VSC-HVDC_bases_network_reinforcement.

VSC-HVDC_bases_network_reinforcement.
VSC-HVDC based Network Reinforcement
Tamiru Woldeyesus Shire
Student No: 1386042
M. Sc. Thesis Electrical power Engineering
Thesis supervisor:
Prof.ir. L. van der Sluis (TUD)
Daily supervisors:
Dr.ir. G. C. Paap (TUD), ir. R. L. Hendriks (TUD),
ir. H. J. M. Arts ( STEDIN), P. Zonneveld ( STEDIN)
Delft University of Technology
Faculty of Electrical Engineering, Mathematics and Computer Science
High-voltage Components and Power Systems
(The work for this thesis has been carried out at STEDIN from October to May 2009)
May 2009
Acknowledgement
The research work is carried out at Network Solutions group, Asset Management
division at STEDIN in conjunction with Delft University of Technology.
First, I would like to express my deepest gratitude to my supervisors prof.ir. L. van
der Sluis, dr.ir. G. C. Paap, ir. H. J. M. Arts, ir. R. L. Hendriks, and P. Zonneveld for
there encouragement and guidance throughout the research work. Their readiness to
help, patience and valuable suggestions were highly appreciated.
I would also like to thank ir. E. J. Coster and dr.ir. A. M. van Voorden for the
indispensable discussions we had throughout the thesis work.
Last but not the least, I want to thank all Stedin B.V. employees for there kind
cooperation.
i
Contents
Acknowledgement .........................................................................................................i
Chapter 1
Introduction..........................................................................................1
1.1 Introduction...........................................................................................................1
1.2 Problem Definition...............................................................................................1
1.3 Outline of the thesis ..............................................................................................2
Chapter 2
2.1
2.2
2.3
2.4
2.5
Introduction...........................................................................................................4
Network model ......................................................................................................4
Network constraints..............................................................................................5
Network Reinforcement .......................................................................................6
HVDC transmission ..............................................................................................8
2.5.1 Arrangement of HVDC systems ...............................................................8
2.5.2 Classical HVDC Systems...........................................................................9
2.5.3 Voltage source converter (VSC) HVDC system....................................10
Chapter 3
3.1
3.2
3.3
3.4
Network constraints and reinforcement ............................................4
Design and operating principle of VSC-HVDC ..............................13
Operating principle of VSC-HVDC ..................................................................13
Capability chart of VSC -HVDC .......................................................................14
PWM ....................................................................................................................15
VSC-HVDC station components .......................................................................15
3.4.1 Converters ................................................................................................15
3.4.2 Converter size...........................................................................................16
3.4.3 Converter transformer ............................................................................16
3.4.4 Direct voltage............................................................................................17
3.4.5 DC capacitor.............................................................................................17
3.4.6 Phase reactor ............................................................................................18
3.4.7 AC filters...................................................................................................19
Chapter 4
Control system of VSC-HVDC .........................................................21
4.1 Introduction.........................................................................................................21
4.2 Direct control.......................................................................................................21
4.3 Vector control......................................................................................................22
4.3.1 Inner Current controller .........................................................................25
4.3.1.1 PWM converter.................................................................................25
4.3.1.2 System transfer function ..................................................................26
4.3.1.3 Regulator ...........................................................................................27
4.3.1.4 Control block diagram .....................................................................27
4.3.2 Outer controllers......................................................................................28
4.3.2.1 Direct voltage control .......................................................................28
4.3.2.2 Active and reactive power control...................................................30
4.3.2.3 Ac voltage control .............................................................................31
4.3.3 Limiting strategies ...................................................................................31
4.3.4 Controller integral windup .....................................................................32
4.3.5 Tuning of PI controllers ..........................................................................32
Chapter 5
Testing of VSC-HVDC model ...........................................................34
5.1 Introductions .......................................................................................................34
ii
5.2 Active power control...........................................................................................35
5.3 Reactive power control .......................................................................................36
5.4 AC voltage control ..............................................................................................37
5.5 Three phase short circuit at Grid side-1 ...........................................................38
5.6 Unbalanced fault conditions: SLGF at Grid side-1 .........................................39
5.7 Three phase short circuit at Grid side-2 ...........................................................40
Conclusion ..................................................................................................................42
Chapter 6
6.1
6.2
6.3
6.4
6.5
6.6
6.7
6.8
Simulation Results .............................................................................43
Introduction.........................................................................................................43
Static voltage stability.........................................................................................44
Effect of VSC- HVDC on rotor angle................................................................47
Short-circuit contribution of VSC-HVDC........................................................49
Loss of DG unit in the distribution grid ...........................................................51
Interaction with wind units ................................................................................54
STATCOM mode of operation ..........................................................................57
Voltage dip...........................................................................................................61
Chapter 7
Conclusions and Further Research ..................................................65
7.1 Conclusions and Recommendations..................................................................65
7.2 Further Research ................................................................................................66
Appendix A .................................................................................................................68
Appendix B .................................................................................................................71
Appendix C .................................................................................................................74
References...................................................................................................................75
iii
Chapter 1
Introduction
1.1 Introduction
Today’s power system operation has changed, which is mainly caused by the
liberalization of the energy market and the incorporation of distributed generation
(DG). The liberalization has led to unbundling of generation and transport of
electrical energy and the establishment of trading markets. Due to this unbundling the
energy flows in the network have become less predictable. Besides the liberalization
of the energy market the introduction of DG to the power system also has a large
influence on the power flow, especially in the distribution networks. It is to be
expected that the penetration level of DG will further increase in the near future. Most
of the DG units are connected to the medium voltage (MV) grid. The incorporation of
DG turns the passive MV grid into an active one. In this active grid some customers
not only consume electricity, but they also generate and if generation exceeds their
demand, they supply the network. In the active grid, the power flow will change from
unidirectional to bidirectional and this changes the traditional loading profile.
In order to secure the supply of power in present and future power systems, grid
operators are now starting to consider installing additional equipment to control the
power flow. A well known control device is the phase-shifting transformer. However,
new power electronic devices, better known as FACTS (Flexible AC Transmission
Systems), have been developed as well. Mainly due to the lack of proven record of
their reliability these devices have not yet been widely applied by the grid operators.
Normally the grid operators solve network constraints by installing additional
transmission lines and transformers, however, in some cases power flow controlling
devices and controllable DG can offer a solution to particularly challenging network
constraints.
The recent development in semiconductor and control equipment has made the highvoltage direct current transmission based on voltage sourced converters (VSCHVDC) feasible. Due to the use of VSC-technology and pulse width modulation
(PWM) the VSC-HVDC has numerous potential advantages such as controllable
short-circuit current contribution, and rapid and independent control of active and
reactive power (ability to absorb/deliver reactive power). With these advantages VSCHVDC can likely be used to solve network constraints efficiently, however, at the
expense of increased losses and investment costs depending on the network topology.
1.2 Problem Definition
In Fig. 1.1A a schematic overview of an existing transmission grid is shown. Two
separate 150 kV grids are connected to the Dutch 380 kV grid. Area 1 (150 kV) is a
greenhouse area including a large penetration of CHP-plants and Area 2 (150 kV)
feeds a 50 kV grid. Due to an outage in the 50 kV grid network constraints occur,
which lead to overloading of some of the remaining circuits or transformers and a
1
Chapter 1
Introduction
violation of voltage limits on some substations. In order to prevent major grid
reinforcements at several locations, the possibility of solving the network constraints
by coupling both areas 1 and 2 on 50 kV level via VSC-HVDC (Fig. 1.1B) is foreseen
as a realistic option and is the research topic of this thesis.
Fig. 1.1 Schematic overview of studied system
A model of a VSC based DC-link is developed. A mathematical model of the control
system is described for the VSC. A control system is developed combining inner
current loop controller and outer controllers. A vector control strategy is studied and
corresponding dynamic performance under step changes and system faults is
investigated. Furthermore, the performance of the model is compared with other
research works.
The dynamics of the studied power system including the VSC-HVDC is thoroughly
investigated through various simulation scenarios. The ability of the VSC-HVDC to
solve the network constraints and its interaction with the CHP and installed wind units
in Area 2 (50 kV) grid is scrutinized.
The research has been done using DIgSILENT PowerFactory software. In this
software package it is possible to represent and solve AC and DC systems
simultaneously. It includes transient analysis tools concerning short-, mid- and
long-term dynamics, with adaptive step-sizes ranging from milliseconds to minutes
[1]. Moreover, the software has a modular built-up and can be extended by user
written models in case the shipped model library proves insufficient.
1.3 Outline of the thesis
Chapter 2 presents the studied grid constraints and the various possible reinforcement
methods. The advantage and disadvantages of the reinforcement methods are
discussed. Furthermore, it addresses why the VSC-HVDC solution is forwarded as a
promising solution.
Chapter 3 presents the VSC-HVDC system in detail. The operating principle and
structure of VSC-HVDC, including its converters, harmonic filters, DC-capacitor,
phase reactors and transformers are described as well as the chosen modeling
2
Chapter 1
Introduction
approach. The design and selection of appropriate parameter values of VSC-HVDC
components is given in detail. The mathematical derivation and overall structure of
the VSC-HVDC control system is described in chapter 4. Chapter 5 discusses the
dynamic performance of the VSC-HVDC under idealized network conditions. In this
chapter, step changes in power and voltage, balanced and unbalanced faults are
simulated using a test network to evaluate the designed VSC-HVDC control systems.
In chapter 6, the VSC-HVDC model is used to couple the two 150 kV Areas on 50 kV
level. Various simulation scenarios are investigated to evaluate the performance of the
VSC-HVDC as a solution of the network constraints. The interaction of VSC-HVDC
with the rest of the grid under various disturbances is also investigated.
Finally, the conclusions of the work and some suggestions for future research to
deploy the VSC-HVDC in the grid are pointed out in Chapter 7.
3
Chapter 2
Network constraints and reinforcement
2.1 Introduction
This chapter presents the modelling of the studied grid and its constraints. Various
possible network reinforcement alternatives to alleviate the network constraints will
be described. Finally, it address why VSC-HVDC could be a potential network
reinforcement method.
2.2 Network model
The detail model of the studied network is shown in Fig. 2.1. Area 1 is a 23 kV
distribution network with many greenhouses and high penetration of DG.
Fig. 2.1 Detailed model of studied network
The distribution system is difficult to analyze due to the large number of active units
and the unavailability of some dynamic data. Therefore, a simplification was made by
representing the distribution network by representative aggregated load and
generation. Moreover, it is will be shown in section 6.5 that it is irrelevant to describe
the distribution system with a high level of detail for the main research question of
4
Chapter 2
Network constraints and reinforcements
this thesis. Thus aggregation is used to model the distribution grid; several generators
with the same or similar dynamic structures are represented by an equivalent
generator model [2, 3]. The equivalent inertia and apparent power are the sum of the
inertias and power ratings of all the individual generators. Thus, the distribution grid
is modeled as a single synchronous generator rated 50 MVA connected to a step-up
transformer of 10/23 kV as shown in Fig. 2.1. Moreover, the equivalent generator
model is operated to maintain the same steady state power flow conditions as the
detailed model; here it is operated in PQ mode with zero reactive power set point as
DG units are operated at unity power factor. The loads are modelled as equivalent
constant impedance load (static load) at each substation.
