC voltage. The
rotational speed of the wind turbines and frequency offshore and will increase. The energy
from the wind is used to accelerate the turbines as long as the fault onshore is too low for the
onshore converter to transfer the power produced by the wind farm.
In the simulation done is this study, it is assumed that the wind speed is constant, and hence
also the mechanical torque on the wind turbines. Because of this assumption, the control
system is set to normal operation during the whole time period in the simulations. If variable
wind speed is to be used as input to the wind turbines, the control system has to be changed
manually from passive mode to normal control mode at the instant the fault onshore occurs.
After the fault onshore is cleared and the voltage has recovered, the control system has to be
changed back to passive mode in order for the offshore converter to function as a swing bus.
The control system can be changed by changing integer constants for the CHVDCL models
for the converters in PSS/E [2].
The wind turbines operate at full power, equal to 1000 MW in total for the offshore wind farm.
It is therefore assumed that the wind turbines have some temporary over load capacity, in
order to transfer the increased rotational power to the onshore grid after the onshore fault is
cleared.
5.3.6 Time step

For the dynamic simulations with HVDC Light, the time step has to be reduced compared to
simulations of normal AC systems. A time step of 10 ms is normally used for AC systems in
PSS/E, while for simulation with HVDC Light a time step of 5 ms will give a good
representation of the dynamic response close to the HVDC Light converters [2]. The reason
for this is the small time constants descried in 5.3.4 associated with HVDC systems.
22
Load flow cases
6 Load flow cases

Before dynamic simulations can be done, a load flow case has to be established as a starting
point. This chapter describes the different load flow situations established for the simulations
done. Two possible connection points of 1000 MW offshore wind power is chosen; Feda and
Kårstø. In order to investigate the transient stability in the two connection points for different
load flow situations as well as reduced short-circuit power, cases with lower load and
production in the system are established.
6.1 Connection point Feda
The case to be studied in this thesis is grid connection of a 1000 MW offshore wind farm in
the North Sea. The wind farm is to be connected to the south of Norway. In order to connect
large amounts of offshore wind to the grid, the transfer capability from the connection point
has to be investigated. One of the connection points for offshore wind power in this thesis is
Feda. This is the only busbar in the very south of Norway close to the sea that has enough
thermal capacity to transfer 1000 MW offshore wind power to the load centers.
In the south of Norway, the highest voltage is 300 kV. The next voltage level is 132 kV. The
132 kV lines in the Norwegian model typically have thermal limits of a couple of hundred
MVA. A 300 kV line can have thermal limit of more than 1000 MW. The voltage at Feda is
300 kV, and there are several lines going out from the busbar. Figure 6.1 shows a section of
the 300 kV grid in the south of Norway, with lines going from Stavanger in the west, Arendal
in the east, Lista in the south and Lyse in the north. Feda is the connection point of the
offshore wind power, marked in red with a circle around. The sum of the capacity of the lines
connected to from Feda is 4710 MVA. This does not indicate how much wind power that can
be fed into the busbar, but indicates that the point is a strong. By performing a short circuit
analysis according to Equation (2.3) in PSS/E on the busbar Feda, the short-circuit capacity is
7147 MVA. As described in 2.4, this is a measure of the strength on the busbar.
23
Load flow cases
Figure 6.1: Overview of the 300 kV power grid in the south of Norway
A screenshot from the original load flow situation described in 5.1 is presented in Figure 6.2.
The number above the line is active power, and the number below the line is reactive power.
Power is defined positive out from the busbar. The boxes on the lines indicate the percentage
of the maximum current loading in the lines. The voltage is given in pu under the busbars. As
described in 5.1, some of the busbar voltages are outside the range of 0,9 – 1,1 pu.
Additionally, some branches are over loaded. It is assumed that the load flow situation is
realistic with these violations. The limit checking reports for the original load flow case is can
be found in Appendix A.1 and A.2.
24
Load flow cases
Figure 6.2: Load flow result at Feda, original load flow
In Figure 6.3 a screenshot of the load flow for the same situation is presented, with 1000 MW
power added to Feda. The offshore wind power is here simply modeled as a voltage
regulating generator connected directly to Feda in order to show the direction of the power in
the surrounding transmission lines. For the dynamic simulations, the offshore wind power will
be modeled differently.
When additional wind power is added to the system, the swing bus adjusts the production in
order to maintain the balance between produced and consumed power in the system. With
1000 MW infeed of offshore wind power, the production at the swing bus becomes negative.
One of the main connections out from Norway is the line from Hasle to Borgvik in Sweden.
All the power imported from the Netherlands and Denmark is assumed exported out from
Norway through this line. A load of 900 MW is added on the 400 kV busbar Hasle. It is
assumed that only active power is transferred to Sweden, even though in a real situation some
reactive power would flow in the line.
25
Load flow cases
Figure 6.3: Load flow result with 1000 offshore wind, connection point Feda
Limit checking reports for voltages and branches for voltages from 132 kV to 400 kV are
made in order to make sure that voltages are within the limits and branches are not over
loaded with full power from the wind farm. The limit checking reports can be found in
Appendix A.3 and A.4. The limit checking reports for the original load flow case in Appendix
A.1 and A.2 are used as references.
6.1.1 Feda with HVDC from the Netherlands and Denmark

Feda is also the connection point for the HVDC cable from the Netherlands, NorNed. This
HVDC system is a traditional system based on thyristors, and the rating of the converter is
700 MW. There are also a total of 1000 MW HVDC links to Denmark. These converters are
placed in Kristiansand. In order to supply the HVDC converters with reactive power, SVC’s
are installed in Kristiansand. Additionally, a synchronous condenser is installed at
Kristiansand in order to increase the short-circuit power at the connection point.
Since the model for the Norwegian power system provided by Statnett only contains data for
Norway isolated, assumptions have to be done in order to estimate the power flow in a real
situation. The HVDC converters importing power from the Netherlands and Denmark can be
modeled as loads with negative active power. Since the converters consume reactive power
corresponding to about 50 % of the active power [17], the complex parts of the loads are
modeled as positive loads.
When additional power is added to the system, the swing bus adjusts the production in order
to maintain the balance between produced and consumed power in the system. With 1700
MW infeed from the Netherlands and Denmark, the production at the swing bus becomes
negative. One of the main connections out from Norway is the line from Hasle to Borgvik in
26
Load flow cases
Sweden. All the power imported from the Netherlands and Denmark is assumed exported out
from Norway through this line. A load of 1600 MW is added on the 400 kV busbar Hasle. It is
assumed that only active power is transferred to Sweden, even though in a real situation some
reactive power would flow in the line.
A screenshot from the original load flow with HVDC added is presented in Figure 6.4.
Figure 6.4: Load flow result at Feda with full import of HVDC
The voltages in the screenshot is within the values of 0,9 – 1,1 pu, but voltages on busbars
other places in the system are too low compared to the original load flow situation with out
any wind power. Additional branches are also slightly over loaded. The 400 kV transmission
link from Teige to Bastø, which is one of the main transmission links from the west to the east,
is at the limit of being over loaded. The limit checking reports for voltages and branches for
voltage from 132 kV to 400 kV can be found in Appendix A.5 and A.6. When adding wind
power to Feda, and increasing the load at Hasle, eventually the load flow will not be solved.
This is assumed to be an unrealistic load flow situation in the Norwegian system. With this
high power production in the system, power will not be imported to Norway from the HVDC
links.
27
Load flow cases
6.2 Connection point Kårstø
Kårstø could also be a potential connection point of 1000 MW offshore wind power. Kårstø is
a gas processing plant in the south west of Norway, north of Stavanger. The voltage is 300 kV,
and the consumption of power from industry in the area is high. The large consumption of
power by the industry nearby will lead to less transmission losses for the system.
Figure 6.5 shows a section of the 300 (blue) and 400 kV (orange) grid in the south-west of
Norway, with lines going from Karmøy in the west, Vemork and Kvilldal in the east, Blåfall
in the north and Saurdal in the south. Kårstø is the connection point of the offshore wind
power, marked in red with a circle around. The short-circuit power at Kårstø is 3986 MVA,
and the sum of capacity of the connected lines is 1717 MVA. This indicates that Kårstø is a
considerable weaker point than Feda.
Figure 6.5: Overview of the 300 kV grid in the southwest of Norway
A screenshot from the original load flow for a peak load situation is presented in Figure 6.6.
28
Load flow cases
Figure 6.6: Load flow result at Kårstø, original load flow
As can be seen from the load flow screenshot, a large amount of power is flowing to the
busbar Haavik. There is a large amount of industry demanding about 500 MW at the neigbour
busbar.
In Figure 6.7 a screenshot of the load flow for the same situation is presented, with 1000 MW
power added to Kårstø. As in the case for Feda, the offshore wind power is modeled as a
voltage regulating generator connected directly to Kårstø in order to investigate the direction
of the power in the surrounding transmission lines. A 900 MW load is added at Hasle in order
to simulate export of power.
The offshore wind power will be modeled differently in the simulations. As can be seen by
comparing Figure 6.6 and Figure 6.7, the direction of the power flow in the transmission line
from Kårstø to Sauda is changed. The line from Sauda to Haavik transfer 300 MW less power
with the offshore wind power added to Kårstø.
29
Load flow cases
Figure 6.7: Load flow result with 1000 MW offshore wind, connection point Kårstø
Limit checking reports for voltages and branches for voltages from 132 kV to 400 kV are
made in order to make sure that voltages are within the limits and branches are not over
loaded with full power from the wind farm. The limit checking reports can be found in
Appendix A.7 and A.8.
30
Load flow cases
6.3 Background for load flow situations
The load flow limit checking reports made in 6.1.1 indicate that the power system is not able
to import wind power in addition to rated power from the HVDC link from the Netherlands
and Denmark in a peak load situation in Norway with full production. With additional
offshore wind power fed into the same busbar as the HVDC power from the Netherlands, the
main transmission lines are even more over loaded, and voltages in the system become too
low. The load flow situations are not realistic. With large enough production from the
Norwegian power plants to cover the peak load, additional power will not be imported from
the HVDC links.
When the production in the system is planned, the cheapest producing unit will be set to
produce power first [22]. Even though the water in hydro power plants is free, water values
are used as a tool for marginal cost of production. Every hydro reservoir has its own water
value, based on the water level in the reservoir, the snow in the mountains, expected
precipitation based on weather forecasting, among other things. If the water value for a hydro
power plant is higher than the spot price, the power plant will store the water for later in order
to maximize the profit for the power producing company. If the water value is lower, the
power plant will produce power. When a reservoir is at the maximum level and it is expected
that more water will come into the reservoir, the value of saving the water becomes equal to
zero. The power plant will then produce as much power as possible. Hydro power plants with
small or no reservoirs, typically have low water values. The saved water will not be possible
to save for later; hence it is better to produce at maximum power and get a low spot price.
Some hydro power plants have large reservoirs that are capable of saving the water from year
to year.
The energy cost for wind power is higher than the hydro power in Norway, because of the
capital cost of the wind farms. Nevertheless, the marginal cost of producing wind power is
practically equal to zero. The wind is free, and the wind farm produce power when the wind
blows with no ability to save the wind for later periods with higher spot price. If a wind farm
does not produce the maximum power for the given wind speed, the production is lost. Wind
power will always be prioritized before hydro power, unless the hydro power plant does not
have the capability to store to water in a reservoir.
If the production in Norway is as high as in the original load flow situation, and the offshore
wind farm produce additionally 1000 MW, the HVDC links will probably export power to the
Netherlands and Denmark or shut down the unit with highest water value in order to
maximize profit. The situation with full HVDC import in the original load flow situation is
assumed not to be realistic.
During winter time the spot prices can be quite high because of low water levels in the
reservoirs, and high power consumption. The HVDC links will import power because the
electricity produced in thermal power plants on the continent is cheaper. Even though the
offshore wind farm produces full power, this does not necessarily reduce the spot price
enough so that the import is stopped. The most realistic case with full import of HVDC and
full power from the offshore wind farm is therefore a peak load situation with lower
production of hydro power than in the original load flow case.
31
Load flow cases
A realistic case could be full import of the HVDC link from Denmark, and no import from the
Netherlands. The Danish power system is dependent of exporting wind power to Norway in
periods with high production of wind. If the production of the offshore wind farm in the North
Sea is at maximum, it is expected that the wind is blowing in Denmark as well. With wind
power fed into Feda instead of HVDC, reactive power is regulated at Feda instead of
consumed by the HVDC converter. The regulation of reactive power for wind power will be
explained in 6.5. Even though the limit checking report from the case with full HVDC import
showed somewhat more over loading of lines in addition to voltages outside the range of 0,9 –
1,1 pu, simulations are done for this load flow situation.
Three different load flow situations for 1000 MW offshore wind production connected to
Feda are chosen for dynamic simulations. Additionally, two different load flow situations are
chosen for Kårstø as connection point. An overview of the different load flow cases are
defined in Table 6.1. Production and load consumption is defined as “original” for the
production and load consumption in the original load flow case provided by Statnett. The load
flow cases are described in detail in detail in 6.4.
Table 6.1: Load flow situations
It has to be emphasized that the power flow situations chosen are supposed to reflect the worst
case scenarios in the sense that no power is exported through the HVDC links to the
Netherlands and Denmark. With high power production in the Norway, it is likely that the
power is exported to the Netherlands and Denmark. For the connection point at Feda, the load
flow changes in the rest of the system will be less affected, since a lot of the offshore wind
power is exported directly from the same busbar.
6.4 Establishing of the load flow cases
The Norwegian model provided by Statnett was a case with peak load in the system and
production corresponding to the load consumption. In order to do dynamic simulations based
on different load flow situations, the load flow had to be changed manually.
In the load flow environment, PSS/E has a function that scales the production and load in
defined percentage or MW steps. However, in large systems drastic changes in the power
production and load consumption become difficult to perform. Even though both the
production and load is scaled down in small steps in the system, at a certain point the load
flow will not be solved without a very large MVA mismatch. Beyond this point, the mismatch
increases drastically for even a small change in the system. All the power plants are set to a
specific voltage set point, corresponding to the load flow situation. When the production and
load is changed, the set points of the voltage in all the plants and reactive power compensating
32
Load flow cases
units should be changed according to the new load flow. Without any information of what the
new set points should be, this becomes a random operation that would not improve the
solution of the load flow.
Because of the above mentioned problems, an assumption had to be made in establishing a