The correct modelling of synchronous generators is a very important issue in all kinds
of studies of electrical power systems. Here, it is taken the advantage of the highly
accurate models [1], which can be used for whole range of different analysis,
provided by DIgSILENT software. For both balanced and unbalanced RMS
simulations, all the generators G1, G2 and G3 will be of 5th order model, stator
transients are ignored and d-q currents remain DC during transients. For EMT
simulations, if any, Generator models of 7th order will be used.
Generators, G2 and G3 are identical synchronous machines of rated capacity of
625 MVA at 21 kV. Both are fitted with standard excitation system of type IEEET1,
governor system of type steam turbine governor (TGOV1) and power system
stabilizer (PSS) of type IEEEST which are available in DIgSILENT PowerFactory
software as standard library blocks. The block diagrams of these controllers and their
parameters are presented in Appendix B. Experimental simulation of a step change in
load and terminal voltage, and short-circuits have been made to test the performance
of the excitation and governor system by changing the gain parameters. The software
default parameters with a slight modification were found to show a reasonably
accurate performance.
Generator, G1 is fitted with excitation system of type IEEET1 and governor system of
type gas turbine generator (pcu_GAST). G1 is not fitted with power system stabilizer
(PSS) as actual distribution network active units, small CHP units, do not have a PSS.
The same experiment as for G2 and G3 has been made to see the performance of these
controllers, DIgSILENT default gain parameters are found to be reasonable.
2.3 Network constraints
As a result of the autonomous load growth, several assets in Area 2 will reach their
maximum loading capacity in the coming years. Besides, an outage in the 50 kV grid
network may cause constraints which lead to overloading of some of the remaining
circuits or transformers, and violation of voltage limits.
An extensive study of load flow calculation and N-1 contingency analysis taking into
account future load growths revealed that the components shown dotted in Fig. 2.2
will be under constraint [4]. Voltage levels within ± 5% are assumed to be acceptable.
5
Chapter 2
Network constraints and reinforcements
Fig. 2.2 Network constraints in Area 2, 50 kV grid
There are wind generators installed at bus 7 & bus 2 and also directly connected to
line 3 & line 2 in Fig. 2.2. These wind units reduce transport of power by supplying
loads locally. This reduces loading of transformers and lines, and will flatten the
voltage profile. However, the wind power production is of stochastic nature, thus one
cannot rely on them without appropriate remedy. In this thesis the worst case, where
the wind units are assumed to be unavailable, is considered. To study the sensitivity of
this assumption, the interaction of the largest wind unit connected to line 3/line 2,
with the grid will be studied in more detail in chapter 6.
2.4 Network Reinforcement
There can be many solutions to alleviate the constraints in the 50 kV grid; here three
possibilities have been investigated. The first solution can be reinforcing each of the
transformers and lines under constraint, and installing voltage support switched
capacitors at selective buses. This solution is obviously very expensive as there are
many components under constraint.
The second solution can be changing the direction of power flows during outage via
operator intervention, i.e. changing network topology during outage. This put stress
on operators besides compromising the reliability of the grid. Moreover, it does not
solve the voltage constraints.
The third solution is to reduce the power transport by the overloaded transformers and
lines rather than reinforcing them. This requires injecting power at lower 50 kV
substations in Area 2. This motivates using the available CHP units in Area 1.
Moreover, this method also reduces the fees charged by the Dutch national
transmission system operator (TenneT) for tie line costs at T 1 and T 2 by the same
amount of injected power or may be more if the new topology reduces overall system
losses. The question now is what is the best way to link the two areas.
To solve most of the constraints, the connection to Areas 2 shall be at bus 7, bus 6,
and bus 2 which are the lower buses in the topology. To solve the constraints on line 2
and line 3, Fig. 2.2, the connection shall be either at bus 2 or bus 7. Moreover to
obtain flat voltage profile, the connection shall be close to the middle bus in the
6
Chapter 2
Network constraints and reinforcements
topology. Thus bus 2 is chosen as a candidate. This has been verified via load flow
calculations and static voltage stability analysis in section 6.2.
An AC cable link between T 3, 50 kV winding and bus 2 shown in Fig. 2.2 was
studied. Load flow calculation showed that the link will result in large power flows
over the new link because of its smaller impedance path compared to Area 2, 50 kV
path. This result in overloading of the 40 MVA rated 50 kV winding of T 3. For the
grid operator it is impractical, due to space limitations, to reinforce T 3 or install
additional transformers. Moreover, the link results in high short-circuit currents
demanding change of the protection scheme and some of the circuit breakers of
Area 2, 50 kV grid because of the bi-directional power flow. This motivates using a
power flow controller which may also have voltage control capability as shown in
Fig. 2.3B.
There is no space available for power flow controller in Area 1, thus it has to be
installed in Area 2.
Fig. 2.3 Network reinforcement
The first obvious choice for grid operators will be reactor coils, capacitor banks and
phase shifting transformers (PST). Usually transformers are used to transport electric
power between different voltage levels. But they may also be used to control the
phase angle between the primary and secondary sides. Such special transformers are
called PST. Load flow calculations including optimal setting of tap-position analysis
showed that the PST alleviates the overloading of the lines and transformers [4]. But
the solution has the disadvantage of reactive power consumption which results in a
worse voltage profile; besides, short circuit current levels still remain to be addressed.
Moreover, it does not allow effective usage of T 3, 50 kV winding to transport only
active power because of the lagging power factor, i.e. independent control of active
and reactive power is not possible. Detailed mathematical analysis and applications of
phase-shifting transformers can be found in [5-7]. This motivated researchers to look
for electronic power flow controllers, which belong to the family of FACTS- devices.
Due to rapid development of the power electronics industry, an increasing number of
high power semi-conductor devices are available for power system applications.
These devices have made it possible to consider new technologies such as FACTS for
controlling power flow, securing line loading, and damping of power system
7
Chapter 2
Network constraints and reinforcements
oscillations [6]. FACTS devices are an attractive alternative for increasing the
transmission capacity of existing grids and enhancing operational flexibility.
HVDC as a power flow controller is seen as a potential solution to alleviate the
network constraints. HVDC in most cases is used to transport bulk power over long
distance by overhead transmission lines or cables. Here, it will be used in a more
innovative way to solve network constraints. Thus in the remaining part of this
chapter HVDC link as a power flow controller will be dealt with in more detail.
2.5 HVDC transmission
HVDC transmission uses power electronics technology with high power and voltage
ratings. It is an efficient and flexible method to transmit large amounts of electrical
power over long distances by overhead transmission lines or underground/submarine
cables. It is also used to interconnect two separate power systems or substations
within one interconnected system where traditional AC solutions cannot be used [8].
HVDC systems have a number of advantages over AC transmission, the important
ones being: converter electronics allows control over the power exchange between
two areas, allow asynchronous links between AC systems where AC ties are not
feasible, HVDC can carry more power per conductor, suitable for many solar and
wind generation solutions and flexibility of HVDC enables improvement of
performance of the overall AC/DC system & thus stabilize a predominantly AC grid.
2.5.1 Arrangement of HVDC systems
HVDC systems can be arranged in a number of configurations. The selection of the
configuration depends on the functions and location of the converter stations. Some of
the typical arrangements are shown in Fig. 2.4.
(a) Two Terminal HVDC system
(b) Back to Back HVDC system
(c) Hybrid AC and DC system
Fig. 2.4 Typical arrangements of AC-DC connections
8
Chapter 2
Network constraints and reinforcements
AC systems at the two sides can be either two asynchronous systems or two
substations within one interconnected system.
When it is economical to transport power through DC transmission from one
geographical location to another, a two terminal or point to point HVDC as shown in
Fig. 2.4(b) is used [9]. The hybrid configuration is mostly used to solve grid stability
problems.
In case of Back-to-Back scheme, the two converter stations are located at the same
site and the converter bridges are directly connected. Because of the unavailability of
installation space in Area 1 and the relatively low DC power level, the back-to-back
scheme will be the obvious choice in the studied case.
In general, the converters in HVDC system can be classified as line-commutated and
self-commutated (or: forced commutated), depending on the type of power electronic
switching elements applied. The line-commutated converters use switching devices
such as thyristor. HVDC systems based on thyristors are called traditional or Classical
HVDC.
Converters
Line commutated
converters
Self commutated
Converters
Voltage Source
Converters (VSC)
Current Source
Converters (CSC)
Fig. 2.5 Possible classification of converters
The self-commutated converters utilize fast switching devices such as IGBT and
GTO, that have controllable turn-off capability. They can be divided into two main
types based on the nature of the DC link: CSC and VSC as shown in Fig. 2.5. The
HVDC system based on VSC is commercially available as HVDC Light and HVDC
PLUS [9]. The classification as shown in Fig. 2.5 will be used to describe the
remaining part of this chapter.
2.5.2 Classical HVDC Systems
A typical classical HVDC system scheme is shown in Fig. 2.6. It consists of AC
filters, shunt capacitor banks or other reactive power-compensation equipments,
converter transformers, converter bridges, DC reactors, and DC lines or cables.
Classical HVDC converters are line commutated current source converters (CSC).
The CSCs perform the conversion from AC to DC (rectifier) at the sending end and
from DC to AC (inverter) at the receiving end. The direct current is kept constant and
magnitude and direction of power flow are controlled by changing the relative
magnitude and direction of direct voltage as can be seen from Eqn. (2.1).
9
Chapter 2
Network constraints and reinforcements
Pd =
U d 2 *(U d 1 − U d 2 )
Rdc
(2.1)
where Pd , U d 1 U d 2 Rdc are as shown in Fig. 2.6.
Terminal A
Pd
Rdc
U d1
Terminal B
U d2
Shunt capacitor
AC filter
AC filter
Shunt capacitor
Fig. 2.6 Basic configuration of classical HVDC system
The current harmonics generated by the converters are suppressed by AC filters. In
the conversion process the converter consumes reactive power which is compensated
in part by the filter banks and in part by capacitor banks. DC filters and smoothing
inductors reduce the ripple produced in DC transmission line current.
The power transmitted over the HVDC link is controlled by the control system where
one of the converters controls the direct voltage and the other converter controls the
current through the DC circuit. The control system acts by firing angle adjustments of
the valves, and tap changer adjustments on the converter transformers to obtain the
desired combination of voltage and current. Power reversal is obtained by reversing
polarity of direct voltages at both ends. The control systems of the two stations of a
bipolar HVDC system usually communicate with each other through a
telecommunication link. A more detailed review of classical HVDC can be found in
[8, 9]. Dynamic reactive power support is not possible in classical HVDC, and thus
the solution does not fully address the requirements described in section 2.3.
2.5.3 Voltage source converter (VSC) HVDC system
VSC-HVDC has been an area of growing interest since recently due to a number of
factors, like its modularity, the possibility to independently control active and reactive
powers, power reversal, good power quality and etc.
VSC can be considered as a controllable voltage source. This high controllability
allows for a wide range of applications. From a system point of view VSC-HVDC
acts as a synchronous machine without mass that can control active and reactive
power almost instantaneously. And as the generated output voltage can be virtually at
any angle and amplitude with respect to the bus voltage, it is possible to control the
active and reactive power flow independently.
10
Chapter 2
Network constraints and reinforcements
A typical back to back VSC-HVDC system, shown in Fig. 2.7, consists of AC filters,
transformers, converters, phase reactors, DC capacitors [8].
2Cdc
u dc
u
2Cdc
v
Fig. 2.7 Basic configuration of back to back VSC-HVDC system
The main operational difference between classical HVDC and VSC-HVDC is the
higher controllability of the latter. This leads to a number of potential advantages and
applications related to power flow flexibility and fast response to disturbances, where
extensive list of them can be found in [9, 10]. The main properties of classic HVDC
and VSC-HVDC are summarized in Table 2.1 below.
Table 2.1 Comparison of characterstics of classcial HVDC and VSC-HVDC
Classical HVDC
•
•
•
•
•
•
•
VSC-HVDC
Acts as a constant current source
on the DC side
The direct current is unidirectional
•
Polarity of direct voltage changes
with DC power flow
DC smoothing reactor maintains
constant DC
DC filter capacitance is used on
line-side of smoothing reactor
Line commutated or forced
commutated
Rarely PWM is applied
•
•
Acts as a constant voltage source
on the DC side
Polarity of direct voltage is
unidirectional
Direction of DC changes with
DC power flow
DC capacitance maintains direct
voltage constant
DC smoothing reactor is used on
line side of DC filter capacitance
Self commutated
•
Often PWM is applied
•
•
•
The high controllability of VSC-HVDC leads to independent control of active, and
reactive power and the possibility to control its short-circuit contributions makes it
technically a promising solution to solve the network constraints described in
section 2.3. However, the engineering and acquisition investment on VSC-HVDC
system is relatively high compared to conventional solutions.
A rough indication of the involved major investments could be: €300k/MVA rating of
back to back VSC-HVDC module and €20k/MW/year tie-line fees at 150 kV
transformers charged by TenneT. Excluding the engineering costs, and assuming
converter loss of 3 % and neglecting the detail of the change in system loss due to the
introduction of the VSC-HVDC; a return on investment of €768k/year can be
expected while the acquisition of the system costs €13.5m. The cost of the involved
cables will be the same as the conventional solutions.
11
Chapter 2
Network constraints and reinforcements
For the VSC-HVDC, it is not yet possible to define the reliability figures based on
operating experience because no recorded figures are publicly available. Therefore, in
the above calculations, availability of 99 % for each conversion station is assumed
(including power electronic converters, transformers, reactors, filters, controls, and
the auxiliaries).
The reinforcement methods discussed in section are summarized in Fig. 2.8 below. In
the subsequent two chapters the design of VSC-HVDC will be addressed in full detail
Network Reinforcement
Component wise
reforcement
Changing grid
topology
Injection of power
at lower bus bars
DG units in Area 1
Power flow controller
Direct Ac- link
HVDC
Phase shifting transformer
Back to Back
Line commutated converters
Self commutated converters
CSC
VSC-HVDC
Forwarded solution
Fig. 2.8 Overview of network reinforcement methods
12
Chapter 3
Design and operating principle of VSC-HVDC
In this chapter the design of VSC-HVDC station components, shown in Fig. 2.7 and
its operating principle are dealt with in full detail.
3.1 Operating principle of VSC-HVDC
The fundamental operation of VSC-HVDC can be explained by considering a voltage
source converter connected to an AC network as shown in Fig. 3.1.
2Cdc
Reactor , X
Δv
2Cdc
v
u
Fig. 3.1 Simplified representation of VSC connected to AC grid
The converter can be thought of as an equivalent AC voltage source where the
amplitude, phase and frequency can be controlled independently, see section 3.3.
Thus the VSC bridge can be seen as a very fast controllable synchronous machine
∧
whose instantaneous phase voltage u , described by
∧
1
u = udc M sin(ωet + δ ) + harmonics
2
(3.1)
where M is the modulation index which is defined as the ratio of the peak value of
the modulating wave and the peak value of the carrier wave, ωe is the fundamental
frequency, and δ is the phase shift of the output voltage.
Variables M and δ can be adjusted independently to obtain any combination of
voltage amplitude and phase shift in relation to the fundamental frequency voltage of
the AC system. Thus the voltage drop Δv across the reactor X can be varied to
control the active and reactive power flows.
The active power flow between the converter and the AC system can be controlled by
controlling the phase angle between the fundamental frequency voltage generated by
the converter and the voltage across the AC-filter. Taking the voltage at the filter bus
as a reference and assuming lossless reactor, the power transfer from the converter to
the AC system will be:
P=
| v || u | sin δ
X
(3.2)
13
Chapter 3
Design and operating principle of VSC-HVDC
The reactive power flow is determined by the relative difference in magnitude
between the converter and filter voltages. The reactive power flow is calculated as:
Q=
| v | (| v | − | u | cos δ )
X
(3.3)
The active power flow on the AC side is equal to the active power transmitted from
the DC side in steady state, disregarding the losses. This can be fulfilled if one of the
two converters controls the active power transmitted and the other controls the direct
voltage. The reactive power generated/consumed by the converter is adjusted to
control AC network voltage or/and reactive power injections.
3.2 Capability chart of VSC -HVDC
It is common to describe the capability of a power apparatus in a number of different
ways, i.e. showing under what conditions it can operate. Active power and reactive
power capability is usually illustrated in the P–Q plane. There are mainly three factors
that limit the active and reactive power output of VSC-HVDC as shown in Fig. 3.2.
The first limiting factor is the maximum current through the IGBT valves. This leads
to the maximum MVA circle in the MVA plane where the maximum current and the
actual AC voltage are multiplied. If the AC voltage decreases, so will the MVA
capability be reduced proportionally to the voltage drop.
| u |= 0.9
| u |= 1
| u |= 1.1
Fig. 3.2 Capability curve of VSC-HVDC
The second limiting factor is the maximum direct voltage. The AC-voltage generated
by the converter is limited by the allowable maximum direct voltage. The reactive
power is mainly dependent on the voltage difference between the AC voltage the VSC
can generate from the direct voltage and the AC grid voltage. If the grid AC voltage is
high the difference between the AC voltage generated by the converter and the grid
AC voltage will be low. The reactive power capability is then moderate but increases
with decreasing AC voltage. This makes sense from a stability point of view.
14
Chapter 3
Design and operating principle of VSC-HVDC
The third limit is the maximum current through the cable if the connection is not back
to back.
For low AC voltages the MVA limit is dominating while for high AC voltages the
DC-limit is quite restrictive but it is not likely that we in that case would like to inject
reactive power when AC voltage is already high. The absorbing reactive capacity
given by MVA circle is hence much more important for high AC voltages. In brief,
MVA capacity limit is in most situations the most restricting one.
3.3 PWM
VSC-HVDC is based on VSC, where the valves are built of IGBTs or GTOs and
PWM is used to create the desired voltage waveform. There are various schemes to
pulse width modulate converter switches in order to shape the output AC voltages to
be as close to a sine wave as possible. Out of the various schemes, sinusoidal pulse
width modulation (SPWM) is discussed here.
In SPWM, to obtain balanced three-phase outputs, a triangular wave form is
compared with three sinusoidal control voltages that are 120º out of phase. In the
linear region of modulation (amplitude modulation index, M ≤ 1 ), the fundamental
frequency component of the output voltage of the converter varies linearly with M .
The line to line voltage at the fundamental frequency can be written as
vLL =
3 M * udc
≈ 0.612M * udc
2 2
(3.4)
By using PWM with high switching frequency, the wave shape of the converter AC
voltage output can be controlled to be almost sinusoidal with the aid of phase reactors
and tuned filters. Changes in waveform, phase angle and magnitude can be made by
changing the PWM pattern, which can be done almost instantaneously. Moreover,
being the focus of this thesis on the RMS dynamics of the studied grid rather than the
switching behaviour of the converters allows modelling the PWM controlled bridge
by an equivalent voltage sources at fundamental frequency (the averaged model), i.e.
controllable voltage source satisfying Eqn. (3.1 & 3.4) [11].
The choice of the modulation index is a trade off between output power and dynamic
response. The higher the modulation index, the higher the output power rating, i.e.
S ∝ M * udc * I rms . Higher M is also preferred from the stand point of harmonics, i.e.
high M for M <1 results in low total harmonic distortion (THD) [11]. On the other
hand, higher M will leave a smaller modulation index margin for dynamic response
[12].
3.4 VSC-HVDC station components
3.4.1 Converters
The converters so far employed in actual transmission applications are composed of a
number of elementary converters such two-level, six-pulse bridges, shown in Fig. 3.3
15
Chapter 3
Design and operating principle of VSC-HVDC
and multi-level topologies such as three-phase, three-level, twelve-pulse bridges [9].
The advantages of multilevel topologies over the two-level converters are improved
voltage waveform quality on the AC side, smaller filter size, and lower switching loss
at the same or better harmonic content.
The two-level bridge is the simplest circuit that can be used for building up a three
phase forced commutated VSC bridges. It consists of six valves and each valve
consists of an IGBT and an anti-parallel diode. IGBTs of nominal current 5001500 A, rated voltage of 2.5 kV and switch frequency of 1-2 kHz are available on
market [13]. In order to use the two or three-level bridge in high power applications
series connection of devices may be necessary and then each valve will be built up of
a number of series connected turn-off devices and anti-parallel diodes. The number of
devices required is determined by the rated power of the bridge and the power
handling capability of the switching devices.
Fig. 3.3 Two-level VSC converter
3.4.2 Converter size
The design of the size of converters in HVDC systems depends basically on the
steady performances requirements, i.e. on scheduled active power transport and
voltage support requirement. During steady state operation the voltages at the
equipment terminals, e.g. converters, shall be within the pre-defined limits. Typical
limits are 95 % to 105 % and 90 % to 110 %. Strictly speaking, the limits are only
applicable to the equipment terminals. However the way power systems are currently
designed and operated requires voltage to be kept within limits in the whole power
system.
Driven by the cost advantage, reduced fees from TenneT due to reduced overall
imported power, the company wants to draw the available power from Area 1.
However the maximum is limited to 40 MW due to the constraint on T3 50 kV
winding, see Fig. 2.1. Moreover, large power transfers may be accompanied by other
network constraints. The reactive power injection at bus 2 required to get acceptable
flat voltage profiles under worst case N-1 outage leads to converter rating of
45 MVA, based on load flow calculations.
3.4.3 Converter transformer
The transformers connect the AC network to the valve bridges and adjust the AC
voltage level to a suitable level for the converters. The transformers can be of
different design depending on the power to be transmitted and possible transport
requirements.
16
Chapter 3
Design and operating principle of VSC-HVDC
Using IGBT valves of nominal current of 500 A, transmitted DC power of 40 MW,
AC-side grid voltage of 50 kV, and steady state modulation index of M = 0.85 for
reasons described in section 3.3, and Eqn. (3.4) and Pdc = I dcudc , the turn ratio of the
transformer will be 1.12. Thus it is decided not to include the transformer in the VSCHVDC model.
In fact converter transformer has other advantage besides just merely transforming the
voltage levels. It has tap changers which can help in regulating the voltage. But within
the time frame of interest of this thesis, the tap-changers are not expected to react.
Thus the absence of the converter transformers does not deteriorate the accuracy of
our VSC- HVDC model, though the ability of managing zero-sequence currents will
be lost. This will be addressed in detail in chapter 5. Moreover, the transfer of active
power is still possible due to the presence of phase reactors, section 3.4.6.
3.4.4 Direct voltage
The large proportion of the cost of HVDC links is the cost of the converter bridges.