load flow for situations with low load consumption and low production, and peak load and
low production. Only the areas surrounding the area of interest were scaled down in order to
be able to solve the load flow. Additionally, assumptions have to be made regarding which
power plants should be disconnected during low production in the system. It is assumed that
the largest power plants will be disconnected first, as they are assumed to have the highest
water values because of storage capacity in the reservoirs. This is an assumption made in this
thesis.
Because the production and load was only scaled down in some parts of the south of Norway,
the load and production of the entire system is not changed much in percentage. The power
flow in the large transmission lines will not be realistic according to the real situation with
load and production levels as it would have been in a summer situation with low load in the
entire system. In order to do realistic simulation in these cases, load flow situation for the
entire system that represents a real situation should be provided. Additional agreements with
Statnett would have to be done in order to get access to these files.
In this paragraph, the procedure of establishing the different load flow cases described in
Table 6.1 is given. The description names used below for the load flow cases will be used to
describe the starting point of the different dynamic simulations performed.
6.4.1 Case 1
The load flow case called Case 1 is the original load flow situation provided from Statnett,
with the offshore wind power connected to Feda. The only change in the power flow is adding
the wind power, and a 900 MW load at Hasle in order to simulate export of power to Sweden.
In Case 1, the short-circuit power at Feda is 7147 MVA. The short-circuit power was found
without the offshore wind power connected, and the results from the short-circuit power
calculation from PSS/E can be found in Appendix B.
6.4.2 Case 2
The load flow case called Case 2 represents a load flow situation with low load and low
production in the system. Feda is the connection point of the offshore wind power. There is
export of power to Sweden. As described above, changes are only made in areas surrounding
the connection point. The consequence of the disconnection of some generators is that the
short-circuit power at Feda is reduced to 5228 MVA, from 7147 MVA in the original
configuration. The short-circuit power was found without the offshore wind power connected,
and the results from the short-circuit power calculation from PSS/E can be found in Appendix
B.
The detailed procedure in establishing the load flow is described in Appendix C. The load is
scaled down with a fixed relationship between P and Q on all busbars in the selected
subsystem, except the busbars with name starting on KKI and IND. These busbars represent
industry, and the power consumption is assumed to be kept constant.
33
Load flow cases
6.4.3 Case 3
The load flow case called Case 3 represents a situation with peak load in the system and
import of 1000 MW HVDC from Denmark. The HVDC converter at Kristiansand is modeled
as a load equal to -1000 + j500 MVA, corresponding to the approximate consumption of
reactive power in the converter. Feda is the connection point of the offshore wind power.
There is no export of power to Sweden, so that the production corresponds to the load minus
the offshore wind power and imported power from Denmark through the HVDC connection.
No generators are disconnected; the production on the largest generators in the south of
Norway is scaled down. The short-circuit power at Feda is 6900 MVA. The short-circuit
power was found without the offshore wind power connected, and the results from the short-
circuit power calculation from PSS/E can be found in Appendix B.
The detailed procedure in establishing the load flow is described in Appendix C.
6.4.4 Case 4
The load flow case called Case 4 is the original load flow situation provided from Statnett,
with the offshore wind power connected to Kårstø. The only change in the power flow is
adding the wind power, and a 900 MW load at Hasle in order to simulate export of power to
Sweden. The short-circuit power at Kårstø is 3986 MVA. The short-circuit power was found
without the offshore wind power connected, and the results from the short-circuit power
calculation from PSS/E can be found in Appendix B.
6.4.5 Case 5
The load flow case called Case 5 represents a situation with peak load in the system. The
offshore wind power is connected to Kårstø. The load is unchanged from the original load
flow situation, but the production is scaled down in the surrounding areas, by disconnecting
the largest generators. The short-circuit power at Kårstø in this case is reduced to 3556 MVA
from 3986 MVA in the original case. The short-circuit power was found without the offshore
wind power connected, and the results from the short-circuit power calculation from PSS/E
can be found in Appendix B. There is no export of power to Sweden.
The detailed procedure in establishing the load flow is described in Appendix C.
6.5 Adding the offshore wind power
In order to do large changes in the load flow, care has to be taken when solving the case. New
busbars added to the system have to be given the same voltage amplitude and angle as the
connection busbar. The power has to be ramped up or down in small steps. As described in
5.2, one equivalent machine represents the entire wind farm. The equivalent generator is
connected to a 20 kV busbar, and the voltage is transformed up to the transmission voltage in
an offshore transformer.
34
Load flow cases
6.5.1 AC cable connection

In all the dynamic simulations performed, a case with AC cable connection is compared to
HVDC Light connection. Transmission with traditional HVDC is not modeled. Statnett has
requirements regarding the power factor of wind farms connected to the grid. All wind farms
larger than 10 MW connected to the regional or central grid, has to be able to operate at a
power factor cos(ĳ) = 0,95 capacitive and inductive at rated power [19]. Up until recently, the
requirement was even stricter, with cos(ĳ) = 0,91.
Transmission with AC cable will require large amounts of reactive power compensation in
order to fulfill Statnett’s requirement of cos(ĳ) = 0,95 capacitive and inductive. In this thesis a
distance of 100 km is assumed with AC cable connection. It has to be emphasized that this
distance to the offshore wind farm is not realistic with today’s technology, because the depths
outside the Norwegian coast would require floating wind turbines at this distance from shore.
In order to transfer 1000 MW of power, two sets of AC cables have to be installed. The data
used for the cables is provided by Nexans. The per phase series impedance is 0,0604 +
j0,1031 ȍ/km, and the shunt capacitance is 0,13 ȝF/km [13]. The copper conductor area is
500 mm2, and two sets of cables have to be used in order to transfer 1000 MW. Cables of this
size are always custom made, so the values for the impedances are typical values [13]. With
the given cable data, the two sets of cables produce a total of 520 Mvar at the connection
point at voltage equal to 1 pu at rated power from the wind farm. This reactive power could be
compensated for with shunt reactors. In the simulations, a shunt reactor at the connection
point corresponding to 500 Mvar reactive power consumption at nominal AC voltage is
installed. The active power received onshore at 1000 MW produced by the wind farm and 100
km AC cable is about 960 MW. In order to operate at cos(ĳ) = 0,95, the necessary capacity
from SVC can be found according to Equation (6.1).
QSVC P tan(cos 1 (M )) 960 tan(cos 1 (0.95)) | 316Mvar (6.1)
Earlier, with cos(ĳ) = 0,91 the necessary capacity from the SVC would have been 437 Mvar,
if all the reactive power produced by the AC cable was compensated for by a shunt reactor.
Two SVCs with a rating of 200 MVA each at the connection point are chosen for simulations
with AC cable transmission. In the model for the Norwegian power system, there is a user
model for SVC called NORSVC, and the dynamic parameters for this model are used in the
simulations. The voltage set point for the SVCs is 1 pu for all the load flow cases.
Figure 6.8 shows the layout of the equivalent offshore wind farm connected to the onshore
grid with AC cables. The green busbar is the 20 kV busbar the wind farm equivalent machine
is connected to, and the blue busbar to the left is the 300 kV connection point onshore. The
two machines onshore represent the SVCs, and the coil is the shunt reactor compensating for
the reactive power produced by the two sets of AC cables.
35
Load flow cases
Figure 6.8: Equivalent offshore wind farm connected with AC cables
6.5.2 HVDC Light connection

For transmission with HVDC Light of a 1000 MW wind farm, the requirement from Statnett
is fulfilled at the connection point without any auxiliary compensating equipment. This can be
verified by inspecting the capability curve of HVDC Light in Figure 5.1. The topology
offshore, as well as the length and dimension of the DC cable will influence the losses and
hence received power onshore. In this thesis a DC cable length of 600 km is used. As
described in 5.3.3, with these assumptions the received power onshore equals about 940 MW,
which corresponds to 0.79 pu active power in Figure 5.1. With this active power, the
converter onshore can produce about 0,375 pu, or 447 Mvar reactive power, and consume 0,5
pu or 596,5 Mvar. This corresponds to about cos(ĳ) = 0,90 capacitive and cos(ĳ) = 0,84
inductive at the connection point onshore. In order to utilize the HVDC Light system maximal
regarding the cos(ĳ) requirement, even more offshore wind power could be installed with the
largest HVDC Light converter.
Figure 6.9 shows the layout of the equivalent offshore wind farm connected to the onshore
grid with HVDC Light. The green busbar is the 20 kV busbar the wind farm equivalent
machine is connected to, and the blue busbar to the left is the 300 kV connection point
onshore. The orange busbars are the converter busbars, with a nominal voltage of 408 kV.
The shunt capacitors represent the AC filters, and the machines represent the HVDC Light
converters.
Figure 6.9: Equivalent offshore wind farm connected with AC cables
36
Dynamic simulations
7 Dynamic simulations
In this chapter dynamic simulations are performed in order to investigate the dynamic
behavior of 1000 MW offshore wind power connected to the grid with HVDC Light. As a
comparison, an identical case with AC cable transmission is established for each simulation
with HVDC Light. In order to fulfill the requirements regarding reactive power control at the
connection point, SVCs are added for the cases with AC cable connection.
Simulations with the two different transmission technologies are done for all five load flow
cases described in 6.4. Three events are defined in order to investigate the dynamic behavior
of the offshore wind farm connected to the grid. All the simulation plots are presented in
Appendix D. Some of the plots from the most significant cases are also given in this chapter
in order to highlight the most significant characteristics.
7.1 Grid code
All power plants, including wind turbines, connected to the grid have requirements regarding
behavior during disturbances in the power system. This is normally referred to as fault ride
through [19]. The power plants shall contribute to the short-circuit power during faults, and
maintain connected after the fault is cleared. Statnett has requirements for all power plants in
Norway connected to directly grounded grid with nominal voltage greater than 200 kV. The
power plant has to stay connected and deliver power when the voltage at the connection point
is above the voltage profile described as follows:
x Voltage reduction to 0 % for up to 150 ms, followed by voltage increase to 25 %

x Linear voltage increase from 25 % to 90 % for 750 ms
x Constant voltage at 90 %
The requirement is shown graphically in Figure 7.1.
Figure 7.1: Fault ride through requirement for power plants [19]
37
Dynamic simulations
Power plants connected to busbars with voltage equal or greater to 200 kV generally have a
large MVA rating. The dynamic behavior of these power plants will have significant
influence on the dynamic behavior of the system. This is the reason why the power plants are
required to stay connected even though the voltage at the connection point is zero for 150 ms,
as the consequences of tripping of such power plant normally will have undesirable effects on
the power system. The fault ride through requirement for power plants connected to voltage
level equal or smaller to 132 kV are allowed to trip if the voltage falls below 0,15 pu [19].
For offshore wind power, the transmission technology will have great influence on the
dynamic behavior of the system. This will be shown in the dynamic simulations done in this
chapter. Protection of the wind turbines is not modeled; it is assumed that the wind turbines
will stay connected as long as the voltage at the connection point onshore is within the profile
in Figure 7.1. The focus of the simulations will be the impact of the transmission technology
on the onshore power system.
7.2 Conversion of loads
Before doing dynamic simulations or fault analyses, the generators and loads have to be
converted to Norton equivalents [18]. The loads can be converted as constant current, constant
admittance or constant power loads. All the loads in the systems will have slightly different
characteristics; hence a few assumptions have to be made in the converting process. Loads in
the Norwegian model are assumed divided into two groups; normal loads and industrial loads.
Industrial loads are recognized in the model by the connected busbar name starting with IND
(industri) or KKI (kraftkrevende industri). The conversion is based on recommendations from
Statnett. Table 7.1 describes how the loads are converted.
Table 7.1: Load conversion
7.3 Definition of disturbances
In this study, the focus is on transient stability. In [12], transient stability is defined as follows:
“Transient stability is the ability of the power system to maintain synchronism when subjected
to a severe transient disturbance… …The contingencies usually considered are short-circuits
of different types.”
The case with voltage equal to zero at the connection point according to the grid code
presented in Figure 7.1 is a worst case scenario. Only a perfect three-phase short-circuit at the
connection point will result in this. In order to investigate the dynamic behavior of HVDC
Light for other situations, simulations are also done with events leading to a smaller voltage
reduction at the connection point.
38
Dynamic simulations
Simulations are done for all the 5 load flow cases described in 6.4; three cases with Feda as
connection point and two cases with Kårstø as connection point. The three events are
simulated for all the load flow cases, with both AC cable and HVDC Light connection of the
offshore wind power. Totally 30 simulations are done, and all the plots can be found in
Appendix D. The description names of the events will be used to describe the dynamic
simulations performed, along with the names of the load flow cases described in 6.4.
7.3.1 Description of Event 1

In Event 1, a fault at a neigbour busbar of the connection point is applied, with shunt
impedance to ground larger than zero. In PSS/E, shunt resistance and reactance are given in
the MW and MVA power for the actual voltage at the busbar. In this event, a power of -
2E+004 Mvar is added to simulate the fault. For nominal voltage equal to 300 kV at the
busbar, the shunt reactance added can be found by Equation (7.1).
U ac2 U ac2 (300 103 ) 2

Q X shunt 4.5: (7.1)
X shunt Q 2 10(4 6)
The bus fault is removed after 150 ms. For connection at Feda the fault appears at Øie, while
for connection at Kårstø the fault appears at Sauda. The busbars can be recognized by
inspection of Figure 6.1 and Figure 6.5.

In Event 2, a three-phase short-circuit at a line going from the connection point to a neighbour
busbar is applied. The short-circuit is simulated with impedance equal to zero, close to the
neigbour busbar. The fault is cleared after 100 ms, and the line is disconnected.
For connection point at Feda, the line with short-circuit is one of the lines going from Tonstad
to Feda. For connection point at Kårstø, the faulted line is the line going from Haavik to
Kårstø. The lines can be recognized by inspection of Figure 6.1 and Figure 6.5.
The choice of faulted lines are based on which line gives the largest contribution to the short-
circuit current at the connection point, and hence short-circuit power. Tripping the line with
the largest contribution is a worst case scenario, in terms of short-circuit capacity at the
connection point. The short-circuit calculation results can be found in Appendix B.