Thus the choice of the voltage levels mainly depends on economical issues. Higher
voltage levels require many valves to be put in series, thus higher costs.
Technically, the minimum direct-voltage level required to avoid converter saturation
while using sinusoidal Pulse width modulation can be calculated from Eqn. (3.4) with
M = 1 , and is given by Eqn. (3.5).
udc min =
vLL min
0.612
(3.5)
The maximum direct voltage depends on the design steady state modulation index. In
most commercial applications M < 0.9 [13] is taken as a design parameter. The
maximum direct voltage level is given by Eqn. (3.6).
udc max =
vLL max
0.612* M
(3.6)
where vLL max and vLL min are the maximum and minimum steady state acceptable ACvoltage level, 105 % and 95 %, respectively.
With a steady state modulation index of M = 0.85 , a direct voltage of 100 kV which
is within the above two limits is chosen as a design value.
3.4.5 DC capacitor
In steady state assuming no losses in the DC link, the instantaneous power on the DC
side must equal the instantaneous power on the AC side. In the moment the power
balance is broken, the instantaneous difference in power is stored in the DC link
capacitor and this leads to fluctuations in the direct voltage. Thus instantaneous
current flows in the DC link given by
17
Chapter 3
Design and operating principle of VSC-HVDC
ic = Cdc
dudc
dt
(3.7)
Due to PWM switching action in VSC-HVDC, the DC link capacitor current contains
harmonics, which will result in a ripple on the DC side voltage. This ripple must be
small enough for the voltage to be virtually constant during switching period. This
sets a lower limit on the capacitor size.
Small voltage ripples require large capacitor, which on the one side has a slow
response to voltage changes, but on the other side has a smaller current and thus
increased lifetime. On the other hand, a small capacitor makes fast changes in the
direct voltage possible allowing fast control of active power at the expense of higher
voltage ripples and reduced lifetime. Selecting the size of the DC capacitor has thus to
be a trade-off between voltage ripples, lifetime and the fast control of the DC link.
The trade off relation for the design of the DC-link capacitor in the back-back
converter is described in [9, 14] as follows
Cdc =
τ=
SN
udcN * Δudc * 2* ωe
(3.8)
0.5Cdc u 2 dcN
SN
(3.9)
where udcN denotes the nominal direct voltage and S N stands for the nominal
apparent power of the converter. The time constant τ is equal to the time needed to
charge the capacitor from zero to rated voltage udcN when the converter is supplied
with a constant active power equal to S N . Δudc denotes the allowed ripple (peak to
peak), and ωe the electrical frequency.
Eqn. (3.8 & 3.9) sets the lower and upper limits of the size of DC-capacitor
respectively. The time constant τ is selected to be less than 10 ms to satisfy small
ripple and transient overvoltage on the DC-link. A capacitance of 37.5 µF which
result in a peak to peak percentage ripple of 18 % is used.
3.4.6 Phase reactor
The phase reactors, as shown in Fig. 2.7, are used for controlling both the active and
the reactive power flow by regulating currents through them. The reactors also
functions as AC filters to reduce the high frequency harmonic content of the AC
currents which are caused by the switching operation of the VSCs. The reactors are
essential for both the active and reactive power flow, since these properties are
determined by the power frequency voltage across the reactors.
The choice of the size of the phase reactor depends on the switching frequency,
converter saturation and control algorithm, converter saturation being the dominant
determinant factor. In vector controlled-VSC-HVDC, section 4.3, the phase reactor L
is chosen such that the minimum reference current tracking time Δt is less than the
18
Chapter 3
Design and operating principle of VSC-HVDC
time constant of the converter current controller [15] , governed by Eqn. (3.10). Thus
converter saturation and control put the upper threshold while current smoothing,
active power and reactive power controls may put lower threshold for the phase
reactor. The phase reactors are usually in the range of 0.1 pu to 0.2 pu [9, 13, 14].
Δt =
0.9 L
0.612 udc
ωe *(vLL − vLL max )
(3.10)
where vLL (pu) denotes the AC- side voltage, L (pu)-smoothing reactor inductance and
vLL max (pu) stands for the theoretical maximum amplitude of the converter
fundamental phase voltage, i.e. 0.612udc for sinusoidal PWM, see Eqn. (3.4). It should
be noted that this value changes with the specific type of PWM used.
3.4.7 AC filters
The AC voltage output contains the fundamental AC component plus higher-order
harmonic, derived from the switching of the IGBT’s. These harmonics have to be
taken care of preventing them from being emitted into the AC system so that
sinusoidal line currents and voltages can be obtained at the point of common coupling
(PCC). High pass filters are installed to take care of these high order harmonics.
Moreover they serve as source of reactive power.
As stated in [11], a PWM output waveform contains harmonics Kf c ± Nf1 where f c is
the carrier frequency, f1 is the fundamental grid frequency. K and N are integers and
their sum is an odd integer. Next to the fundamental frequency component, the
spectrum of the output voltages contains components around the carrier frequency of
the PWM and multiples of the carrier frequency.
With the use of PWM, passive high-pass damped filters are selected to filter the high
order harmonics. Normally, a second order high-pass filter (see Fig. 3.4), the
characteristic frequency of which is selected based on the switching [9], is used in
VSC-HVDC systems. In RMS type simulations the filter solely injects reactive power
at fundamental frequency and does not need to be represented in full detail.
Cfilter
Lfilter
R filter
Fig. 3.4 passive second order high pass filter
Quality factor Q f of typical values between 0.5 % and 5 % [9], AC-filter rating,
Q filter and harmonic order h are used as a design parameters. The resistance R filter ,
19
Chapter 3
Design and operating principle of VSC-HVDC
capacitance C filter and inductances L filter are calculated based on the following
equations
C filter =
L filter
(h 2 − 1)Q filter
(3.11)
h 2ωe vLL 2
1
=
C filter h 2ωe 2
R filter = Q f
(3.12)
L filter
(3.13)
C filter
In this thesis typical values of Q filter =15 % of converter rating [13], Q f =3 % [9, 13],
and h=35 are used as design inputs. Remark should be taken that these values depends
on harmonic requirements of the specific system under study.
20
Chapter 4
Control system of VSC-HVDC
4.1 Introduction
The control of a VSC-HVDC system is basically the control of the transfer of energy.
The aim of the control in VSC based HVDC transmission is thus the accurate control
of transmitted active and reactive power. Moreover, the VSC controls are often used
to provide ancillary services, such as improve the dynamics of AC grids.
Different control strategies are found in literature for the control of VSC-HVDC.
Direct control and vector control methods which are based on voltage controlled VSC
and current controlled VSC schemes respectively are the most widely used methods.
In voltage-controlled schemes, the active and reactive power is controlled directly by
controlling the phase angle and amplitude of the converter output voltage. On the
other hand, the current controlled scheme utilizes the converters as a controllable
current source, where the injected current vector follows a reference current vector.
The current-controlled VSC offers potential advantages over the voltage-controlled
VSC. The mains advantages being: 1) better power quality as the current-controller
converter is less affected by grid harmonics and disturbances, 2) decoupled control of
active and reactive power, 3) inherent protection against over currents, and 4) the
control mode can be easily extended to compensate for line harmonics and other
power quality issues [16]. The vector control method is widely used in VSC-HVDC
and will also be adopted in this thesis.
4.2 Direct control
The direct control method uses voltage control of the VSC. The active and reactive
power flows are controlled by directly altering the phase shift δ , and the modulation
index M thus the magnitude of the converter voltage, see Eqns. (3.2 & 3.3). The
actual power angle is calculated from the terminal quantities and compared to the
desired power angle, which is calculated from the active power order. The error in the
power angle is processed by a power angle controller to generate the reference phase
angle of the modulating signal. In a similar manner, the error between the actual and
desired reactive power is processed by a reactive power controller to generate the
magnitude reference of the modulating signal. A phase-locked loop (PLL) circuit is
responsible for synchronizing the converter output voltage with the AC grid. The
control scheme is shown in Fig. 4.1.
21
Chapter 4
Control system of VSC-HVDC
Reactor , X
Δv
u
v
u*ref ∠δ *ref
Q*
+
Q
u*ref
−
ωe
V
P*
v
u
δ*
+
δ *ref
−
δ
u
v
Q
Fig. 4.1 Direct control principle of VSC-HVDC
A change in the converter voltage angle does influence both P and Q so does a
change in magnitude of the converter voltage, u ; see Eqn. (3.2 & 3.3). Thus
independent control of active and reactive power is not possible.
4.3 Vector control
The most widely used control scheme for VSC-HVDC is vector control. This method
controls the converter voltage to track a current order injected into the AC network.
The vector control scheme involves representation of three-phase quantities in the dq
synchronous reference frame. The transformation of phase quantities to dq coordinates involves two steps: a transformation from the three-phase stationary
coordinate system to the two-phase αβ stationary coordinate system and a
transformation from the αβ stationary coordinate to the dq rotating coordinate
system. Power invariant Clark and Park transformation are used to convert between
the reference frames. The zero-sequence components will not be considered in the
coordinate transformation as balanced three-phase modeling is adopted. Details can
be found in Appendix A.
One of the most advantageous characteristics of vector control is that vectors of AC
currents and voltages occur as constant vectors in the steady state, and hence static
errors in the control system can be successfully removed by applying PI controllers.
The dynamic model of a three phase grid interfaced VSC, as shown in Fig. 4.2,
consists of models of the AC and DC sides, and equations to link them.
22
Chapter 4
Control system of VSC-HVDC
iL
idc
+udc
icap
2Cdc
u dc
i
L
R
v
u
2Cdc
−udc
Fig. 4.2 Single line diagram representation of VSC
In stationary coordinates, seen from the filter bus voltage towards the converter, the
AC-dynamics are given by the dynamics of the phase reactors.
L
di αβ
dt
= v αβ − uαβ − Ri αβ
(4.1)
Transforming Eqn. (4.1) to synchronous coordinates,
L
di dq
dt
= v dq - udq − ( R + jωe L)i dq
(4.2)
The term jωe Lidq in Eqn. (4.2) represents the time derivative of the synchronous
rotation of the dq reference frame. Eqn. (4.2) can be rewritten component wise as
did
= − Rid + ωe Liq − ud + vd
dt
di
L q = − Riq − ωe Lid − uq + vq
dt
L
(4.3)
From Eqn. (4.3), the equivalent circuit of the VSC in the synchronized dq -reference
frame will be as shown in Fig. 4.3.
id
vd
ωe Liq
iq
ωe Lid
vq
ud
uq
Fig. 4.3 Equivalent circuit of VSC in synchronous dq reference frame
The dq -reference frame is chosen such that its d -axis is defined to be along AC filter
voltage. With this alignment,
23
Chapter 4
Control system of VSC-HVDC
vq = 0 and vd = v
(4.4)
Using Eqn. (4.4) and Appendix A, the instantaneous real and reactive power absorbed
from the AC system will be
p = vd id
(4.5)
q = −vd iq
To complete the dynamic model of the VSC, the dynamics of the DC link are given
by
Cdc
dudc
= idc − iL
dt
(4.6)
and
pdc = udc idc
(4.7)
Eqns. (4.3, 4.5, 4.6, & 4.7) fully describe the VSC in Fig. 4.2.
As can be seen, the transformation into rotating dq coordinate system leads to the
possibility to control the two current components, id and iq independently. Thus
independent control of active and reactive power is possible; see Eqn. (4.5), assuming
the PLL is performing well.
As the vector control technique offers decoupled control of active and reactive power
and fast dynamics, it makes the realization of system control in the form of a cascade
structure possible, with two control loops in cascade: an outer control loop that
provides the current set points and the inner current control loop described above. The
outer controllers include the direct voltage controller, the active power controller, the
reactive power controller, and the AC voltage controller, depending on the
application. The reference value for active current can be provided by the direct
voltage controller or the active power controller, while the reference value for reactive
current is provided by AC voltage controller or reactive power controller.
In all possible combinations of outer controllers, the direct voltage controller is
always necessary to ensure an active power balance in the system. Active power taken
out of the network must equal the active power fed into the network minus the losses
in the DC system; any difference would mean that the direct voltage in the system
will rapidly change to intolerable levels.
Fig. 4.4 shows the various controllers of VSC-HVDC. The control system of VSC-2
is not shown explicitly, but is similar to that of VSC-1.
24
Chapter 4
Control system of VSC-HVDC
id c
VSC-1
Sending bus
L
p
v
id ref
Dc voltage
controller
q
Reactive power
controller
q ref
v
PLL
θ1
udcref
Inner current
controller
Ac voltage
controller
v ref
Receiving
bus
2Cdc
PWM
Active power
controller
u dc
R
v
p
VSC-2
2Cdc
u
ref
iL
idq vdq
iq ref
Outer
controller
αβ → dq
iαβ vαβ
Fig. 4.4 Vector control scheme of VSC-HVDC
The inner current controller and the various outer controllers will be described in
detail in the remaining part of this chapter. Simulation results illustrating the
performance of the current control scheme will be given in chapter 5.
4.3.1 Inner Current controller
The inner current control loop can be implemented in the dq -frame, based on the
basic relationship of the system model. The control loop consists of controllers,
decoupling factors and feed-forward terms as will be described further. The current
control block is represented by the following general block diagram.
i dqref
udq ref
+
udq
i dq
−
Fig. 4.5 General block diagram of inner current controller
Inside the controller block, there are two regulators, respectively for d and q axis
current control. In order to have a detailed overview of the control system, each block
of the control system is discussed below.
4.3.1.1 PWM converter
The inner current control loop generates a voltage set point udq ref , transforming to
αβ reference frame,
uαβ ref = e jθ1 udq ref
(4.8)
25
Chapter 4
Control system of VSC-HVDC
where θ1 is the angle of the dq frame used by the control system, which is obtained
from the PLL. The reference vector uαβ ref serves as input signal to the PWM of the
VSC. The PWM can be regarded as fast and accurate; as long as the reference vector
does not exceed the maximum modulus or PWM is in its linear range, i.e.
| udq ref | = | u αβ ref | ≤ umax
(4.9)
where umax is proportional to the direct voltage, it can be assumed that the actual
converter voltage follows the reference without time delay. The PWM only adds
switching harmonics, i.e.
u dq = u dq ref + harmonics
(4.10)
However, the phase reactors and tuned filters remove virtually all switching
harmonics as seen from the grid interface. If the PWM is made to remain in its linear
range, this assumption leads to
u dq = u dq ref
(4.11)
4.3.1.2 System transfer function
The system behaviour is governed by Eqn. (4.3) which is rewritten as:
did
+ Rid − ωe Liq
dt
di
vq − uq = L q + Riq + ωe Lid
dt
vd − ud = L
(4.12)
Eqn. (4.12) shows that the model of the VSC in the synchronous reference frame is a
multiple-input multiple output, strongly coupled nonlinear system. Thus it will be
difficult to realize the exact decoupled control system with general linear control
strategies. The transformed voltage equations of each axis have speed/frequency
induced term ( ωe Lid and ωe Liq ) that gives a cross-coupling between the two axes. For
each axis, the cross-coupling term can be considered as disturbance from a control
point of view. Thus, a close-loop direct current controller with decoupled current
compensation and voltage feed-forward compensation is required to obtain a good
control performance.
Using Eqn. (4.11), Fig. 4.5 and Laplace transformation leads to
u dq ( s ) = (i dq ref ( s ) − i dq ( s ) ) H ( s )
(4.13)
The inputs to the system is modified to include a component obtained from the
converter and feed-forward terms to eliminate the cross-coupling as shown below
26
Chapter 4
Control system of VSC-HVDC
ud ref (s) = −(id ref (s) − id (s))H(s) + ωe Liq (s) + vd (s)
uqref (s) = −(iqref (s) − iq (s) )H(s) − ωe Lid (s) + vq (s)
Manipulating Eqn. (4.11), Eqn. (4.12) and Eqn. (4.13), gives
did
+ Rid = ud
dt
di
L q + Riq = uq
dt
L
(4.15)
Eqn. (4.15) shows that the cross coupling terms are cancelled out and independent
control in d and q axis is achieved. Moreover, the equations in d and q axis show
the same form. Thus analysis of the regulator in the d -axis is enough.
Laplace transforming Eqn. (4.15),
i dq (s) =
1
udq (s)
sL + R
(4.16)
Hence the system transfer function becomes:
G ( s) =
1 1
R 1 + sτ
(4.17)
where the time constant, τ is defined as τ = L
R
4.3.1.3 Regulator
Eqn. (4.15) shows that the resulting system is composed of two independent first
order systems, thus it is sufficient to use a PI controller as a regulator. Hence H(s) will
be
H (s) = K p +
⎛ 1 + Ti s ⎞
Ki
= Kp ⎜
⎟
s
⎝ Ti s ⎠
(4.18)
where K p is the proportional gain and Ti is the integral time constant.
4.3.1.4 Control block diagram
The detailed block diagram of the complete system is developed based on Eqns. (4.11,
4.14, 4.17 & 4.18) and is shown in Fig. 4.6.
27
Chapter 4
Control system of VSC-HVDC
vd
id
+
ref
+
ud
−
PI
−
ref
+
PWM
converter
ud = ud ref
vd
+
ud −
+
1 1
R 1 + sτ
id
ωe L
ωe L
iq ref
−
+
−
PI
−
+
u q ref
vq
PWM
converter
ref
q
q
u q−
u =u
−
+
iq
1 1
R 1 + sτ
vq
Fig. 4.6 Detailed block diagram of complete system
4.3.2 Outer controllers
As previously mentioned, the outer controllers consist of direct voltage controller, AC
voltage controller, active power controller and reactive power controller. The
simplified diagram of the cascaded control system is shown in Fig. 4.7.
X
ref
+
−
Outer
controller
i dq ref
Inner
controller
udq
System transfer
function
X
Fig. 4.7 Outer controller scheme
where X ref denotes the desired set point of the outer controllers, and X is the actual
value of the controlled variable.
In the cascaded control system, the outer controllers must be much slower than the
inner current controller in order to insure stability [9]. Thus, for the design of the
outer controllers, the response of the current control loops may be assumed
instantaneous, i.e. (Fig. 4.7)
i dq = i dq ref
(4.19)
4.3.2.1 Direct voltage control
A direct voltage controller is needed to maintain the active power exchange between
the converters. Using Eqns. (4.5 & 4.19), the expression for the active and reactive
powers becomes
p = vd id ref
(4.20)
q = −vd iq ref
28
Chapter 4
Control system of VSC-HVDC
Neglecting losses in the converter and phase reactor, equating the power on the AC
and DC sides of the converter using Eqns. (4.7 & 4.20):
idc =
vd ref
id
udc
(3.21)
Any unbalance between AC and DC powers leads to a change in voltage over the DC
link capacitor, equating Eqn. (4.6) and Eqn. (4.21)
Cdc
dudc vd
=
id − iL
dt
udc
(4.22)
We see that Eqn. (4.22) is non-linear with respect to udc . Linearizing Eqn. (4.22)
around steady state operating point as outlined in [8] and considering i L as a
disturbance signal, the transfer function from udc to i d becomes
G ( s) =
vd 1
udc 0 s Cdc
(4.23)
where udc 0 is the steady state DC-link capacitor voltage.
Transfer function, G ( s ) , has a pole at the origin, thus it will be difficult to control it.
Introducing an inner feedback loop for active power damping as outlined in [17],
id ref = id ' − Ga udc
(4.24)
Substituting Eqn. (4.24) into Eqn. (4.22) gives
Cdc
dudc
v
= d (id ' − Ga udc )
dt
udc 0
(4.25)
The Laplace transformation from id ' to udc becomes
G '( s ) =
vd udc 0
sCdc + Ga (vd udc 0 )
(4.26)
Since Eqn. (4.26) is a first order system, a PI controller can be used to control the
direct voltage. Under balanced operating conditions idc = iL , thus the reference value
u
of the d -axis current i d ref shall be dc iL which the feed-forward term, ensuring
vd
exact compensation for load variation, in Fig. 4.8.
29
Chapter 4
Control system of VSC-HVDC
u dc
iL
vd
udc
ref
+
id '
+
id
−
−
ref
imax
id lim
−imax
Ga
u dc
Fig. 4.8 Direct voltage controller structure
4.3.2.2 Active and reactive power control
If vd in Eqn. (4.20) is assumed to be constant, then the active and reactive powers will
be correlated with the active and reactive current references respectively. The simplest
method to control the active and reactive powers will therefore be an open-loop
controller, Eqn. (4.27).
id ref =
iq
ref
P ref
vd
(4.27)
q ref
=−
vd
More accurate control can be achieved if a feedback loop is employed, using PI
control, in combination with the open loop [8, 9] as shown in Fig. 4.9.
p
ref
p ref
vd
+
+
+
id
−
q
id lim
− i m ax
p
q ref
imax
ref
−
iq
+
+
−
q ref
vd
ref
imax
iq lim
− i m ax
Fig. 4.9 (a) Active power controller (b) Reactive power controller
In this thesis, the above topology without the open loop will be adopted. This makes
relatively slow active and reactive power controller possible without using large
integral gains.
30
Chapter 4
Control system of VSC-HVDC
4.3.2.3 Ac voltage control
The voltage drop Δv over the phase reactor in Fig. 4.4 can be approximated as [9]:
Δv = v − u ≈
R * p + ωe L * q
v
(4.28)
Assuming ωe L >> R for the phase reactor, the voltage drop over the reactor depends
only on the reactive power flow. With this assumption, the variation of the AC
voltage depends only on the reactive power flow, meaning the voltage can be
regulated by controlling the q -component of the current. From Eqns. (4.27 & 4.28),
the block diagram of the AC voltage controller can be obtained as shown in Fig. 4.10.
imax
v ref +
v
iq
−
ref
iq lim
−imax
Fig. 4.10 Ac voltage controller
4.3.3 Limiting strategies
Since the VSC-HVDC does not have any overload capability as synchronous
generators have, over-currents due to disturbances will lead to thermal degradation of
the valves or instantly to permanent damage. The VSC-HVDC shall be operated
within its capability limits, see section 3.2. Therefore, a current limiter must be
implemented in the control system. Moreover, in order to maintain a proper control
and reduce the lower order frequency harmonics due to over-modulation, the
maximum reference voltage generated by the inner current loop shall be limited. The
maximum value is such that modulation index is less or equal to one for SPWM [11].
The current limit imax is compared with the current magnitude computed from the
active and reactive reference currents. When the current limit is exceeded, both the
active and reactive reference currents will be limited to
id lim and iq lim . The choice of
the limiting strategy depends on the application.