In Event 3, a three-phase short-circuit is applied to the connection busbar. The short-circuit is
modeled with impedance equal to zero, so that the voltage at the connection point is zero
during the fault. The fault is cleared after 150 ms.
39
Dynamic simulations
7.4 Simulation results
The wind turbine model used in the simulations is simply a PSS/E model of induction
generator called CIMTR3. No protection of the wind farm is modeled. It is assumed that the
wind farm will stay connected as long as the system has a stable behavior after a disturbance.
This is an approximation, but the main focus of this thesis is to describe the dynamic behavior
on the onshore power system with different transmission technologies.
Table 7.2 presents the main results from the simulations.
Table 7.2: Simulations results
As can be seen from the table, all simulations with HVDC Light show stable behavior. For all
the simulations marked as stable, the onshore voltage recovers in such a manner that the
offshore wind farm has to stay connected based on the fault ride through requirement.
For all simulations with AC cable connection the following variables are plotted. System base
is equal to 1000 MVA.
x Voltage at the connection point [pu]

x Voltage at the offshore busbar [pu]
x Deviation from synchronous speed of the wind turbine generators [pu]
x Reactive power from one of two 200 Mvar SVC at the connection point [pu on system
base]
For all simulations with HVDC Light connection, the following variables are plotted:
x Voltage at the connection point [pu]

x Voltage at the offshore busbar [pu]
x Deviation from synchronous speed of the wind turbine generators [pu]
x DC voltage at the voltage regulating HVDC Light converter [kV]
x Reactive power from the HVDC Light converter at the connection point [pu on system
base]
40
Dynamic simulations
In order to present the main characteristics of the dynamic response both with AC cable and
HVDC Light, all the plots described above are given for Case 1 and Event 1 in 7.5 and 7.6.
In the rest of the simulations, only the most significant plots are presented in this chapter. The
complete sets of plots are found in Appendix D. For the simulations with unstable behavior,
only the voltage at the connection point and reactive power from one SVC are plotted.
7.5 Case 1, AC cable
Simulations with the three events described in 7.3 are done for load flow Case 1, with AC
cable connection of the offshore wind farm. Case 1 is the original load flow situation, with
1000 MW offshore wind power added at Feda, and 900 MW load at Hasle to simulate export
to Sweden. With AC cable connection, SVCs are added at connection point Feda in order to
fulfill the requirements from Statnett regarding cos(ĳ) at the connection point.
7.5.1 Event 1, bus fault at Øie

The fault at Øie lasting for 150 ms reduces the voltage at Feda to about 0,26 pu. The grid code
requires that the wind farm should not disconnect as long as the fault is cleared within 150 ms
and the voltage recovers according to Figure 7.1. After the fault is cleared, the system has to
be stable. The plots for voltage at the onshore connection point Feda, offshore voltage, speed
deviation from synchronous speed of the wind turbine generators and reactive power from one
SVC at Feda are given in Plot 7.1 to Plot 7.4.
Plot 7.1: Voltage at Feda [pu]
41
Dynamic simulations
Plot 7.2: Offshore voltage [pu]
As can be seen from the simulation plots in Plot 7.1 and Plot 7.2, the voltage is reduced both
onshore and offshore during the fault. The voltage is recovered at the connection point
onshore about 0,5 second after the fault is cleared. Based on the fault ride through
requirement described in 7.1, this implies that the wind turbines have to stay connected.
42
Dynamic simulations
Plot 7.3: Deviation from synchronous speed of the wind turbine generators [pu]
The speed of the equivalent wind turbine increases during the fault because of the reduced
electric torque caused by the low voltage onshore. When the fault is cleared, the speed rapidly
decreases because of the electric torque, and swings back to the pre fault speed.
43
Dynamic simulations
Plot 7.4: Reactive power from one SVC [pu on system base]
The SVCs produce as much reactive power as possible in order to raise the voltage. However,
as can be seen in Plot 7.4, one SVC is not able to supply its maximum value of 200 Mvar
reactive power until the voltage at the connection point has recovered at approximately t=1,25
s. The reason for this is that the produced reactive power from the SVC is proportional to the
voltage squared outside the working range of the SVC, and the voltage is low at the
connection point after the fault. After the SVC reaches 200 Mvar at t=1,25 s, the reactive
power is quickly regulated down in order to maintain the voltage at approximately 1 pu.
7.5.2 Event 2, line fault and tripping of line
The dynamic behavior after Event 2 (line fault and tripping of line) is similar to those from
Event 1 (bus fault at Øie). The voltage onshore and offshore is however reduced more during
the fault. Additionally, it can be observed from Plot 7.5 that the speed of the wind turbine
generator swings to a speed lower than the pre fault speed.
44
Dynamic simulations
This can be explained from the fact that the frequency in the whole system is reduced for
Event 2. The generator speed is dependent on the frequency for induction generators directly
connected to the grid. After the line is tripped, the current in the other lines become too large,
and some generators trip because of over current protection. With the same load in the system
for reduced production, the system frequency will drop according to the speed-droop
characteristics of the system [15].
7.5.3 Event 3, three-phase short-circuit at Feda

The voltage onshore never recovers after the three-phase short-circuit at Feda, as can be seen
in Plot 7.6. Based on the fault ride through requirement, the wind farm is allowed to trip at
approximately t = 0,7 s, just after the fault is cleared. The SVC is not able to supply more than
half of the reactive power rating after the fault is cleared, because of the low voltage.
45
Dynamic simulations
7.6 Case 1, HVDC Light
Simulations for the three events described in 7.3 are done for the load flow situation Case 1
with HVDC Light connection of the offshore wind farm. Case 1 is the original load flow
situation, with 1000 MW offshore wind power added at Feda, and 900 MW load at Hasle to
simulate export to Sweden.

The fault at Øie lasting for 150 ms reduces the voltage at Feda to about 0,26 pu. The grid code
requires that the wind farm should not disconnect as log as the fault is cleared within 150 ms.
After the fault is cleared, the system should be stable. The plots for voltage at the onshore
connection point Feda, offshore voltage, reactive power from the HVDC Light converter at
Feda, the DC voltage and the speed of the wind turbines are given in Plot 7.7 to Plot 7.11.
46
Dynamic simulations
Even though the onshore voltage is reduced approximately to the same value as with AC
cable connection, the voltage recovers much faster with HVDC Light. The fault ride through
requirement requires the wind turbines to stay connected for the simulated event. A small
local peak in the voltage can be observed at the instant the fault is cleared. This corresponds
in time to the large peak in reactive power from the onshore converter given in Plot 7.11.
47
Dynamic simulations
Plot 7.8: Offshore voltage [pu]
The offshore voltage is almost unaffected by the fault onshore. Only a small temporary
change is observed in Plot 7.8. The wind turbines can stay online without noticing the fault
onshore. However, the problem with this is described in 5.3.5. In order to avoid the DC
voltage to increase drastically, power cannot be transferred into the DC side during the fault.
The offshore converter is in Udc control mode, and increases the frequency, and hence also the
rotational speed of the wind turbines as shown in Plot 7.9. This allows the wind turbines to
store the energy produced during the fault as rotational kinetic energy. When the fault is
cleared, the rotational speed decreases in order to convert the additional rotational energy to
electric energy transferred to the onshore grid through the HVDC Light system. The faster the
speed decreases after the fault, the larger the additional power from the wind turbines will be.
The HVDC Light converter is loaded relatively close to rated power; hence the decrease in
speed can not take place too fast as the IGBTs have limited capacity.
48
Dynamic simulations
Plot 7.10: DC voltage at the voltage regulating HVDC Light converter [kV]
49
Dynamic simulations
Plot 7.11: Reactive power from the HVDC Light converter at the connection point [pu on
system base]
Unlike the situation with SVCs at the connection point, the HVDC Light converter can supply
a large amount of reactive power at the instant the fault is disconnected. A high peak of
reactive power can be observed from Plot 7.11, causing the voltage to recover almost to the
pre fault value. The fact that the HVDC Light converter can supply reactive power during the
fault, means that the HVDC Light converter contributes to the short-circuit power [10]. Even
so, the converter can only operate inside the PQ diagram in Figure 5.1.
The peak of reactive power supplied from the HVDC Light converter at the instant the fault is
cleared should be investigated. Certain aspects of the model could be idealized so that the
dynamic behavior is not realistic compared to a physical system. Comparison with
measurements from a real HVDC Light system would be useful in order to verify the response
shown in the simulation.

Event 1 (bus fault at Øie). The voltage onshore and offshore is reduced more during the line
fault, as was the case with AC cable connection. The system frequency is reduced because of
tripping of generators from over-current protection.

The simulation results are similar as to Event 1 and 2. After the three-phase short-circuit at
Feda, the voltage is fully recovered at t = 1 s, as can be seen from Plot 7.12.
50
Dynamic simulations
Based on the fault ride through requirement the wind farm must stay connected. The system is
stable after the fault is cleared, unlike the same situation with AC cable connection.
Simulations for the three events described in 7.3 are done in load flow Case 2 with AC cable
connection of the offshore wind farm. In Case 2 both load and production in the areas
surrounding Feda is ramped down. Generators are disconnected in order to establish a load
flow situation with lower short circuit capacity at the connection point. 1000 MW offshore
wind power is assed to Feda, and a 900 MW load is assed at Hasle in order to simulate export
of power to Sweden. The short-circuit capacity at Feda is reduced from 7147 MVA in Case 1
to 5228 MVA in Case 2 due to disconnection of generators.
The simulations with AC cable show a stable behavior of all the disturbances. The intention
of establishing the load flow situation called Case 2 was to trig unstable dynamic behavior
because of reduced short-circuit capacity. The reduction of the loads in the areas however
seems to have a bigger influence on the dynamic behavior than the reduction in the short-
circuit power.
51
Dynamic simulations

The fault at Øie lasting for 150 ms reduces the voltage at Feda to approximately 0,24 pu. The
grid code requires that the wind farm should not disconnect as long as the fault is cleared
within 150 ms and the onshore voltage recovers according to Figure 7.1. After the fault is
cleared, the system should be stable. The plot for voltage at Feda is given in Plot 7.13.
By comparing Plot 7.1 to Plot 7.13, it can be observed that the voltage recovers slightly faster
than for the load flow case with all the generators connected.
As short-circuit power at the connection point indicates the strength at the busbar, other
factors influence the stability as well. In Case 2, load was scaled down in order to disconnect
generators. The active and reactive power load was scaled down at a fixed P/Q ratio. The
reduced load in the area surrounding Feda contributes the voltage recovery.
As for Event 1 with bus fault at Øie, the voltage onshore recovers faster than in Case 1. This
is despite the reduced short-circuit power at the point, and can be explained by lighter loading
in the area from the load flow.
52
Dynamic simulations

For the case with a three-phase short-circuit with no impedance at Feda, the voltage recovers
after the fault is cleared. This can be observed in Plot 7.14.
According the fault ride through requirement in Figure 7.1, the voltage recovers fast enough
so that the wind farm must stay connected. 750 ms after the fault is cleared, the voltage is
higher than 0,9 pu.
Simulations for the three events described in 7.3 are done in load flow Case 2 with HVDC
Light connection of the offshore wind farm. In Case 2 both load and production in the areas
surrounding Feda is ramped down. Generators are disconnected in order to establish a load
flow situation with lower short circuit capacity at the connection point. 1000 MW offshore
wind power is assed to Feda, and a 900 MW load is assed at Hasle in order to simulate export
of power to Sweden.

The fault at Øie lasting for 150 ms reduces the voltage at Feda to about 0,24 pu. The voltage
recovers almost immediately after the fault is cleared, and the dynamic response is similar the
one presented for the same event in Case 1.
53
Dynamic simulations

Event 1 (bus fault at Øie).

After the fault is cleared at Feda, the voltage recovers almost immediately. The voltage
recovery is slightly faster than in Case 1, as can be seen by comparing Plot 7.12 and Plot 7.15.
connection of the offshore wind farm. In Case 3 the load is kept at the maximum value
corresponding to the original load flow situation, but the production is ramped down, without
disconnection any generators. 1000 MW power from Denmark is imported through the
HVDC links at Kristiansand, modeled as a load with value -1000 + j 500 MVA. 1000 MW
offshore wind power is assed to Feda, and there is no export of power to Sweden.
54
Dynamic simulations

1000 MW HVDC import at Kristiansand in addition to 1000 MW offshore wind power added
at Feda result in heavy loading of the transmission lines. After the fault at Øie is cleared, the
voltage at Feda recovers. As can be seen from Plot 7.16, the voltage 750 ms after the fault is
cleared is just over 0,9 pu. This is the limit for which the wind farm has to stay connected.
The short-circuit power is larger than in Case 2, but the loading in the lines surrounding Feda
is much higher because of the peak load and import of 1000 MW HVDC at Kristiansand. 500
Mvar reactive power is drawn by the HVDC converter at Kristiansand. The system frequency
is reduced after the disturbance because of tripping of generators by over current protection.
A similar response as the one observed for Event 1 (line fault at Øie) can be seen from Event
2 (line fault and tripping of line). The voltage recovers in such a manner that the wind farm
has to stay connected. A marginally slower recovery of the voltage would allow the wind
farm to trip.

The voltage at Feda never recovers after the three-phase short-circuit at Feda. Based on the
fault ride through requirement, the wind farm is allowed to trip just after the fault is cleared.
55
Dynamic simulations
The SVC is not able to supply more than half of the reactive power rating after the fault is
cleared, because of the low voltage.
Light connection of the offshore wind farm. In Case 3 the load is kept at the maximum value
corresponding to the original load flow situation, but the production is ramped down, without
disconnection any generators. 1000 MW power from Denmark is imported through the
HVDC links at Kristiansand, modeled as a load with value -1000 + j 500 MVA. 1000 MW
offshore wind power is assed to Feda, and there is no export of power to Sweden.

After the fault at Øie is cleared, the voltage at Feda recovers almost immediately. The
response is similar to the one in Case 1 for the same event. Generators are tripped because of
the heavy loading of the lines, and the system frequency decreases.

The response after the line is tripped is very similar the one in Case 1. Generators are tripped
because of the heavy loading of the lines, and the system frequency decreases.