The first strategy could be to give the active reference current high priority when the
current limit is exceeded as shown in Fig. 4.11 [15]. This strategy is used for instance
when the converter is connected to a strong grid to produce more power [9, 18].
The second strategy could be, for instance when the converter is connected to a weak
grid or used to supply an industrial plant, to give high priority to the reactive reference
current when the current limit is exceeded as shown in Fig. 4.11 (b) [9]. This strategy
helps to support the AC-side voltage by allowing the converter to increase it reactive
power support equal to its rating during voltage dips.
31
Chapter 4
Control system of VSC-HVDC
The last strategy could be to give equal scaling to the active and reactive current
references when the current limit is exceeded as shown in Fig. 4.11 (c) [18].
In this thesis current limiting strategy as in Fig. 4.11(a) will be adopted, giving
precedence to the active current component.
idref
id lim
i max
( id
ref
) + ( iq
2
ref
)
idref
2
( id
id lim
iq ref
iq lim
ref
) + ( iq
2
iq ref
iq lim
i max
idref
ref
)
2
id lim
i max
( id ref ) 2 + ( iq ref ) 2
iq ref
iq lim
Fig. 4.11 Current limiting strategies
4.3.4 Controller integral windup
When designing the control laws, the control signal cannot be arbitrary large due to
design limitation of the converters. Therefore the control signal must be limited
(saturated) as discussed above. This causes the integral part of the PI controller to
accumulate the control error during limiting of the output signal, so called integral
windup. This might cause overshoots in the controlled variable [19].
Integral windup can be avoided by making sure that the integral is kept to a proper
value or disconnected during saturation, so that the controller is ready to resume
action as soon as the control error changes.
4.3.5 Tuning of PI controllers
In tuning of PI controllers for VSC-HVDC, the tuning is done following the criteria
adopted for electrical drives. Cascaded control requires the speed of response to
increase towards the inner loop.
The well known tuning rule, internal model control (IMC) is adopted for tuning the
inner current controller. With the analysis outlined in [8, 9, 15, 20], the gain
parameters of the inner current loop are calculated using Eqn. (4.29). Moreover, care
must be taken in the proportional constants in RMS calculations for iteration
convergence.
Kp =αL
Ti = L
KI = α R
tr =
R
ln 9
(4.29)
α
where α is a design parameter that is equal to the closed loop bandwidth and tr is the
rise time, the time needed to take a step response from 10 % to 90 %. A rise time of
32
Chapter 4
Control system of VSC-HVDC
tr =2 ms, equivalently α =1098 rad/s will be used for the inner current loop. This is in
agreement with the phase reactor inductance value requirement of Section 3.6,
Eqn. (3.10). However, the performance of the control system will get better if the
inner loop is made to respond instantaneous. But in reality instantaneous response
means usually infinite amplification of noise.
IMC is also used to tune the direct voltage controller. The gain parameters are
calculated based on analysis outlined in [9, 15, 17] where the equations are
reproduced here as Eqn. (4.30).
K p = α d Cdc
Ga = α d Cdc
K I = α d 2Cdc
(4.31)
where α d =220 rad/ s is used as an initial value. The actual gains used are a bit
modified from the ones calculated based on Eqn. (4.30) to get the desired response
time, slower than the inner current controller.
For the AC voltage and Reactive power controllers, there are no general tuning rules
as the controller gain depends on the network impedance [20]. Therefore it is made
via trial and error to get a reasonable speed of response, slower than the inner current
and direct voltage controllers.
33
Chapter 5
Testing of VSC-HVDC model
5.1 Introductions
For the purpose of analyzing the performance of the designed controllers of the VSCHVDC system, it is not necessary to represent the studied grid in full detail.
Therefore, the connection to PCC at bus-1 and bus-2 in Fig. 2.1 are replaced by slack
buses with the same short-circuit characteristics as the full detailed grid, see Fig. 5.1.
However, this leads to optimistic results as the inherent coupling of both PCCs, due to
meshed grid topology, is lost. Moreover, the dynamics of the loads and generators is
lost in the reduced test model which helps to see the dynamics of the VSC-HVDC
clearly. The applied settings of the slack buses and the system parameters are
provided in Appendix C. The voltage at the terminals of both network equivalents is
set 1 pu. Because of the DC-link, angle, power exchange between the slack nodes is
not dependent on the relative angle between there voltages. Thus the reference angle
of both slack buses is set to zero.
The test network shown in Fig. 5.1 is simulated using DIgSILENT PowerFactory
software. Simulation results are presented in this chapter, the focus being the
performance of the VSC-HVDC at steady state, load changes and AC side
disturbances.
2Cdc
u dc
2Cdc
Fig. 5.1 Grid set up for testing the VSC-HVDC controller
As the intent of this thesis is the study of the integration of VSC-HVDC to alleviate
network constraints in the planning stage, it is assumed that RMS simulations are
sufficiently detailed. However, EMT simulations are made for some special study
cases. Moreover switching actions of the valves are neglected in both EMT and RMS
simulations. It has been shown in [9] that this assumption does not result in a
considerable loss of accuracy.
As discussed in chapter 4, the choice of outer controllers depends on the application.
In order to effectively use the 50 kV winding of T 3 solely for active power transport
without overloading, in Fig. 2.1, VSC-1 shall be in reactive power control mode
where the reference set point is zero. Moreover, in order to improve the voltage
profile of Area 2 50 kV grid, VSC-2 shall control the AC voltage at bus-2. The direct
voltage and the active power control can be achieved by either of the two converters.
34
Chapter 5
Testing of VSC-HVDC model controller
Thus, we have two possible control strategies:
Strategy 1:
VSC-1 controls the reactive power and the active power
VSC-2 controls the AC voltage and the direct voltage
Strategy 2:
VSC-1 controls the reactive power and the direct voltage
VSC-2 controls the active power and the AC voltage
The choice between the two strategies depends mainly on two factors. Firstly, the
accurate control of the active power exchange between the AC and DC sides without
implementing loss dependent active power set point in the control system to avoid
overloading of T 3, 50 kV winding in Fig. 2.1, Strategy-2 will be more appropriate
than Strategy-1.
Area 2 consists of cables of longer length in its distribution and sub-transmission grid
than Area 1. In addition, there are overhead lines in Area 2 but not in Area 1. Thus the
frequency of disturbances in Area 2 is expected to be higher than in Area 1.
Furthermore, disturbances in Area 2 cause larger voltage dips at bus-2 than
disturbances in Area 1 cause at bus 8. As voltage dips hinder the performance of the
VSC-HVDC controllers, the preferred strategy will be to have VSC-1 control the
direct voltage, which is Strategy-2.
In this thesis the control system is in Strategy-1 under-normal condition but will
switch to Strategy-2 in case of some disturbances, see Section 5.7.
5.2 Active power control
With the AC-voltage reference set point at 1 pu, and the reactive power set point at
0.0 pu, the setting of the active power controller VSC-1 is instantaneously changed
from +0.5 pu to -0.5 pu at t=0.1 s and then set to +0.5 pu at t=0.25 s. In real operation,
grid operators change the active power rather slowly than instantaneously via power
run up and run back controllers. Thus simulation results presented here show transient
changes in active power rather than grid operator interventions. The results are shown
in Fig. 5.2.
From the simulation results, it can be seen that the system works stably with reversal
of power. When the steps are applied the active power flow is adjusted to the new
setting within 2.5 line cycles. The phase currents at both sides are affected after the
step change is applied. The change in active power is reflected in the d -components
of the AC currents, and the direct current. Because of the decoupled control almost no
effect is seen in the q -component VSC-1 current, i.e. reactive power control. The q component of the VSC-2 current has changed to keep the AC-voltage at bus-2 at its
set point. The response of the direct voltage only shows some minor transients at the
beginning of the step change of the active power. The step change also results in
minor transients in the voltages of bus-1 and bus-2, the transient at bus-2 is large
because of the slow AC-voltage controller compared to the active power controller.
35
Chapter 5
Testing of VSC-HVDC model controller
1.02
1.02
1.01
1.01
1.00
0.99
0.00
1.00
0.10
0.20
bus-1: Voltage Magnitude in p.u.
bus-2: Voltage Magnitude in p.u.
0.30
[s]
0.40
0.99
0.00
0.75
0.75
0.50
0.50
0.25
0.25
0.20
0.30
[s]
0.40
0.10
0.20
0.30
VSC-1: Current, q-Axis in p.u.
VSC-1: Current, d-Axis in p.u.
VSC-1: Phase Current, Magnitude in p.u.
[s]
0.40
0.10
0.20
VSC-2: Current, d-Axis in p.u.
VSC-2: Current, q-Axis in p.u.
Dc Bus: current, idc in p.u.
[s]
0.40
0.00
0.00
-0.25
-0.25
-0.50
-0.50
-0.75
0.00
0.10
Dc Bus: Vdc
0.10
0.20
Grid side-1: Active Power in p.u.
Grid side-2: Active Power in p.u.
0.30
[s]
0.40
-0.75
0.00
0.75
0.10
0.50
0.00
0.25
-0.10
0.00
-0.20
-0.25
-0.50
-0.30
-0.40
0.00
-0.75
0.00
0.10
0.20
Grid side-1: Reactive Power in p.u.
Grid side-2: Reactive Power in p.u.
0.30
[s]
0.40
0.30
Fig. 5.2 Transient responses for a step change in active power
5.3 Reactive power control
In this case, AC-voltage reference is held at 1 pu. The system is initially operated with
reactive power reference of +0.3 pu and active power reference of +0.5 pu. At t=0.1 s,
the reactive power reference is set to -0.3 pu and then set back to +0.3 pu at t=0.25 s.
The response of the system is shown in Fig. 5.3.
1.05
1.0015
1.04
1.0010
1.03
1.0005
1.02
1.0000
1.01
1.00
0.99
0.00
0.9995
0.10
0.20
bus-1: Voltage Magnitude in p.u.
bus-2: Voltage Magnitude in p.u.
0.30
[s]
0.40
0.9990
0.00
0.75
0.90
0.50
0.60
0.25
0.10
Dc Bus: Vdc
0.20
0.30
[s]
0.40
0.10
0.20
0.30
VSC-1: Current, q-Axis in p.u.
VSC-1: Current, d-Axis in p.u.
VSC-1: Phase Current, Magnitude in p.u.
[s]
0.40
0.10
0.20
VSC-2: Current, d-Axis in p.u.
VSC-2: Current, q-Axis in p.u.
Dc Bus: current, idc in p.u.
[s]
0.40
0.30
0.00
0.00
-0.25
-0.30
-0.50
-0.75
0.00
0.10
0.20
Grid side-1: Active Power in p.u.
Grid side-2: Active Power in p.u.
0.30
[s]
0.40
-0.60
0.00
0.60
0.40
0.50
0.20
0.40
0.30
0.00
0.20
0.10
-0.20
-0.40
0.00
0.00
0.00
0.10
0.20
Grid side-1: Reactive Power in p.u.
Grid side-2: Reactive Power in p.u.
0.30
[s]
0.40
Fig. 5.3 Transient responses for a step change in reactive power
36
0.30
Chapter 5
Testing of VSC-HVDC model controller
It is seen that the system works well even in case of change of direction of reactive
power flow. The active power is maintained at the reference values despite the step in
reactive power. The change in reactive power is reflected in the change of the q component of AC current, where as the d -component remains constant, showing the
decoupling of the d and q-axis components. Moreover, the reactive power at VSC-2
side is not affected by the change in the reactive power at the VSC-1, showing VSC-1
and VSC-2 can control their reactive powers independently. The change in the
reactive power also affects the voltage at bus-1 in such a way that the voltage
increases as reactive power is injected at bus-1 by VSC-1 at t=0.1 s and decreases as
reactive power is absorbed by VSC-1.
The step change causes transients on the direct voltage, but, as expected, the step
change of reactive power causes a much smaller transient than that with the step
change in the active power.
5.4 AC voltage control
In order to test the operation of the VSC-HVDC as an AC-voltage controller, another
test case has been studied. The setting of the AC-voltage controller for VSC-2 is
instantaneously decreased from 1 pu to 0.95 pu at t=0.1 s and then set to 1.05 pu at
t=0.25 s. The set active power flow is 0.5 pu, which is transmitted from VSC-1 to
VSC-2 and is not changed when the step is applied. The results are shown in Fig. 5.4.
1.10
1.02
1.06
1.01
1.02
0.98
1.00
0.94
0.90
0.00
0.10
0.20
bus-1: Voltage Magnitude in p.u.
bus-2: Voltage Magnitude in p.u.
0.30
[s]
0.40
0.99
0.00
0.75
0.75
0.50
0.50
0.25
0.25
0.10
Dc Bus: Vdc
0.20
0.30
[s]
0.40
0.10
0.20
VSC-1: Current, q-Axis in p.u.
VSC-1: Current, d-Axis in p.u.
VSC-1: Phase Current, Magnitude in p.u.
0.30
[s]
0.40
0.10
0.20
VSC-2: Current, d-Axis in p.u.
VSC-2: Current, q-Axis in p.u.
Dc Bus: current, idc in p.u.
0.30
[s]
0.40
0.00
0.00
-0.25
-0.25
-0.50
-0.50
-0.75
-0.75
0.00
0.10
0.20
Grid side-1: Active Power in p.u.
Grid side-2: Active Power in p.u.
0.30
[s]
0.40
0.00
0.40
0.60
0.20
0.40
0.20
0.00
0.00
-0.20
-0.20
-0.40
-0.40
-0.60
0.00
0.00
0.10
0.20
Grid side-1: Reactive Power in p.u.
Grid side-2: Reactive Power in p.u.
0.30
[s]
0.40
0.10
0.20
VSC-1: Reactive power in p.u.
VSC-2: Reactive power in p.u.
0.30
[s]
0.40
0.60
0.40
0.20
0.00
-0.20
-0.40
0.00
Fig. 5.4 Transient responses for a step change in AC voltage
37
Chapter 5
Testing of VSC-HVDC model controller
It can be seen that the step change in the reference AC voltage from 0.95 pu to
1.05 pu causes a change of VSC-2 operating point from reactive power absorption to
generation. As soon as the step change in the reference voltage is applied around
250 ms, the AC voltage is increased to the AC voltage reference value 1.05 pu after
approximately one line cycle. From the phase voltage at both sides, it can be seen that
the step change does not affect the phase voltage at VSC-1. This also shows that
VSC-1 and VSC-2 handle reactive power flows independently, as expected. The step
voltage affects the q -component of the current and not the d -component, conceding
decoupled d and q -axis controls. However, there is a change in the d -component of
the current of Grid side-1 in response to the voltage changes in order to keep the
power at its reference point, Eqn. (4.20). There is a very small transient in the direct
voltage. The transient will be larger if a small DC-capacitor would have been used.
5.5 Three phase short circuit at Grid side-1
Short circuit faults in the grid are likely the most severe disturbances for the VSCHVDC link. Thus a three-phase fault is analyzed to investigate the performance of
VSC-HVDC during such faults. A three-phase short circuit having a fault impedance
of 0 + j 1 Ω is applied at Grid side-1 at 0.1 s and is cleared at 0.2 s. The initial
operating state is such that the direct voltage reference at 1 pu, reactive power at
0.0 pu, AC-voltage at 1 pu and active power flow of +0.5 pu from VSC-1 to VSC-2.
The simulation results are shown in Fig. 5.5.
1.20
1.010
1.00
1.006
0.80
1.002
0.60
0.998
0.40
0.994
0.20
0.00
0.00
0.10
0.20
bus-1: Voltage Magnitude in p.u.
0.30
[s]
0.40
1.20
0.990
0.00
0.10
Dc Bus: Vdc
0.20
0.30
[s]
0.40
0.10
0.20
bus-2: Voltage Magnitude in p.u.
0.30
[s]
0.40
0.10
0.20
VSC-1: Reactive power in p.u.
VSC-2: Reactive power in p.u.
0.30
[s]
0.40
1.02
0.80
0.40
1.01
0.00
-0.40
1.00
-0.80
-1.20
0.00
0.10
0.20
VSC-1: Active Power in p.u.
VSC-2: Active Power in p.u.
0.30
[s]
0.40
0.99
0.00
1.25
0.30
1.00
0.20
0.75
0.10
0.50
0.00
0.25
-0.10
0.00
0.00
0.10
0.20
VSC-1: AC Magnitude in p.u.
VSC-2: AC Magnitude in p.u.
0.30
[s]
0.40
-0.20
0.00
Fig. 5.5 Responses for three phase short circuit at Grid side-1
The AC-voltage at bus-2, which is controlled to keep its terminal voltage at 1 pu, is
maintained at 1 pu except some minor oscillations at the beginning of the fault and at
38
Chapter 5
Testing of VSC-HVDC model controller
clearing the fault. The AC-voltage at bus-1 decreases to 0.2 pu during the fault and
recovers to 1 pu after clearing the fault. The current at VSC-1 increases to its
maximum limit setting, 1 pu in our case, in an effort to keep the scheduled power
transfer. The current at VSC-2 decreases to a lower value to maintain power balance.
The real power flow is reduced to a lower value during the fault and recovers to
0.5 pu successfully after the fault is cleared. The DC-link voltage shows a minor
increase in voltage level during the fault due the increase in the current magnitude of
VSC-1. VSC-1’s reactive power increases in order to keep the reactive power at the
set point, i.e. absorbs reactive power.
The above results highly depend on the current limiting strategy, section 4.3.3,
adopted and whether voltage controlled current limiters are used or not. Voltage
controlled current limiters are used when the rated current of the converters is in the
order of the short circuit current level of the AC-system to avoid special protection
co-ordination, dealt in detail in chapter 6.
5.6 Unbalanced fault conditions: SLGF at Grid side-1
In this case, the behavior of the VSC-HVDC system during unbalanced faults is
investigated. The initial operating state is such that the system is in balanced
conditions with reactive power reference at 0.0 pu, AC-voltage reference at 1 pu and
active power flow of +0.5 pu from VSC-1 to VSC-2. At t=0.1 s, a fault with a
duration of 100 ms is applied at Grid side-1. EMT simulations have been made to
obtain a better understanding of the effect of the imbalance and the assumption of a
balanced system in the design of the control system. Fig. 5.6 presents the simulation
results.
It is seen that under unbalanced conditions, the control scheme produces distorted
currents and direct voltage. During the fault, ripples at the double system frequency
(100 Hz) appear in the direct voltage. The d and q components of currents are also
oscillating at 100 Hz and are no longer decoupled constant DC quantities. The voltage
at VSC-2 is not affected by the unbalanced voltages at VSC-1. The active and reactive
powers at VSC-1 also contain the 100 Hz oscillations due to the unbalanced fault. The
reactive power at VSC-2 only shows minor transients. All oscillations in voltages and
currents at both sides, in the dq -coordinate system, means that the phase voltages and
currents at both sides are unbalanced. Furthermore note that, because converter
transformers are not included in VSC-HVDC model, there appear zero sequence
currents at VSC-1 as shown in Fig. 5.6.
The results also show that the control scheme requires some adjustments to get better
performance during unbalanced system operating conditions. The poor performance
of the current controller during system imbalance is due to the negative sequence
components in the grid voltage [8, 9, 21]. To suppress the ripple in the direct voltage
and the 100 Hz oscillation both positive and negative sequence currents need to be
controlled simultaneous and independently though these still may not totally remove
the oscillations. Mathematical origin of the oscillations is described in detail in [9].
No further attempt will be made in this thesis to study the effect of negative sequence
currents and there control scheme.
39
Chapter 5
Testing of VSC-HVDC model controller
1.40
1.016
1.20
1.010
1.00
1.004
0.80
0.998
0.60
0.992
0.40
0.20
0.00
0.10
0.20
bus-1: Voltage Magnitude in p.u.
0.30
[s]
0.40
0.986
-0.10
1.20
0.40
0.80
0.20
0.40
0.00
0.00
-0.20
-0.40
-0.40
-0.80
-0.60
-1.20
0.00
0.10
0.20
VSC-1: Active Power in p.u.
VSC-2: Active Power in p.u.
0.30
[s]
0.40
-0.80
0.00
0.00
0.10
Dc Bus: Vdc
0.20
0.30
[s]
0.40
0.10
0.20
VSC-1: Current, d-Axis in p.u.
VSC-1: Current, q-Axis in p.u.
0.30
[s]
0.40
0.10
0.20
bus-2: Voltage Magnitude in p.u.
0.30
[s]
0.40
1.50
0.40
1.00
0.20
0.50
0.00
0.00
-0.50
-0.20
-1.00
-0.40
0.00
0.10
0.20
VSC-1: Reactive Power in p.u.
VSC-2: Reactive Power in p.u.
0.30
[s]
0.40
-1.50
0.00
0.80
4.00
0.60
2.00
0.40
0.00
0.20
-2.00
0.00
-0.20
0.00
-4.00
0.00
0.10
0.20
VSC-2: Current, d-Axis in p.u.
VSC-2: Current, q-Axis in p.u.
0.30
[s]
0.40
0.10
0.20
0.30
[s]
VSC-1: +ve sequence current Magnitude in p.u.
VSC-1: Zero sequence current magnitude
VSC-1: -ve sequence current, RMS magitude in p.u.
0.40
Fig. 5.6 Transient responses for single line to ground fault at Grid side-1
5.7 Three phase short circuit at Grid side-2
A 100 ms long three phase short-circuit of fault impedance 0 + j 5 Ω is applied at
Grid side-2 at 0.1 s. The initial operating states are such that Ac voltage reference at
1 pu and the active power flow of 40 MW from VSC-1 to VSC-2. The simulation
results are shown in Fig. 5.7.
During the fault, the voltage at the grid side of VSC-2 reduced by half, 0.34 pu. The
current controller of VSC-2 increases its current in an effort to keep the active power
and AC voltage of bus-2 at there reference set point. However, because of the current
limiter of the current controller and the voltage dip at bus-2, VSC-2 is not able to
exchange sufficient power between the AC and the DC sides. Hence the direct voltage
either increases or decreases quickly depending on whether the steady sate active
power flow direction was from VSC-1 to VSC-2 or the other way. In this case, as can
be seen from Fig. 5.7, the direct voltage rises to unacceptable level.
40
Chapter 5
Testing of VSC-HVDC model controller
1.100
3.903
1.078
3.295
1.056
2.687
1.034
2.078
1.012
1.470
0.990
0.00
0.10
0.20
bus-1: Voltage Magnitude in p.u.