After the fault is cleared at Feda, the dynamic response is very similar to the one in Case 1. As
for the other events in Case 3, the system frequency is reduced after the fault is cleared
because of tripping of generators.
connection of the offshore wind farm. Case 4 is the original load flow situation, with 1000
MW offshore wind power added at Kårstø, and 900 MW load at Hasle to simulate export to
Sweden.
7.11.1 Event 1, bus fault at Sauda

The fault at Sauda lasting for 150 ms reduces the voltage at Kårstø to about 0,44 pu. As long
as the voltage is above 0,9 pu 750 ms after the fault is cleared, the wind farm has to stay
connected. This correspond to t = 1,4 s in Plot 7.17, and the voltage has recovered to a value
higher than 0,9 pu at this time.
56
Dynamic simulations
Plot 7.17: Voltage at Kårstø [pu]
The wind farm has to stay connected. However, by inspecting the plots for Case 1, it can be
observed that the voltage at Kårstø recovers more slowly in Case 4. This is as expected
because of the considerably lower short-circuit power at Kårstø.
Because the voltage offshore does not decrease as much as at Feda, the wind turbine
generators do not accelerate as much. The SVCs at Kårstø show the same behavior as the case
at Feda, the supplied reactive power does not reach the maximum value of 200 Mvar before
the voltage has recovered.

After the fault at Haavik, and then clearing of the line from Haavik to Kårstø, the voltage
onshore does not recover. The voltage at Kårstø is shown in Plot 7.18.
57
Dynamic simulations
The wind farm is allowed to trip shortly after the line is tripped based on the fault ride through
requirement described in Figure 7.1.
7.11.3 Event 3, three-phase short-circuit at Kårstø

As for Event 2, the voltage does not recover after the fault at Kårstø is cleared at t = 0,65 s.
The voltage at Kårstø is showed in Plot 7.19. The wind farm is allowed to trip.
58
Dynamic simulations
Light connection of the offshore wind farm. Case 4 is the original load flow situation, with
1000 MW offshore wind power added at Kårstø, and 900 MW load at Hasle to simulate
export to Sweden.

With HVDC Light connection, the onshore voltage recovers practically immediately after the
fault is cleared. The voltage at Kårstø is showed in Plot 7.20.
59
Dynamic simulations
The dynamic response is quite similar to the response in Case 1, 2 and 3, even though the
connection point in these cases was Feda. A small maximum corresponding to approximately
1,05 pu in the voltage can however be observed at t=1s.

The response after the line is tripped is very similar the one in Case 1.

After the fault is cleared at Feda, the dynamic response is very similar to the one in Case 1.
connection of the offshore wind farm. In Case 5 1000 MW offshore wind power is connected
at Kårstø. The load is kept at the maximum value corresponding to the original load flow
situation, but the production is reduced by disconnecting large generators close to Kårstø.
There is no export of power to Sweden, and no import of power from Denmark through the
HVDC links.
60
Dynamic simulations

The fault at Sauda lasting for 150 ms reduces the voltage at Feda to about 0,35 pu, shown in
Plot 7.21. After the fault is cleared, the voltage does not recover. The wind farm is allowed to
trip shortly after the fault is cleared based on the fault ride through requirement.
Before the disturbance is applied, the SVCs at Kårstø produces rated reactive power. This is
shown in Plot 7.22. Because the voltage never recovers, the SVCs are not able to deliver more
than half the rated reactive power after the fault is cleared.
61
Dynamic simulations
Plot 7.22: Reactive power from one SVC [pu on system base]

As for the case with bus fault at Sauda, the system is not stable after the faulted line is tripped.
The wind farm is allowed to trip shortly after the line is tripped.

After the short-circuit at Kårstø, the voltage does not recover. The wind farm is allowed to
trip shortly after the short-circuit is cleared.
Light connection of the offshore wind farm. In Case 5 1000 MW offshore wind power is
connected at Kårstø. The load is kept at the maximum value corresponding to the original
load flow situation, but the production is reduced by disconnecting large generators close to
Kårstø. There is no export of power to Sweden, and no import of power from Denmark
through the HVDC links.

The dynamic response is stable for the case with a bus fault at Sauda. The voltage at Kårstø
recovers almost immediately after the fault is cleared, similar the response in Case 1.
62
Dynamic simulations

The response after the line is tripped is very similar the one in Case 1.

A perfect three-phase short-circuit at Kårstø with reduced short-circuit power is the most
critical simulations done in this thesis. Kårstø is the connection point with lowest short-circuit
power, and in Case 5 the short-circuit power is reduced additionally. A perfect three-phase
short-circuit is the most severe disturbance, as the voltage at the connection point is zero
during the fault.
With AC cable connection, all simulations in Case 5 were unstable, and the wind farm had to
disconnect. For the case with HVDC Light, the system is stable for all simulations. The
voltage at Kårstø recovers almost immediately after the fault is cleared, and the response is
similar to the other cases. The voltage at Kårstø is shown in Plot 7.23 for Event 3. The wind
farm has to stay connected based on the fault ride through requirement.
63
Improvements of dynamic behavior
8 Improvements of dynamic behavior

In this chapter, suggestions are made to improve the stability of some of the simulations with
unstable behavior. Only the simulations with AC cable connection showed unstable behavior.
Simulations are done for Case 5 in order to verify the improvement of the dynamic behavior.
Case 5 was the load flow case with the offshore wind power connected to Kårstø and reduced
short-circuit power at Kårstø.
Two suggestions are made in this chapter in order to improve the stability for the three events
described in 7.3 for Case 5 with AC cable connection.
8.1 Changing of the power factor

A 500 Mvar shunt reactor is used for compensation of the reactive power produced by the AC
cables in all simulations in 7.4. 500 Mvar is approximately the amount of reactive power
produced by the AC cables with voltage equal 1 pu at the connection point. In order to fulfill
the requirement regarding cos(ĳ) = 0,95 capacitive and inductive at the connection point, two
SVCs with a total capacity of 400 MVA are installed. ± 400 Mvar from the SVCs correspond
to almost cos(ĳ) = 0,91 inductive and capacitive for rated power from the offshore wind farm.
The requirement from Statnett is over fulfilled.
In Case 5, the SVCs produce close to the maximum limit of reactive power in order to keep
the voltage at 1 pu before the disturbance. This can be verified by inspection of Plot 7.22 in
7.13.1. The power factor for the connected wind farm at Kårstø is about 0,91 capacitive. The
inductance of the shunt reactor could easily be increased, causing the cos(ĳ) capacitive at the
connection point to decrease even more. However, the requirement of cos(ĳ) = 0,95 inductive
would not be fulfilled. A switched shunt could be used in order to operate at a lower power
factor than cos(ĳ) = 0,91 capacitive, but still fulfill the power factor requirement of cos(ĳ) =
0,95 inductive.
In this paragraph, the shunt reactor is adjusted as a suggestion to improve the dynamic
behavior in case 5. The shunt reactor is adjusted so that the reactive power from the SVCs in
the load flow was equal to zero. The reactive power from the shunt reactor is then 106,2 Mvar,
as can be seen from Figure 8.1.
64
Figure 8.1: Load flow at Kårstø (Case 5), reduced cos(ĳ)
Plots for the voltage at Kårstø are given for Event 1, 2 and 3 in Plot 8.1, Plot 8.2 and Plot 8.3
respectively.
Plot 8.1: Voltage at Kårstø, Case 1 [pu]
65
66
As can be observed from the plots, the effect of decreasing the reactive power consumption of
the shunt reactor improved the stability for bus fault at Sauda, Event 1. The wind farm has to
stay connected. For the other cases, the voltage onshore does not recover, and the wind farm
can disconnect.
8.2 Adding synchronous condensers

At Kårstø, the short-circuit power is considerably lower than at Feda. In addition to operate at
a lower cos(ĳ) capacitive than required as described in 6.5.1, synchronous condensers were
installed at the connection point instead of the SVCs in order to increase the short-circuit
power. In Kristiansand, a 140 MVA synchronous condenser is installed in order to increase
the short-circuit power at the connection point of the traditional HVDC converters. Data from
this machine is used in the simulations. As data from this machine size was the only one
available, three machines had to be modeled in order to fulfill the requirement of cos(ĳ ) =
0,95, corresponding to ±316 Mvar at rated power. The synchronous condensers are connected
to a 17 kV busbar, as shown in Figure 8.2.
Figure 8.2: Load flow at Kårstø (Case 5), synchronous condensers
Plots for the voltage at Kårstø are given for Event 1, 2 and 3 in Plot 8.4, Plot 8.5 and Plot 8.6,
respectively.
67
68
As can be seen from the plots, exchanging the SVCs with a total rating of 400 Mvar with
three synchronous condensers with a total rating of 420 Mvar, improved the stability further.
Event 2, which is the case with line fault and disconnection of the line, is stable. However, the
voltage 750 ms after the line is tripped, has not recovered to 0,9 pu. The fault ride through
requirement does not require the wind farm to stay connected. For the case with a three-phase
short-circuit at Kårstø, the synchronous condenser did not contribute enough to the recovery
of the voltage.
It was expected that the contribution from the synchronous condensers would improve the
stability more than indicated in the simulations in this paragraph. One possible reason for this
is the modeling of the excitation system [14]. The reactive power supplied from one of the
three synchronous condensers at the instant the fault is applied is about 160 Mvar. When the
fault is cleared, the reactive power supplied is only about 40 Mvar. The plot of the supplied
reactive power for Event 3 is given in Plot 8.7.
69
Plot 8.7: Reactive power from one three 140 Mvar synchronous condenser, Event 3 [pu on
system base]
Because of the rotating mass involved, the synchronous condenser should be able to
contribute with reactive power during the fault and after the fault is cleared, unlike an SVC
which is dependent of the voltage squared outside the working area. The rotating field
winding induces voltage in the stator of the synchronous machine. By applying a large DC
field current the machine will be over magnetized, corresponding to production of reactive
power [8]. When the voltage at the connection busbar is low because of the fault, the current
from the stator and hence reactive power will be high because of the voltage difference
between the stator winding and the connection busbar. This can be seen from Plot 8.7.
However, after the fault is cleared the machine only contributes to about 0,3 pu.
A synchronous machine should be able to contribute to the short-circuit power corresponding

to 2-3 times in the transient period [14].
The model for the excitation system is a user model called NOREX1. With no user manual for
the model, the parameters could not be tuned in the dynamic description file. However, with
the description of the dynamic variables for the excitation system, the transient reactances and
time constants could be changed in order to provoke a different dynamic behavior. It is
expected that a synchronous condenser would contribute with reactive power in such a
manner that the voltage recovers. Further work should be done investigating the possibility of
using synchronous condensers with AC cable connection of large scale offshore wind power.
70
9 Discussion
Certain aspects of the models and simulations are discussed consecutively in the previous
chapters. An overall discussion is made in this chapter as a summary.
As described in 6.5, the cases with AC cable connection of the offshore wind farm was
mainly established in order to have something to compare the response from HVDC Light to.
However, if the technology for floating wind turbines became commercial, it is possible to
install a large wind farm at a distance from shore that makes AC cable transmission an option.
Studies can be made to investigate the possibility of compensating the reactive power
produced by the AC cables between the offshore wind farm and the power system onshore. In
principal, advanced technology should not be necessary in order to accomplish this.
Case 2 was established in order to provoke unstable behavior caused by reduced short-circuit
power at Feda. This was done by reducing the load in the area and disconnection generators.
The reduced short-circuit power at Feda was however still higher than the original case at
Kårstø. Results from the simulations shows that the stability is not affected in the intended
way. In order to disconnect generators, both active and reactive power load was reduced in the
areas surrounding Feda. Even though the short-circuit power was reduced, the voltage drop
onshore is not as severe as intended because of the lower power consumption by the reduced
loads.
For the load flow cases with Feda as connection point for the offshore wind power, only in
Case 3 import of HVDC from Denmark is modeled. Import from the Netherlands was not
modeled because of the high loading of the lines in the peak load situation. Further work
should be done in order to investigate for which load flow situations it is relevant to model
full import of HVDC. Since the model of the Norwegian power system provided by Statnett
only was an isolated model, the HVDC converters were not modeled. As an assumption, the
converters at Kristiansand were modeled as a load equal to -1000 + 500 MVA. The load was
converted as 50 % current and 50 % constant power for both active and reactive power.
Simulation should be done for the complete model with the HVDC converters both at Feda
and Kristiansand in order to get a more realistic dynamic response close to the HVDC
converters.
Future plans for upgrades and new production units in the Norwegian power system are not
accounted for in any simulations. Such plans should be implemented in a complete model of
the power system with the Sweden and the HVDC converters modelled, in order to get a
realistic dynamic behavior. The new gas power plant at Kårstø at 420 MW is not modelled.
The contribution of the gas power plant will lead to increased short-circuit power at Kårstø,
making AC cable connection with SVC more favourable for the kind of events simulated, as
the generator can contribute to the short-circuit power during the fault.
The plots for reactive power from the HVDC Light converters should be subject of further
investigation. The onshore converter supplies a very fast peak of reactive power at the instant
the fault is cleared, contributing to the voltage recovery. In order to verify that the HVDC
Light model gives a correct dynamic response of reactive power, simulation results should be
compared with real measurements of an existing HVDC Light system.
71
In the existing plans for offshore wind power application with HVDC Light connection, DC
choppers are planned. The excess power produced during the fault onshore will be dissipated
in the resistance. In this way, the offshore wind farm will not notice the fault onshore. The
focus on fault ride through capability for the wind turbines themselves will not be an issue for
faults in the onshore power system. As can be seen from the simulation results with HVDC
Light, the voltage offshore is practically unaffected by the fault onshore. Because of the above
mentioned reason, the focus in this thesis has therefore been on the impact of the transmission
technologies onshore. For simulations with AC cable connection however, the protection of
the wind turbines itself should be modeled for different wind turbine technologies.
When synchronous condensers were added at Kårstø, the voltage recovery improved
compared to the solution with SVCs. The user model of the excitation system for the
synchronous condensers used in the simulations was taken from a synchronous condenser
from the Norwegian power system model, without any documentation. It is expected that the
dynamic response could improve the voltage recovery even more than showed in the
simulations, based on the plot for supplied reactive power after the three-phase short-circuit at
Kårstø is cleared.
72
10 Conclusion
The power production in Norway mainly consists of hydro power, and up until recently there
has been a net export of power in years with average precipitation. This situation has now
changed, and with the ambitious national goal of no net CO2 emission in 2050, great effort
has to be made. Because of the great wind resources in the sea areas surrounding Norway,
offshore wind power seems like a promising alternative.
In this study, grid integration of a 1000 MW offshore wind farm was studied. Two possible
connection points in the Norwegian power system were found; Feda and Kårstø. Feda is the
point where the HVDC cable to the Netherlands is connected, and the voltage at the busbar is
300 kV. Kårstø is the other connection point, also with a voltage of 300 kV. At Kårstø, and
Karmøy west of Kårstø, there is a lot of power consuming industry. The offshore wind power
connected to Kårstø will change the direction of the power flow in the lines going to the area.
There are several challenges with offshore wind power. Depths just outside the Norwegian
coast reach several hundred meters, indicating that floating turbines should be used. This
technology is not yet commercial, but there are several concepts under development. It is
however possible to build large offshore wind farms in areas a couple of hundred kilometers
south of Norway. Depths are suitable for technology available today, with wind turbines
foundations on the seabed. The distance makes AC cable connection difficult because of the
reactive power produced by the cables, and HVDC should therefore be used. HVDC Light is a
VSC HVDC technology well suited for offshore wind power, and with rating up to 1174 MW
connection of large wind farms are possible. The equipment is compact, and there is no
requirement for short-circuit capacity either onshore or offshore.
Dynamic simulations have been done for the two connection points both with AC cables of
100 km and HVDC Light of 600 km, for different faults close to the connection points. The
HVDC Light converter contributes to the voltage recovery onshore so that the fault ride
through requirement requires the wind turbines to stay connected. This counts for all
simulations in this thesis. Immediately after the fault is applied, the onshore HVDC Light
converter supplies a large amount of reactive power reactive power. The contribution to the
fault current during the fault, as well as the reactive power produced after the fault is cleared,
supports the voltage at the connection point onshore. The reactive power is regulated so that
the onshore voltage is approximately maintained at 1 pu in the time after the fault is cleared.
For the cases with AC cable connection, the contribution of reactive power from the SVCs is
limited, depending on how low the voltage is during the fault. After the fault is cleared, the
SVCs produce as mush reactive power as possible. Because the SVCs are dependent of the
voltage squared outside the working area, the amount of reactive power is not large enough in
some of the simulations. In these cases, the voltage doesn’t recover, and the system becomes
unstable. The wind farm is allowed to trip, and the system has to restart in order to return to
the pre fault condition.
From the simulations results, the conclusion can be drawn that system stability with AC cable
transmission with SVC as reactive power compensation is highly dependent of short-circuit
power, load flow situation in the power system and power system topology. With HVDC
Light, the system stability is not affected for any of these factors in the simulations done.
Measures can however be done regarding the system stability with AC cable transmission.
Synchronous condensers can be added at the connection point in order to contribute to the
73
short-circuit power. Simulations in this thesis with synchronous condensers did show some
improvement compared to the solution with SVCs, but the voltage did not recover after a
three-phase short-circuit at Kårstø. The model of the excitation system might need to be tuned
in order to give a proper representation of the dynamic response of the synchronous condenser.
The voltage offshore is low during the onshore faults with AC cable connection. The wind
turbine protection equipment is not modeled, but they are assumed to stay connected during
the fault. With HVDC Light, the voltage offshore remains almost constant during the fault.
The energy produced during the fault is stored as rotational kinetic energy in the turbines. A
DC chopper will be used in existing plans for wind applications with HVDC Light, so that the
wind turbines will be unaffected by a fault onshore.
From the TSO’s point of view, the simulations indicate that HVDC Light is the favorable
solution compared to AC. The reason for this is the ability of the HVDC Light converter to
support the voltage during a fault onshore close to the connection point of the offshore wind
farm. For the wind farm owner, the fact that the offshore wind turbines will not notice an
onshore fault with a HVDC Light system with DC chopper, HVDC Light seems to be the
most favourable transmission technology. Since the focus on fault ride through requirement
for wind turbines connected to the onshore grid with HVDC Light is insignificant, the
possibility of using less sophisticated wind turbines in terms of electrical properties arises.
74
11 Further work
This thesis is the end of the work started during the autumn 2007 with the project, and
suggestions for further work are presented in this chapter. Statkraft will continue to offer
students the possibility to write their master’s thesis in cooperation with them, and the
suggestions made in this chapter can be used as input for formulating new projects.
There are several challenges with grid integration of large scale offshore wind power. VSC
HVDC will probably have to be used, based on the distances to the areas where it is possible
to install offshore wind turbines with today’s technology. Base on this assumption, power
system studies with PSS/E and the user model of HVDC Light from ABB is a good starting
point. As the focus in this thesis has been on the onshore grid side, work can be done
investigating the offshore power system. The next version of the PSS/E user model for HVDC
Light will have implemented a complete control system especially for offshore wind power
with an isolated power system.
An interesting case to investigate is the connection of small wind farms to oil platforms.
Models of existing platforms should be used in order to investigate how the equipment on the
platforms behaves together with the small wind farms, and vice versa. The groups of
platforms and wind farms could be could be connected in an offshore grid, with one common
HVDC Light converter offshore. The impact of different wind turbine technologies can be
investigated for such a case. General models for full frequency turbines and DFIG turbines
can be provided by Statkraft, and the models can be tuned in order to correspond to the wind
turbines from different manufacturers.
A model of Siemens’s VSC HVDC technology HVDC PLUS is also available in PSS/E.
Simulations can be done with HVDC Light and HVDC PLUS in order to compare the two
technologies.
75
12 References
[1] ABB, Its time to connect – technical description of HVDC Light® technology (2007)
[2] ABB, User manual HVDC Light (2008)
[3] ABB, www.abb.com/hvdc
[4] Barberis N., Todorovic J., Ackermann T. (2005), Loss evaluation of HVAC and HVDC
transmission solutions for large offshore wind farms, Science Direct
[5] Chengyoug, Z., A Control Strategy for VSC-HVDC System Based on Analytic
Expression, IEEE
[6] Gustafsson, M. (Statnett), conversations
[7] Holen, A. T., Fosso, O. B., Olsen, K. J. (2006), TET4115 Power system analysis, Tapir
akademisk forlag
[8] Hubert, C. I. (2002), ELECTRIC MACHINES Theory, Operation, Applications,

Adjustment, and Control, Prentice Hall
[9] IEEE working group (1997), IEEE Guide for Planning DC Links Terminating at AC
Locations Having Low Short-Circuit Capacities, IEEE
[10] Jiang-Häfner Y., Hyttinen M., Pääjärvi B., On the Short Circuit Current Contribution of
HVDC Light, IEEE
[11] Koutiva, X. I., Vrionis, T. D., Vovos, N. A., Giannakopoulos, G. B., Optimal
Integration of an offshore wind farm to a weak grid, IEEE
[12] Kundur, P. (1994), Power System Stability and Control, The McGraw-Hill Companies
[13] Larssen, V. (Nexans), conversations
[14] Leirbukt, A. (ABB), conversations
[15] Machowski, J., Bialek, J. W., Bumby, J. R. (1997), Power System Dynamics and
Stability, John Wiley & Sons
[16] Manwell J. F., McGowan J. G., Rogers A. L. (2002), Wind Energy Explained – Theory,
Design and Application, John Wiley & Sons
[17] Mohan, N., Undeland, T. M., Robbins W. P. (2003), Power electronics – Converters,
Applications and Design, John Wiley & Sons, Inc.
[18] PSS/E user manual
76
[19] Statnett SF, FIKS Funksjonskrav i kraftsystemet
[20] Statnett, www.statnett.no
[21] Wang, C., Wang, L., Shi, L., A survey on Wind Power Technologies in Power Systems,
IEEE
[22] Wangensteen, I. (2006), Power System Economics – Nordic Electricity Marked, Tapir
Academic Press
[23] Weimers, L. New Markets need new Technology, IEEE
77
Appendix A: Limit checking reports
I
A.1 Limit checking report for busbar voltages
between 130 and 400 kV
Original load flow case
--------------------------------------------------------------------------------
PTI INTERACTIVE POWER SYSTEM SIMULATOR--PSS/E TUE, MAY 20 2008 10:55

STATNETTS DRIFTDATASETT FOR NORGE
BUSES WITH VOLTAGE GREATER THAN 1.1000:
BUS# X-- NAME --X BASKV AREA V(PU) V(KV) BUS# X-- NAME --X BASKV AREA V(PU) V(KV)
50419 DUM-KVA2 130.00 68 1.1121 144.57 50556 DUM-T8- 220.00 67 1.1250 247.51
51168 BERGER 130.00 51 1.1133 144.73 51186 FUSDAL-A 130.00 51 1.1130 144.69
57200 NR-RAZFE 220.00 67 1.1250 247.49 57204 N.ROSS2 220.00 67 1.1250 247.49
57299 SUNDFJ-B 130.00 67 1.1035 143.46 57683 GLOMFJD1 130.00 67 1.1022 143.29
58049 KVAN-NOR 130.00 68 1.1120 144.56 58353 STRAUM1A 130.00 69 1.1075 143.97
58363 INNSET1 130.00 69 1.1119 144.54 58702 VARANGER 220.00 69 1.1169 245.72
BUSES WITH VOLTAGE LESS THAN 0.9000:
51137 SMESTD1B 130.00 51 0.8906 115.78 51178 PRESTEGA 130.00 51 0.8882 115.47
51179 BRISKEBY 130.00 51 0.8893 115.60 51183 SKOYEN 130.00 51 0.8901 115.72
52403 SAVALEN1 130.00 52 0.8923 116.00 52413 TYNSET1 130.00 52 0.8799 114.38
52433 TRYSIL1 130.00 52 0.8980 116.74 57548 ULSET-B- 130.00 66 0.8942 116.24
A.2 Limit checking report for branches between 130
and 400 kV
Original load flow case
--------------------------------------------------------------------------------
PTI INTERACTIVE POWER SYSTEM SIMULATOR--PSS/E TUE, MAY 20 2008 10:54

BRANCH LOADINGS ABOVE 100.0 % OF RATING SET A:
X--------- FROM BUS ----------X X---------- TO BUS -----------X CURRENT(MVA)

BUS# X-- NAME --X BASKV AREA BUS# X-- NAME --X BASKV AREA CKT LOADING RATING PERCENT
50029 KKI-THA2 20.000* 66 56573 THAMSHAM 130.00 66 1 37.2 29.1 127.7
50032 KKI-FES1 20.000 67 57623 SVABO-A- 130.00* 67 5 47.2 46.4 101.8
50035 KKI-FINN 20.000 69 58393 FINNFJOR 130.00* 69 1 52.3 52.2 100.2
50280 MF-O.VIN 300.00 52 50539 DUM-VINS 12.000* 52 1 160.8 142.9 112.5
50483 DUM-LUN1 50.000* 56 54463 LUND1 130.00 56 1 77.1 55.0 140.2
50492 DUM-HAVI 66.000* 61 55152 HAAVIK3 300.00 61 1 80.8 79.5 101.6
50495 DUM-T1- 130.00 65 57015 AURA-TEA 17.000* 65 1 77.4 69.6 111.2
50563 DUM-KLE2 66.000 66 57032 KLEBU3 300.00* 66 1 158.1 157.4 100.4
50574 DUM-FAN2 130.00 62 55275 FAN-T2-T 17.000* 62 1 58.9 58.2 101.2
50588 DUM-STOR 66.000 61 55172 STORD3 300.00* 61 1 108.8 99.3 109.6
50605 DUM-FLE1 66.000 55 51203 FLESAK1A 130.00* 55 1 108.1 75.4 143.3
50608 DUM-FLE4 130.00 55 51202 FLESAK3 300.00* 55 1 125.6 116.0 108.2
50608 DUM-FLE4 130.00 55 51203 FLESAK1A 130.00* 55 1 119.3 106.0 112.5
50635 DUM-MYKS 11.000* 53 51393 MYKSTF1A 130.00 53 1 55.4 49.7 111.5
50637 DUM-POR1 130.00 55 51244 PORSGR0 17.000* 55 1 109.3 100.0 109.3
50638 DUM-POR2 130.00 55 51245 PORSGRKA 17.000* 55 1 109.5 100.0 109.5
50654 DUM-T7- 130.00 51 51132 SMESTAD3 300.00* 51 1 268.3 246.2 109.0
50654 DUM-T7- 130.00 51 51137 SMESTD1B 130.00* 51 1 259.8 246.2 105.5
50678 DUM-ULV1 50.000 51 51102 ULVEN3 300.00* 51 1 148.7 138.6 107.3
50679 DUM-ULV2 50.000 51 51102 ULVEN3 300.00* 51 1 149.1 138.6 107.6
50751 DUM-RJUK 130.00 54 51711 RJUKAN 400.00* 54 1 297.4 282.6 105.2
50751 DUM-RJUK 130.00 54 51713 RJUKAN-T 130.00* 54 1 294.1 282.6 104.1
50760 DUM-HAMM 66.000* 69 58543 HAMMERFE 130.00 69 1 56.1 49.2 114.0
51174 PRESTEGA 11.000 51 51178 PRESTEGA 130.00* 51 1 43.5 39.4 110.4
51175 BRISKEBY 11.000 51 51179 BRISKEBY 130.00* 51 1 44.2 39.4 112.3
51180 SKOYEN 11.000 51 51183 SKOYEN 130.00* 51 1 44.9 39.4 114.0
51180 SKOYEN 11.000 51 51183 SKOYEN 130.00* 51 2 44.9 39.4 114.0
51196 AKERSBER 130.00* 51 51197 AKERSBER 11.000 51 1 44.3 39.4 112.5
51326 SPIKKEST 130.00* 51 51330 SPIKKEST 20.000 51 1 18.3 18.2 100.5
51403 NORE2-A- 130.00 53 51407 NORE2-G2 11.600* 53 1 29.0 28.7 101.1
51483 MAR-A- 130.00 55 51485 MAR-G1- 6.0000* 55 1 37.9 36.1 105.0
51483 MAR-A- 130.00 55 51486 MAR-G2- 6.0000* 55 1 38.8 36.1 107.5
51483 MAR-A- 130.00 55 51487 MAR-G3- 6.0000* 55 1 40.0 36.1 110.8
51484 MAR-B- 130.00 55 51488 MAR-G4- 6.0000* 55 1 39.8 36.1 110.1
51484 MAR-B- 130.00 55 51489 MAR-G5- 6.0000* 55 1 39.8 36.1 110.1
51484 MAR-B- 130.00* 55 51718 RJUKAN-B 130.00 54 1 290.9 218.0 133.4
51593 SUNDLAND 130.00* 55 51595 SUNDLAND 20.000 55 1 34.8 17.4 199.8
51593 SUNDLAND 130.00 55 51595 SUNDLAND 20.000* 55 3 43.5 17.4 250.0
51713 RJUKAN-T 130.00* 54 51718 RJUKAN-B 130.00 54 1 294.1 184.2 159.7
52323 ABJORA1 130.00 52 52329 AABJ-G4- 11.000* 52 1 100.8 94.3 106.9
53021 DAGALI4 400.00* 53 53025 DAGALI-R 20.000 53 1 112.3 104.4 107.5
53133 UVDAL1-1 130.00 53 53136 UVDA1-G1 8.0000* 53 1 49.1 46.4 105.8
53133 UVDAL1-1 130.00 53 53137 UVDA1-G2 8.0000* 53 1 49.6 46.4 107.0
54022 KR.SAND3 300.00* 56 54025 KR.SAND 110.00 56 3 120.3 110.0 109.4
54152 TJORHOM3 300.00 57 54156 TJORHOM 13.000* 57 1 134.3 131.3 102.3
54193 LYSEBTN1 130.00 57 54197 LYSEB-G4 9.0000* 57 4 48.4 46.4 104.4
54503 AKLAND1 130.00* 56 54543 HOLT1 130.00 56 1 52.2 45.0 115.9
54583 BOYLEFOS 130.00 56 54584 BOYLEFOS 66.000* 56 1 88.7 78.8 112.6
55112 NESFLAT3 300.00 61 55118 SULD2-G1 10.000* 61 1 76.9 76.2 101.0
55513 MIDTBYG1 130.00 62 55519 MIDTBYGD 11.000* 62 1 20.1 20.0 100.5
55825 BJOL-G1 12.000 62 55829 NYEBJOLV 130.00* 62 1 101.4 95.7 105.9
56032 LEIRDOL3 300.00 63 56042 FORTUN3 300.00* 63 1 272.6 259.8 104.9
56403 MOSKOG 130.00* 64 56404 STAK-66A 66.000 64 1 65.2 49.2 132.6
56743 TAF-K4- 130.00 65 56746 TA-K4-G1 11.000* 65 1 55.2 53.1 103.9
57005 SUNNDALS 20.000 65 57013 AURA1 130.00* 65 1 23.9 22.2 107.5
57053 NEA1 130.00* 66 57054 NEA66- 66.000 66 1 29.4 27.8 105.8
57132 TUNSJDAL 300.00 67 57136 TUNN-G12 8.5000* 67 1 88.8 85.7 103.6
57132 TUNSJDAL 300.00 67 57137 TUNN-G34 8.5000* 67 1 89.4 85.7 104.3
57203 N.ROSS1A 130.00 67 57214 NROSS-G1 11.000* 67 1 44.8 44.8 100.0
57203 N.ROSS1A 130.00 67 57215 NROSS-G2 11.000* 67 1 44.8 44.8 100.0
57883 LOMI1 130.00 67 57887 LOMI-G2- 10.500* 67 1 63.9 62.8 101.8
58463 KVALOYA1 130.00 69 58464 KVALOY66 66.000* 69 1 48.7 47.1 103.5
58463 KVALOYA1 130.00 69 58464 KVALOY66 66.000* 69 2 49.0 47.1 104.0
1000 MW offshore wind connected to Feda, 900 MW load at Hasle
--------------------------------------------------------------------------------
PTI INTERACTIVE POWER SYSTEM SIMULATOR--PSS/E FRI, MAY 16 2008 11:58

50419 DUM-KVA2 130.00 68 1.1121 144.57 50556 DUM-T8- 220.00 67 1.1250 247.51
57200 NR-RAZFE 220.00 67 1.1250 247.49 57204 N.ROSS2 220.00 67 1.1250 247.49
57299 SUNDFJ-B 130.00 67 1.1035 143.46 57683 GLOMFJD1 130.00 67 1.1022 143.29
58049 KVAN-NOR 130.00 68 1.1120 144.56 58353 STRAUM1A 130.00 69 1.1075 143.97
58363 INNSET1 130.00 69 1.1119 144.54 58702 VARANGER 220.00 69 1.1169 245.72
51179 BRISKEBY 130.00 51 0.8738 113.59 51183 SKOYEN 130.00 51 0.8746 113.70
52403 SAVALEN1 130.00 52 0.8908 115.81 52413 TYNSET1 130.00 52 0.8783 114.18
52433 TRYSIL1 130.00 52 0.8963 116.52 57548 ULSET-B- 130.00 66 0.8926 116.04
and 400 kV
1000 MW offshore wind connected to Feda, 900 MW load at Hasle
--------------------------------------------------------------------------------
PTI INTERACTIVE POWER SYSTEM SIMULATOR--PSS/E FRI, MAY 16 2008 12:00


50029 KKI-THA2 20.000* 66 56573 THAMSHAM 130.00 66 1 37.1 29.1 127.6
50032 KKI-FES1 20.000 67 57623 SVABO-A- 130.00* 67 5 47.2 46.4 101.8
50280 MF-O.VIN 300.00 52 50539 DUM-VINS 12.000* 52 1 161.5 142.9 113.0
50483 DUM-LUN1 50.000* 56 54463 LUND1 130.00 56 1 77.0 55.0 140.0
50495 DUM-T1- 130.00 65 57015 AURA-TEA 17.000* 65 1 77.3 69.6 111.1
50563 DUM-KLE2 66.000 66 57032 KLEBU3 300.00* 66 1 158.1 157.4 100.5
50574 DUM-FAN2 130.00 62 55275 FAN-T2-T 17.000* 62 1 59.0 58.2 101.5
50588 DUM-STOR 66.000 61 55172 STORD3 300.00* 61 1 109.6 99.3 110.3
50605 DUM-FLE1 66.000 55 51203 FLESAK1A 130.00* 55 1 104.3 75.4 138.3
50607 DUM-FLE3 130.00 55 51203 FLESAK1A 130.00* 55 1 108.5 106.0 102.3
50608 DUM-FLE4 130.00 55 51202 FLESAK3 300.00* 55 1 125.4 116.0 108.1
50608 DUM-FLE4 130.00 55 51203 FLESAK1A 130.00* 55 1 119.1 106.0 112.4
50635 DUM-MYKS 11.000* 53 51393 MYKSTF1A 130.00 53 1 56.5 49.7 113.6
50637 DUM-POR1 130.00 55 51244 PORSGR0 17.000* 55 1 108.9 100.0 108.9
50638 DUM-POR2 130.00 55 51245 PORSGRKA 17.000* 55 1 109.0 100.0 109.0
50654 DUM-T7- 130.00 51 51132 SMESTAD3 300.00* 51 1 273.8 246.2 111.2
50654 DUM-T7- 130.00 51 51137 SMESTD1B 130.00* 51 1 265.1 246.2 107.7
50678 DUM-ULV1 50.000 51 51102 ULVEN3 300.00* 51 1 150.8 138.6 108.8
50679 DUM-ULV2 50.000 51 51102 ULVEN3 300.00* 51 1 151.2 138.6 109.1
50751 DUM-RJUK 130.00 54 51711 RJUKAN 400.00* 54 1 298.8 282.6 105.7
50751 DUM-RJUK 130.00 54 51713 RJUKAN-T 130.00* 54 1 295.5 282.6 104.6
51180 SKOYEN 11.000 51 51183 SKOYEN 130.00* 51 1 45.9 39.4 116.4
51180 SKOYEN 11.000 51 51183 SKOYEN 130.00* 51 2 45.9 39.4 116.4
51403 NORE2-A- 130.00 53 51407 NORE2-G2 11.600* 53 1 29.5 28.7 102.8
51483 MAR-A- 130.00 55 51485 MAR-G1- 6.0000* 55 1 38.4 36.1 106.5
51483 MAR-A- 130.00 55 51486 MAR-G2- 6.0000* 55 1 39.0 36.1 108.0
51483 MAR-A- 130.00 55 51487 MAR-G3- 6.0000* 55 1 40.2 36.1 111.3
51484 MAR-B- 130.00 55 51488 MAR-G4- 6.0000* 55 1 39.6 36.1 109.6
51484 MAR-B- 130.00 55 51489 MAR-G5- 6.0000* 55 1 39.6 36.1 109.6
51484 MAR-B- 130.00 55 51718 RJUKAN-B 130.00* 54 1 292.2 218.0 134.1
51713 RJUKAN-T 130.00* 54 51718 RJUKAN-B 130.00 54 1 295.5 184.2 160.4
52052 FABERG3 300.00* 52 52072 N.VINST3 300.00 52 1 312.2 311.8 100.2
52062 RENDAL3 300.00 52 52066 RENDA-G1 14.000* 52 1 106.7 106.1 100.6
52307 DOKKA-G2 8.0000* 52 52309 DOKK-KRV 130.00 52 2 23.2 23.2 100.1
52323 ABJORA1 130.00 52 52329 AABJ-G4- 11.000* 52 1 100.9 94.3 107.0
53133 UVDAL1-1 130.00 53 53136 UVDA1-G1 8.0000* 53 1 49.2 46.4 106.0
53133 UVDAL1-1 130.00 53 53137 UVDA1-G2 8.0000* 53 1 49.7 46.4 107.1
54022 KR.SAND3 300.00* 56 54025 KR.SAND 110.00 56 3 114.9 110.0 104.5
54152 TJORHOM3 300.00 57 54156 TJORHOM 13.000* 57 1 134.3 131.3 102.3
54193 LYSEBTN1 130.00 57 54197 LYSEB-G4 9.0000* 57 4 48.4 46.4 104.3
54503 AKLAND1 130.00 56 54543 HOLT1 130.00* 56 1 91.5 45.0 203.2
55112 NESFLAT3 300.00 61 55118 SULD2-G1 10.000* 61 1 76.9 76.2 100.9
55825 BJOL-G1 12.000 62 55829 NYEBJOLV 130.00* 62 1 101.6 95.7 106.1
56032 LEIRDOL3 300.00 63 56042 FORTUN3 300.00* 63 1 272.6 259.8 104.9
56403 MOSKOG 130.00* 64 56404 STAK-66A 66.000 64 1 65.2 49.2 132.6
56743 TAF-K4- 130.00 65 56746 TA-K4-G1 11.000* 65 1 55.2 53.1 104.0
57005 SUNNDALS 20.000 65 57013 AURA1 130.00* 65 1 23.9 22.2 107.5
57053 NEA1 130.00* 66 57054 NEA66- 66.000 66 1 29.4 27.8 105.8
57132 TUNSJDAL 300.00 67 57136 TUNN-G12 8.5000* 67 1 88.8 85.7 103.6
57132 TUNSJDAL 300.00 67 57137 TUNN-G34 8.5000* 67 1 89.4 85.7 104.3
57203 N.ROSS1A 130.00 67 57214 NROSS-G1 11.000* 67 1 44.8 44.8 100.0
57203 N.ROSS1A 130.00 67 57215 NROSS-G2 11.000* 67 1 44.8 44.8 100.0
57883 LOMI1 130.00 67 57887 LOMI-G2- 10.500* 67 1 63.9 62.8 101.8
58463 KVALOYA1 130.00 69 58464 KVALOY66 66.000* 69 1 48.7 47.1 103.5
58463 KVALOYA1 130.00 69 58464 KVALOY66 66.000* 69 2 49.0 47.1 104.0
58683 TANA1 130.00* 69 58685 TANA22 20.000 69 1 10.1 8.3 121.3
1700 MW import of HVDC, 1600 MW load at Hasle
--------------------------------------------------------------------------------
PTI INTERACTIVE POWER SYSTEM SIMULATOR--PSS/E WED, MAY 14 2008 15:01

50419 DUM-KVA2 130.00 68 1.1121 144.57 50556 DUM-T8- 220.00 67 1.1250 247.51
57200 NR-RAZFE 220.00 67 1.1250 247.49 57204 N.ROSS2 220.00 67 1.1250 247.49
57299 SUNDFJ-B 130.00 67 1.1035 143.46 57683 GLOMFJD1 130.00 67 1.1022 143.29
58049 KVAN-NOR 130.00 68 1.1120 144.56 58353 STRAUM1A 130.00 69 1.1075 143.97
58363 INNSET1 130.00 69 1.1119 144.54 58702 VARANGER 220.00 69 1.1169 245.72
51092 ROYKAS3 300.00 51 0.8957 268.72 51102 ULVEN3 300.00 51 0.8892 266.76
51112 FURUSET3 300.00 51 0.8913 267.40 51122 SOGN3 300.00 51 0.8857 265.72
51123 SOGN1 130.00 51 0.8879 115.43 51132 SMESTAD3 300.00 51 0.8838 265.14
51137 SMESTD1B 130.00 51 0.8443 109.76 51140 HAM-REAK 300.00 51 0.8876 266.27
51177 SAGENE 130.00 51 0.8855 115.12 51178 PRESTEGA 130.00 51 0.8416 109.41
51179 BRISKEBY 130.00 51 0.8428 109.56 51182 BERUM3 300.00 51 0.8853 265.60
51183 SKOYEN 130.00 51 0.8437 109.69 51196 AKERSBER 130.00 51 0.8799 114.39
51302 HADELAND 300.00 51 0.8999 269.97 52012 MINNE3 300.00 52 0.8992 269.77
52403 SAVALEN1 130.00 52 0.8859 115.17 52413 TYNSET1 130.00 52 0.8734 113.54
52423 ENGERD1 130.00 52 0.8963 116.52 52433 TRYSIL1 130.00 52 0.8909 115.82
57548 ULSET-B- 130.00 66 0.8877 115.41
and 400 kV
1700 MW import of HVDC, 1600 MW load at Hasle
--------------------------------------------------------------------------------


50029 KKI-THA2 20.000* 66 56573 THAMSHAM 130.00 66 1 37.1 29.1 127.5
50032 KKI-FES1 20.000 67 57623 SVABO-A- 130.00* 67 5 47.2 46.4 101.8
50280 MF-O.VIN 300.00 52 50539 DUM-VINS 12.000* 52 1 163.6 142.9 114.5
50483 DUM-LUN1 50.000* 56 54463 LUND1 130.00 56 1 78.3 55.0 142.4
50492 DUM-HAVI 66.000* 61 55152 HAAVIK3 300.00 61 1 80.1 79.5 100.7
50495 DUM-T1- 130.00 65 57015 AURA-TEA 17.000* 65 1 77.2 69.6 110.9
50563 DUM-KLE2 66.000 66 57032 KLEBU3 300.00* 66 1 158.4 157.4 100.6
50574 DUM-FAN2 130.00 62 55275 FAN-T2-T 17.000* 62 1 59.0 58.2 101.3
50588 DUM-STOR 66.000 61 55172 STORD3 300.00* 61 1 109.1 99.3 109.8
50605 DUM-FLE1 66.000 55 51203 FLESAK1A 130.00* 55 1 91.5 75.4 121.4
50607 DUM-FLE3 130.00* 55 51202 FLESAK3 300.00 55 1 117.4 116.0 101.2
50607 DUM-FLE3 130.00 55 51203 FLESAK1A 130.00* 55 1 117.4 106.0 110.8
50608 DUM-FLE4 130.00 55 51202 FLESAK3 300.00* 55 1 118.1 116.0 101.8
50608 DUM-FLE4 130.00* 55 51203 FLESAK1A 130.00 55 1 112.2 106.0 105.9
50619 DUM-HAM1 50.000 51 51142 HAMANG3 300.00* 51 1 134.7 133.0 101.3
50620 DUM-HAM2 50.000 51 51142 HAMANG3 300.00* 51 1 136.4 133.0 102.6
50635 DUM-MYKS 11.000* 53 51393 MYKSTF1A 130.00 53 1 57.9 49.7 116.6
50637 DUM-POR1 130.00 55 51244 PORSGR0 17.000* 55 1 107.4 100.0 107.4
50638 DUM-POR2 130.00 55 51245 PORSGRKA 17.000* 55 1 107.5 100.0 107.5
50641 DUM-RING 130.00 51 51161 RINGERI4 400.00* 51 1 248.0 246.2 100.7
50652 DUM-T4- 50.000* 51 51132 SMESTAD3 300.00 51 1 156.4 154.4 101.3
50653 DUM-T5- 50.000* 51 51132 SMESTAD3 300.00 51 1 157.1 154.4 101.7
50654 DUM-T7- 130.00 51 51132 SMESTAD3 300.00* 51 1 290.4 246.2 117.9
50654 DUM-T7- 130.00 51 51137 SMESTD1B 130.00* 51 1 277.0 246.2 112.5
50678 DUM-ULV1 50.000 51 51102 ULVEN3 300.00* 51 1 157.1 138.6 113.3
50679 DUM-ULV2 50.000 51 51102 ULVEN3 300.00* 51 1 157.4 138.6 113.5
50751 DUM-RJUK 130.00 54 51711 RJUKAN 400.00* 54 1 303.6 282.6 107.4
50751 DUM-RJUK 130.00* 54 51713 RJUKAN-T 130.00 54 1 300.3 282.6 106.3
51081 FROGNER4 400.00 51 51086 FROGN-F1 17.000* 51 1 231.5 220.8 104.8
51180 SKOYEN 11.000 51 51183 SKOYEN 130.00* 51 1 48.0 39.4 121.7
51180 SKOYEN 11.000 51 51183 SKOYEN 130.00* 51 2 48.0 39.4 121.7
51276 GOKSTAD 11.000 55 51278 GOKSTAD1 130.00* 55 1 24.3 23.9 101.8
51277 HOLMESTR 20.000 55 51408 HOLMESI1 130.00* 55 1 26.3 26.1 100.8
51403 NORE2-A- 130.00 53 51407 NORE2-G2 11.600* 53 1 30.2 28.7 105.3
51404 SLAGEN22 20.000 55 51418 SLAGEN1 130.00* 55 1 26.3 26.1 100.8
51483 MAR-A- 130.00 55 51485 MAR-G1- 6.0000* 55 1 39.5 36.1 109.3
51483 MAR-A- 130.00 55 51486 MAR-G2- 6.0000* 55 1 39.4 36.1 109.2
51483 MAR-A- 130.00 55 51487 MAR-G3- 6.0000* 55 1 40.7 36.1 112.6
51484 MAR-B- 130.00 55 51488 MAR-G4- 6.0000* 55 1 39.3 36.1 109.0
51484 MAR-B- 130.00 55 51489 MAR-G5- 6.0000* 55 1 39.3 36.1 109.0
51484 MAR-B- 130.00 55 51718 RJUKAN-B 130.00* 54 1 297.0 218.0 136.2
51693 HOLTANE1 130.00 55 54507 GJERST_T 130.00* 56 1 151.4 112.6 134.5
51713 RJUKAN-T 130.00* 54 51718 RJUKAN-B 130.00 54 1 300.3 184.2 163.0
52052 FABERG3 300.00* 52 52072 N.VINST3 300.00 52 1 325.7 311.8 104.5
52062 RENDAL3 300.00 52 52066 RENDA-G1 14.000* 52 1 111.4 106.1 105.0
52307 DOKKA-G2 8.0000* 52 52309 DOKK-KRV 130.00 52 2 23.9 23.2 103.0
52323 ABJORA1 130.00 52 52329 AABJ-G4- 11.000* 52 1 101.5 94.3 107.7
53133 UVDAL1-1 130.00 53 53136 UVDA1-G1 8.0000* 53 1 49.5 46.4 106.8
53133 UVDAL1-1 130.00 53 53137 UVDA1-G2 8.0000* 53 1 50.1 46.4 107.9
54020 KR.SA-F1 17.000* 56 54022 KR.SAND3 300.00 56 1 125.9 125.8 100.1
54022 KR.SAND3 300.00* 56 54025 KR.SAND 110.00 56 3 118.8 110.0 108.0
54152 TJORHOM3 300.00 57 54156 TJORHOM 13.000* 57 1 138.8 131.3 105.7
54193 LYSEBTN1 130.00 57 54197 LYSEB-G4 9.0000* 57 4 51.3 46.4 110.5
54503 AKLAND1 130.00* 56 54507 GJERST_T 130.00 56 1 160.1 135.1 118.5
54503 AKLAND1 130.00 56 54543 HOLT1 130.00* 56 1 136.6 45.0 303.4
55112 NESFLAT3 300.00 61 55118 SULD2-G1 10.000* 61 1 77.9 76.2 102.3
55825 BJOL-G1 12.000 62 55829 NYEBJOLV 130.00* 62 1 101.4 95.7 105.9
56032 LEIRDOL3 300.00 63 56042 FORTUN3 300.00* 63 1 273.0 259.8 105.1
56356 GRYTT-G1 16.000* 65 56359 GR.KRV 130.00 65 1 143.6 143.5 100.1
56403 MOSKOG 130.00* 64 56404 STAK-66A 66.000 64 1 65.3 49.2 132.7
56743 TAF-K4- 130.00 65 56746 TA-K4-G1 11.000* 65 1 55.3 53.1 104.1
57005 SUNNDALS 20.000 65 57013 AURA1 130.00* 65 1 23.9 22.2 107.8
57053 NEA1 130.00* 66 57054 NEA66- 66.000 66 1 29.4 27.8 105.7
57132 TUNSJDAL 300.00 67 57136 TUNN-G12 8.5000* 67 1 88.8 85.7 103.6
57132 TUNSJDAL 300.00 67 57137 TUNN-G34 8.5000* 67 1 89.4 85.7 104.3
57203 N.ROSS1A 130.00 67 57214 NROSS-G1 11.000* 67 1 44.8 44.8 100.0
57203 N.ROSS1A 130.00 67 57215 NROSS-G2 11.000* 67 1 44.8 44.8 100.0
57883 LOMI1 130.00 67 57887 LOMI-G2- 10.500* 67 1 63.9 62.8 101.8
58463 KVALOYA1 130.00 69 58464 KVALOY66 66.000* 69 1 48.7 47.1 103.5
58463 KVALOYA1 130.00 69 58464 KVALOY66 66.000* 69 2 49.0 47.1 104.0
58683 TANA1 130.00* 69 58685 TANA22 20.000 69 1 10.1 8.3 121.3
1000 MW offshore wind connected to Kårstø, 900 MW load at Hasle
--------------------------------------------------------------------------------

50419 DUM-KVA2 130.00 68 1.1121 144.57 50556 DUM-T8- 220.00 67 1.1250 247.51
57200 NR-RAZFE 220.00 67 1.1250 247.49 57204 N.ROSS2 220.00 67 1.1250 247.49
57299 SUNDFJ-B 130.00 67 1.1035 143.46 57683 GLOMFJD1 130.00 67 1.1022 143.29
58049 KVAN-NOR 130.00 68 1.1120 144.56 58353 STRAUM1A 130.00 69 1.1075 143.97
58363 INNSET1 130.00 69 1.1119 144.54 58702 VARANGER 220.00 69 1.1169 245.72
51179 BRISKEBY 130.00 51 0.8696 113.04 51183 SKOYEN 130.00 51 0.8705 113.16
52403 SAVALEN1 130.00 52 0.8905 115.76 52413 TYNSET1 130.00 52 0.8780 114.13
52433 TRYSIL1 130.00 52 0.8960 116.47 57548 ULSET-B- 130.00 66 0.8923 116.00
and 400 kV
1000 MW offshore wind connected to Kårstø, 900 MW load at Hasle
--------------------------------------------------------------------------------


50029 KKI-THA2 20.000* 66 56573 THAMSHAM 130.00 66 1 37.1 29.1 127.6
50032 KKI-FES1 20.000 67 57623 SVABO-A- 130.00* 67 5 47.2 46.4 101.8
50280 MF-O.VIN 300.00 52 50539 DUM-VINS 12.000* 52 1 161.6 142.9 113.1
50483 DUM-LUN1 50.000* 56 54463 LUND1 130.00 56 1 77.1 55.0 140.2
50495 DUM-T1- 130.00 65 57015 AURA-TEA 17.000* 65 1 77.3 69.6 111.1
50563 DUM-KLE2 66.000 66 57032 KLEBU3 300.00* 66 1 158.1 157.4 100.5
50574 DUM-FAN2 130.00 62 55275 FAN-T2-T 17.000* 62 1 59.2 58.2 101.7
50588 DUM-STOR 66.000 61 55172 STORD3 300.00* 61 1 110.3 99.3 111.1
50605 DUM-FLE1 66.000 55 51203 FLESAK1A 130.00* 55 1 102.8 75.4 136.3
50607 DUM-FLE3 130.00 55 51203 FLESAK1A 130.00* 55 1 109.3 106.0 103.1
50608 DUM-FLE4 130.00 55 51202 FLESAK3 300.00* 55 1 124.1 116.0 107.0
50608 DUM-FLE4 130.00 55 51203 FLESAK1A 130.00* 55 1 117.9 106.0 111.2
50635 DUM-MYKS 11.000* 53 51393 MYKSTF1A 130.00 53 1 56.6 49.7 113.9
50637 DUM-POR1 130.00 55 51244 PORSGR0 17.000* 55 1 109.0 100.0 109.0
50638 DUM-POR2 130.00* 55 51245 PORSGRKA 17.000 55 1 109.1 100.0 109.1
50654 DUM-T7- 130.00 51 51132 SMESTAD3 300.00* 51 1 275.4 246.2 111.8
50654 DUM-T7- 130.00 51 51137 SMESTD1B 130.00* 51 1 266.7 246.2 108.3
50678 DUM-ULV1 50.000 51 51102 ULVEN3 300.00* 51 1 151.4 138.6 109.2
50679 DUM-ULV2 50.000 51 51102 ULVEN3 300.00* 51 1 151.7 138.6 109.5
50751 DUM-RJUK 130.00 54 51711 RJUKAN 400.00* 54 1 298.8 282.6 105.7
50751 DUM-RJUK 130.00 54 51713 RJUKAN-T 130.00* 54 1 295.5 282.6 104.6
51180 SKOYEN 11.000 51 51183 SKOYEN 130.00* 51 1 46.1 39.4 117.1
51180 SKOYEN 11.000 51 51183 SKOYEN 130.00* 51 2 46.1 39.4 117.1
51403 NORE2-A- 130.00 53 51407 NORE2-G2 11.600* 53 1 29.5 28.7 102.9
51483 MAR-A- 130.00 55 51485 MAR-G1- 6.0000* 55 1 38.4 36.1 106.5
51483 MAR-A- 130.00 55 51486 MAR-G2- 6.0000* 55 1 39.0 36.1 108.0
51483 MAR-A- 130.00 55 51487 MAR-G3- 6.0000* 55 1 40.2 36.1 111.3
51484 MAR-B- 130.00 55 51488 MAR-G4- 6.0000* 55 1 39.6 36.1 109.6
51484 MAR-B- 130.00 55 51489 MAR-G5- 6.0000* 55 1 39.6 36.1 109.6
51484 MAR-B- 130.00 55 51718 RJUKAN-B 130.00* 54 1 292.3 218.0 134.1
51713 RJUKAN-T 130.00* 54 51718 RJUKAN-B 130.00 54 1 295.5 184.2 160.4
52052 FABERG3 300.00* 52 52072 N.VINST3 300.00 52 1 313.1 311.8 100.4
52062 RENDAL3 300.00 52 52066 RENDA-G1 14.000* 52 1 107.0 106.1 100.8
52307 DOKKA-G2 8.0000* 52 52309 DOKK-KRV 130.00 52 2 23.3 23.2 100.3
52323 ABJORA1 130.00 52 52329 AABJ-G4- 11.000* 52 1 101.0 94.3 107.1
53133 UVDAL1-1 130.00 53 53136 UVDA1-G1 8.0000* 53 1 49.2 46.4 106.1
53133 UVDAL1-1 130.00 53 53137 UVDA1-G2 8.0000* 53 1 49.7 46.4 107.2
54022 KR.SAND3 300.00* 56 54025 KR.SAND 110.00 56 3 120.5 110.0 109.5
54152 TJORHOM3 300.00 57 54156 TJORHOM 13.000* 57 1 134.4 131.3 102.3
54193 LYSEBTN1 130.00 57 54197 LYSEB-G4 9.0000* 57 4 48.5 46.4 104.5
54503 AKLAND1 130.00 56 54543 HOLT1 130.00* 56 1 64.0 45.0 142.2
55112 NESFLAT3 300.00 61 55118 SULD2-G1 10.000* 61 1 76.6 76.2 100.5
55825 BJOL-G1 12.000 62 55829 NYEBJOLV 130.00* 62 1 101.8 95.7 106.4
56032 LEIRDOL3 300.00 63 56042 FORTUN3 300.00* 63 1 272.7 259.8 105.0
56403 MOSKOG 130.00* 64 56404 STAK-66A 66.000 64 1 65.2 49.2 132.6
56743 TAF-K4- 130.00 65 56746 TA-K4-G1 11.000* 65 1 55.2 53.1 104.0
57005 SUNNDALS 20.000 65 57013 AURA1 130.00* 65 1 23.9 22.2 107.6
57053 NEA1 130.00* 66 57054 NEA66- 66.000 66 1 29.4 27.8 105.8
57132 TUNSJDAL 300.00 67 57136 TUNN-G12 8.5000* 67 1 88.8 85.7 103.6
57132 TUNSJDAL 300.00 67 57137 TUNN-G34 8.5000* 67 1 89.4 85.7 104.3
57203 N.ROSS1A 130.00 67 57214 NROSS-G1 11.000* 67 1 44.8 44.8 100.0
57203 N.ROSS1A 130.00 67 57215 NROSS-G2 11.000* 67 1 44.8 44.8 100.0
57883 LOMI1 130.00 67 57887 LOMI-G2- 10.500* 67 1 63.9 62.8 101.8
58463 KVALOYA1 130.00 69 58464 KVALOY66 66.000* 69 1 48.7 47.1 103.5
58463 KVALOYA1 130.00 69 58464 KVALOY66 66.000* 69 2 49.0 47.1 104.0
58683 TANA1 130.00* 69 58685 TANA22 20.000 69 1 10.1 8.3 121.3
Appendix B: Short-circuit power calculations
The short-circuit power can be found in two ways from the bottom line of the print from
PSS/E.
FAULT MVA A [MW] B [MVAR] C [A]
1:
Fault MVA A2 B 2
2:
Fault MVA 3 300 C 103
Case 1
--------------------------------------------------------------------------------
PTI INTERACTIVE POWER SYSTEM SIMULATOR--PSS/E TUE, JUN 03 2008 10:21

X------- FROM BUS ------X AREA VOLT GEN LOAD SHUNT X---------- TO BUS ----------X TRANSFORMER
RATING
BUS# X-- NAME --X BASKV ZONE PU/KV ANGLE MW/MVAR MW/MVAR MW/MVAR BUS# X-- NAME --X BASKV AREA
CKT MW MVAR RATIO ANGLE AMPS %I SET A
54032 FEDA3 300.00 57 0.0000 156.6 0.0 0.0 0.0 ---------------------------------------------------------------------------------

510 0.0011 0.0 0.0 0.0 54022 KR.SAND3 300.00 56 1 0.0 0.0 3604 180 2000A
54042 LISTA3 300.00 57 1 0.0 0.0 0 0 500A
54042 LISTA3 300.00 57 2 0.0 0.0 0 0 500A
54052 ANA-SIR3 300.00 57 1 0.0 0.0 1811 145 1250A
54102 TONSTAD3 300.00 57 1 0.0 0.0 3378 186 1815A
54102 TONSTAD3 300.00 57 2 0.0 0.0 4474 224 2000A
54592 OIE 300.00 57 1 0.0 0.0 585 59 1000A
FAULT MVA 2839.2 6558.3 13753
Case 2
--------------------------------------------------------------------------------

RATING
54032 FEDA3 300.00 57 0.0000 160.2 0.0 0.0 0.0 ---------------------------------------------------------------------------------

510 0.0008 0.0 0.0 0.0 54022 KR.SAND3 300.00 56 1 0.0 0.0 2974 149 2000A
54042 LISTA3 300.00 57 1 0.0 0.0 0 0 500A
54042 LISTA3 300.00 57 2 0.0 0.0 0 0 500A
54052 ANA-SIR3 300.00 57 1 0.0 0.0 1680 134 1250A
54102 TONSTAD3 300.00 57 1 0.0 0.0 2106 116 1815A
54102 TONSTAD3 300.00 57 2 0.0 0.0 2789 139 2000A
54592 OIE 300.00 57 1 0.0 0.0 538 54 1000A
FAULT MVA 1774.5 4917.5 10061
XI
Case 3
--------------------------------------------------------------------------------

RATING
54032 FEDA3 300.00 57 0.0000 166.9 0.0 0.0 0.0 ---------------------------------------------------------------------------------

510 0.0010 0.0 0.0 0.0 54022 KR.SAND3 300.00 56 1 0.0 0.0 3379 169 2000A
54042 LISTA3 300.00 57 1 0.0 0.0 0 0 500A
54042 LISTA3 300.00 57 2 0.0 0.0 0 0 500A
54052 ANA-SIR3 300.00 57 1 0.0 0.0 1746 140 1250A
54102 TONSTAD3 300.00 57 1 0.0 0.0 3276 180 1815A
54102 TONSTAD3 300.00 57 2 0.0 0.0 4338 217 2000A
54592 OIE 300.00 57 1 0.0 0.0 580 58 1000A
FAULT MVA 1560.3 6721.4 13279
Case 4
--------------------------------------------------------------------------------

RATING
55232 KÅRSTØ 300.00 61 0.0000 152.6 0.0 0.0 0.0 ---------------------------------------------------------------------------------

512 0.0006 0.0 0.0 0.0 50070 IND-KARS 20.000 61 1 0.0 0.0 0.935RG 65 87 39M
50070 IND-KARS 20.000 61 2 0.0 0.0 0.935RG 67 89 39M
50070 IND-KARS 20.000 61 3 0.0 0.0 0.945RG 65 74 46M
50070 IND-KARS 20.000 61 4 0.0 0.0 0.947RG 65 74 46M
55102 SAUDA3 300.00 61 1 0.0 0.0 3554 239 1490A
55152 HAAVIK3 300.00 61 1 0.0 0.0 3890 214 1815A
55233 KÅRSTØ 130.00 61 1 0.0 0.0 0.985RG 0 0 172M
FAULT MVA 1832.1 3539.8 7671
Case 5
--------------------------------------------------------------------------------

RATING
55232 KÅRSTØ 300.00 61 0.0000 111.9 0.0 0.0 0.0 ---------------------------------------------------------------------------------

512 0.0005 0.0 0.0 0.0 50070 IND-KARS 20.000 61 1 0.0 0.0 0.815RG 70 93 39M
50070 IND-KARS 20.000 61 2 0.0 0.0 0.815RG 71 95 39M
50070 IND-KARS 20.000 61 3 0.0 0.0 0.827RG 69 79 46M
50070 IND-KARS 20.000 61 4 0.0 0.0 0.909LO 63 72 46M
55102 SAUDA3 300.00 61 1 0.0 0.0 3190 214 1490A
55152 HAAVIK3 300.00 61 1 0.0 0.0 3445 190 1815A
55233 KÅRSTØ 130.00 61 1 0.0 0.0 0.876RG 0 0 172M
FAULT MVA 3300.4 1323.7 6844
XII
Appendix C: Establishing of load flow cases
Case 2
Connection point of offshore wind power: Feda (54032)
The production and load consumption in the system is scaled down. The load is not scaled
down at busbars starting with KKI- or IND-. These busbars represent industrial loads. The
load is scaled down with a constant P/Q ratio. Generators are disconnected from the system.
Because of difficulties of doing drastic changes throughout the entire system, only the areas
55, 56, 57, 60 and 61 are chosen for the scaling. The result is not a realistic situation, but the
areas surrounding the connection point of the offshore wind power have a low load with
generators disconnected. Generators are given by the busbar number of the generator. The
short circuit power at the connection point will be reduced as a consequence of the
disconnection of generators.
x Load is scaled down: -500 MW

x Generators are disconnected: 51204, 51356, 51507, 54106 - 54108
x Generators are disconnected: 54109, 54116, 54117
x Generators are disconnected: 54126, 54146, 54147, 54156 (id 1 and 2)
x Generators are disconnected: 54186, 54246, 54326 - 54329
x Generators are disconnected: 54406, 54585, 54619, 55026 - 55029
x Generators are disconnected: 55056 - 55058
x Generator is disconnected: 55146
The offshore wind power equals 1000 MW.

No import or export of HVDC.
Export of 900 MW to Sweden through Hasle (51031)
XIII
Case 3
Connection point of offshore wind power: Feda (54032)
The production in the system is scaled down. The load consumption in the system is at its
maximum. No generators are disconnected.
x Production in all areas is scaled down: -500 MW

x Value of load at Kristiansand (55032): -500 + j 250 MVA
x Production in all areas is scaled down: -100 MW
x New value at (53206 - 53207): 200 MW each
x New value at (54109): 150 MW
x New value of load at Kristiansand (55032): -1000 + j 500 MVA
The offshore wind power equals 1000 MW.

1000 MW of HVDC from Denmark.
No export of power to Sweden.
XIV
Case 5
Connection point of offshore wind power: Kårstø (55232)
The generation in the surrounding area of the connection point is reduced by disconnection
generators. Generators are given by the busbar number of the generator. The load is not scaled
down.
x Generators are disconnected: 55026 - 55029 (Saurdal)

x Generators are disconnected: 55076 - 55077 (Hylen)
x Generators are disconnected: 55104 (id 1 - 4) (Sauda)
x Generators are disconnected: 55194 (id 1 - 4) (Blåfall)
x Generators are disconnected: 55206 - 55207 (Mauranger)
x Generators are disconnected: 55226 - 55227 (Blåfall)
x Generators is disconnected: 55376 (Stord)
The offshore wind power equal 1000 MW.

No import or export of HVDC.
No export of power to Sweden.
XV
Appendix D: Dynamic simulation results
Summary of load flow cases and events:
Summary of disturbances for dynamic simulations:
System frequency: 50 Hz
System base: 1000 MVA
XVI
Case 1, AC cable
Event 1, bus fault at Øie
Voltage at Feda [pu]
Voltage offshore [pu]
XVII
Speed deviation of wind turbine generators [pu]
Reactive power from one SVC [pu on system base]
XVIII
Event 2, line fault and tripping of line
XIX
XX
Event 3, three-phase short-circuit at Feda
XXI
Case 1, HVDC Light
XXII
DC voltage at the voltage regulating HVDC Light converter [kV]
XXIII
Reactive power from the HVDC Light converter at the connection point [pu on system base]
XXIV
XXV
XXVI
XXVII
XXVIII
XXIX
XXX
Case 2, AC cable
XXXI
XXXII
XXXIII
XXXIV
XXXV
XXXVI
Case 2, HVDC Light
XXXVII
XXXVIII
XXXIX
XL
XLI
XLII
XLIII
XLIV
XLV
Case 3, AC cable
XLVI
XLVII
XLVIII
XLIX
L
Case 3, HVDC Light
LI
LII
LIII
LIV
LV
LVI
LVII
LVIII
LIX
Case 4, AC cable
Event 1, bus fault at Sauda
Voltage at Kårstø [pu]
LX
LXI
LXII
Event 3, three-phase short-circuit at Kårstø
LXIII
Case 4, HVDC Light
LXIV
LXV
LXVI
LXVII
LXVIII
LXIX
LXX
LXXI
LXXII
Case 5, AC cable
LXXIII
LXXIV
LXXV
Case 5, HVDC Light
LXXVI
LXXVII
LXXVIII
LXXIX
LXXX
LXXXI
LXXXII
LXXXIII
LXXXIV

PSSE Offshore Wind HVDC

Uploaded by

Copyright:

Available Formats

PSSE Offshore Wind HVDC

Uploaded by

Document Information

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

PSSE Offshore Wind HVDC

Uploaded by

Copyright:

Available Formats

Offshore Wind Power in the North Sea

Grid Integration of 1000 MW Offshore Wind Power into the

Knut Magnus Sommerfelt

Master of Science in Energy and Environment

Norwegian University of Science and Technology

Assignment given: 15. January 2008

Knut Magnus Sommerfelt

2.1 Overhead lines

Table 2.1: Typical parameters for overhead transmission lines [12]

Table 2.2: Typical parameters for cables (PILC) [12]

2.3 Surge Impedance Loading

Transmission lines and cables can be described by their characteristic impedance ZC

2.4 Strength of AC systems

A normal parameter for measuring the strength at a busbar in an AC network is short-circuit

In fault studies, a generator is represented as Norton equivalent, with a current source in

In order to connect a HVDC link based on thyristor technology to an AC system, two

total rotational inertia of the AC system [MW  s]

The inertia constant is recommended to be at least 2 - 3 s in order to have satisfactory

3 Background for HVDC

3.1 HVDC in general

Figure 3.1: Maximal transmitted power through an AC-cable [4]

3.2 HVDC Light

4 VSC HVDC technology

4.1 Single phase VSC

Figure 4.1: Block diagram of a single-phase inverter [17]

Figure 4.3: One-leg switch mode inverter [17]

Figure 4.4: Pulse-width modulation of a single-phase converter [17]

4.2 Three-phase VSC

Figure 4.5: Three-phase switch-mode converter [17]

Figure 4.6: Pulse-width modulation of a three-phase converter [17]

4.3 Power flow in Voltage Source Converters

Consider a system topology as shown in Figure 4.7.

Figure 4.7: System topology of a VSC HVDC system [5]

ªV cos(G )  VC º VCVS cos(G )  VC2

VS2  VSVC cos(G )

4.4 Control modes

4.4.1 Normal operation

4.4.2 Offshore wind power

5.1 The Norwegian model

5.2 Wind turbine model

5.3 The HVDC Light model

Table 5.1: Modules in HVDC Light [1]

5.3.2 Power flow representation

Figure 5.1: PQ diagram of a HVDC Light converter [1]

Ploss  Prectifier  Pinverter (5.1)

Table 5.2: Losses in HVDC Light [1]

Psending  Preceiving 1147.9  1110.1

In percentage, the converter losses for the modules M7 to M9 are identical.

Ploss ,cable Rdc  I dc2 7.08 16352 | 19 MW (5.4)

5.3.4 Dynamic model

Figure 5.2: HVDC Light model in PSS/E [2]

5.3.5 Offshore wind power application

5.3.6 Time step

6 Load flow cases

6.1 Connection point Feda

Figure 6.2: Load flow result at Feda, original load flow

6.1.1 Feda with HVDC from the Netherlands and Denmark

6.2 Connection point Kårstø

Figure 6.5: Overview of the 300 kV grid in the southwest of Norway

total rotational inertia of the AC system [MW s]

ªV cos(G ) VC º VCVS cos(G ) VC2

VS2 VSVC cos(G )

Ploss Prectifier Pinverter (5.1)

Psending Preceiving 1147.9 1110.1

Ploss ,cable Rdc I dc2 7.08 16352 | 19 MW (5.4)

QSVC P tan(cos 1 (M )) 960 tan(cos 1 (0.95)) | 316Mvar (6.1)

U ac2 U ac2 (300 103 ) 2