0.30
[s]
0.862
0.00
0.40
0.10
Dc Bus: Vdc
0.20
0.30
[s]
0.40
0.10
0.20
bus-2: Voltage Magnitude in p.u.
0.30
[s]
0.40
0.10
0.20
VSC- 1: AC Magnitude in p.u.
VSC-2: AC Magnitude in p.u.
0.30
[s]
0.40
0.999
2.00
0.861
1.00
0.722
0.00
0.583
-1.00
0.445
-2.00
0.00
0.10
0.20
VSC- 1: Active power in p.u.
VSC-2: Active power in p.u.
0.30
[s]
0.306
0.00
0.40
0.500
1.80
0.360
1.60
0.220
1.40
0.080
1.20
-0.060
1.00
-0.200
0.00
0.10
0.20
VSC- 1: Reactive power in p.u.
VSC-2: Reactive power in p.u.
0.30
[s]
0.80
0.00
0.40
Fig. 5.7 Response for three phase to ground fault at Grid side-2
In order to control the direct voltage within acceptable limits, in our case between
0.9 pu and 1.1 pu, an extra direct-voltage controller and active-power controller are
added to VSC-1 and VSC-2 respectively, which will be activated only when the direct
voltage leaves the specified range for continuous operation. VSC-1 switches from
active power control mode to direct-voltage control mode while the outer controller of
VSC-2 switches from direct-voltage control mode to active power control mode
smoothly when the direct voltage level is out of range. This will control the direct
voltage by reducing the power transferred from the Grid side-1 to the DC bus, so as
the capacitor discharges due to the new active power imbalance. Another way to
discharge the capacitor could be to temporary reverse the power exchange between
Grid side-1 and the DC-link while keeping the direction of power exchange between
Grid side-2 and the DC-link the same as pre-fault. The working principle of this
control scheme is summarized in Fig. 5.8.
Fig. 5.8 summary of direct voltage dependent control scheme
41
Chapter 5
Testing of VSC-HVDC model controller
Fig. 5.9 shows the simulation results after adopting the control modes as shown in
Fig. 5.8. It can be seen that the direct voltage is effectively controlled to be with in its
allowed range. The voltage at bus-1 only shows some minor transients at the moment
of fault applying and clearing, and so does the reactive power at VSC-1 side. The
reactive power from VSC-2 increases to support the voltage at bus-2 due to the outer
AC-voltage controller.
1.100
1.111
1.070
1.071
1.040
1.031
1.010
0.991
0.980
0.952
0.950
0.00
0.10
0.20
bus-1: Voltage Magnitude in p.u.
0.30
[s]
0.912
0.00
0.40
1.50
1.120
1.00
0.945
0.50
0.10
Dc Bus: Vdc
0.20
0.30
[s]
0.40
0.10
0.20
bus-2: Voltage Magnitude in p.u.
0.30
[s]
0.40
0.10
0.20
VSC- 1: AC Magnitude in p.u.
VSC-2: AC Magnitude in p.u.
0.30
[s]
0.40
0.770
0.00
0.595
-0.50
-1.00
-1.50
0.00
0.420
0.10
0.20
VSC- 1: Active power in p.u.
VSC-2: Active power in p.u.
0.30
[s]
0.245
0.00
0.40
0.800
1.25
0.560
1.00
0.320
0.75
0.080
0.50
-0.160
0.25
-0.400
0.00
0.10
0.20
VSC- 1: Reactive power in p.u.
VSC-2: Reactive power in p.u.
0.30
[s]
0.00
0.00
0.40
Fig. 5.9 Transient response to three phase to ground fault at Grid side-2 after control scheme
revision
Conclusion
The above simulation results have been compared with various research works [8, 9,
22] and are found to be in agreement with most of them in key aspects. The main
deviation, compared to some papers, is found on the DC-link transient peak values
where the results in this thesis are found to be smaller. This may be accounted to the
followed optimization rule of controller to satisfy the objectives at hand, the rating of
the converters compared to the AC systems, the current limiting strategies adopted
and actual control parameters.
This chapter has presented the dynamic performance of the VSC-HVDC model
during step changes of active and reactive power orders, and during balanced and
unbalanced faults. From the simulation results it can be concluded that the VSCHVDC can fulfill fast and bi-directional power transfers and AC-voltage adjustments.
In chapter 6, the VSC-HVDC model will be used as circuit element in the full grid to
study its positive and negative impacts via extensive simulation scenarios.
42
Chapter 6
Simulation Results
6.1 Introduction
In previous chapters, the need for VSC-HVDC to solve the network constraints has
been justified. Moreover, in chapter 5 the performance of the designed VSC-HVDC
has been studied using a simplified model of the studied grid, and simulation results
showed that the performance of the VSC-HVDC model is reasonable and inagreement with other researchers works.
In this chapter various simulations scenarios will be carried out to investigate the
possibility of installing VSC-HVDC to mitigate network constraints. The study is
focused on: identifying the appropriate PCC of the VSC-HVDC through voltage
sensitivity analysis, the change in the network topology as a result of the VSC-HVDC
link, the effect of disturbances in the distribution network on the dynamics of the
VSC-HVDC, the short-circuit contribution of the VSC-HVDC to investigate if there
is a need to change the rating of circuit breakers and protection scheme revision,
interaction of VSC-HVDC with the wind units and the dynamics of the network and
VSC-HVDC during various faults.
The studied grid including the VSC-HVDC link is shown in Fig. 6.1 below.
2Cdc
u dc
2Cdc
Fig. 6.1 Studied grid including the VSC-HVDC link
43
Chapter 6
Simulation Results
6.2 Static voltage stability
Voltage stability is the ability of power system to maintain adequate voltage
magnitude so that when the system nominal load increases the actual power
transferred to that load will also increase. The main factor causing voltage instability
is lack of adequate reactive power supply in the system. Voltage stability can be
broadly classified into two categories: static voltage stability and dynamic voltage
stability.
In static voltage stability, slowly developing changes in the power system occur that
eventually leads to a shortage of reactive power and declining voltage. This
phenomenon can be seen from the plot of voltage versus transferred power at a load
bus. The plots are popularly known as PV curves. When the loading of the system is
increased incrementally the voltage at a specific bus decreases. Eventually, the critical
point, at which the system reactive power is short in supply, is reached where any
further increase in active power transfer will lead to very rapid decrease in voltage.
The maximum load at the verge of voltage collapse can be determined from the knee
point of the PV curve. The power at the knee point is called critical power Pcr and the
corresponding voltage is called the critical voltage Vcr [5, 23].
The voltage stability of the studied grid is investigated by incrementally and slowly
increasing the load at bus-2, PL and QL in Fig.6.2, at constant power-factor.
v0
v2
bus
2
Pdc , Qconv
P ,Q
System
PL , QL
Fig. 6.2 Simplified equivalent grid model at bus 2
Based on the analysis given in [4], assuming Pdc and Qconv to be both equal to zero for
the moment, Pcr and Vcr can be given as:
Vcr =
Pcr =
v0
(6.1)
2 + 2sin φ
v0 2 cos φ
X (2 + 2sin φ )
(6.2)
where cos φ is the power-factor of the transferred power ( P and Q ), and X is the
reactance of the assumed equivalent line.
Eqn. (6.1 & 6.2) show that the value of Pcr depends on φ and v0 . The value of Pcr
increase as the power factor (lagging) is increased and it is further increased for
leading power factor. As mentioned earlier, VSC-HVDC is capable of supplying
44
Chapter 6
Simulation Results
(absorbing) reactive power. Thus it can increase the value of critical power by
changing the effective power factor.
Usually, placing adequate reactive power support at the weakest bus enhances staticvoltage stability margins more effectively. The weakest bus is defined as the bus,
which is nearest to experiencing a voltage collapse. Equivalently, the weakest bus is
the one that has a large ratio of differential change in voltage to differential change in
∂v
load ( ).
∂p
Fig. 6.3a illustrates the voltage stability PV curves at bus-2, bus-6 and bus-7, which
are considered as the weak area of the 50 kV grid of Area-2 due to the long overhead
transmission lines, line 7 and line 8. Moreover, Fig. 6.3b shows the voltage sensitivity
of the three buses. When the load is increased continuously from its nominal value the
voltage starts to decrease slowly. The slope of the PV curve increases sharply around
its knee. It can be seen from Fig. 6.3 that bus-2 and bus-7 are the weakest buses, i.e.
∂v
largest absolute value of
. Thus bus-2 and bus-7 are good candidates for placing
∂p
reactive power support equipments, in our case VSC-HVDC point of common
coupling (PCC) at Area 2, to improve the system static voltage stability. The system
losses will be higher when PCC is at bus-7 than at bus-2 due to the lower load at bus7 compared to the planned DC-power injection at the PCC.
PV curve for basecase
Voltage in p.u.
1
0.95
bus-2
0.9
bus-6
0.85
bus-7
0.8
knee
0.75
0.7
0.65
15
20
25
30
35
40
45
P:Load at bus-2 (MW)
(a)
V/P sensitivity
∂ v/∂ p
0
-0.01
bus-2
-0.02
bus-6
-0.03
bus-7
-0.04
-0.05
-0.06
-0.07
-0.08
15
20
25
30
35
40
45
P:Load at bus-2 (MW)
(b)
Fig. 6.3 Voltage sensitivity
In order to study the impact of the VSC-HVDC link on the static voltage stability,
three different scenarios are investigated and compared: base case (without the VSCHVDC link), VSC-HVDC as reactive power compensator, i.e. STATCOM mode of
operation ( Pdc = 0 ), and VSC-HVDC with Pdc = 40 MW.
The PV curves at bus-2, bus-6 and bus-7 for each of the above three scenarios are
illustrated in Fig. 6.4.
45
Chapter 6
Simulation Results
PV curve of bus-2
1
0.95
Voltage (p.u.)
0.9
0.85
0.8
0.75
0.7
0.65
10
Base case
VSC-HVDC (Pdc=0)
VSC-HVDC (Pdc=40MW)
low voltage limit
Nominal load
knee
20
30
40
50
60
70
50
60
70
50
60
70
P: load at bus-2
(a)
PV curve of bus-6
1
Voltage(p.u)
0.95
0.9
0.85
0.8
0.75
10
Base case
VSC-HVDC (Pdc=0)
VSC-HVDC (Pdc=40MW)
low voltage limit
Nominal load
knee
20
30
40
P: load at bus-2 (MW)
(b)
PV curve of bus-7
1
0.95
Voltage (p.u.)
0.9
0.85
0.8
0.75
0.7
0.65
10
Base case
VSC-HVDC (Pdc=0)
VSC-HVDC (Pdc=40MW)
low voltage limit
Nominal load
knee
20
30
40
P: load at bus-2 (MW)
(c)
Fig. 6.4 PV curves (a) for bus 2 (b) for bus 6 (c) bus 7
In maximum loading condition, the magnitude of bus-2, bus-6 and bus-7 voltage, i.e.
the critical voltages for each of the scenarios is shown in Fig. 6.5. It indicates that
with the application of VSC-HVDC, the voltage profile in all the three buses has
improved significantly, i.e. the transport capacity of the lines has increased because of
VSC-HVDC link; the critical voltage and critical power are higher for the case of
VSC-HVDC( Pdc = 40 MW). Furthermore, the improvement in the voltage stability
due to the DC-active power injection is smaller compared to the reactive power
injection. This shows that voltage is firmly related to reactive power flows rather than
active power.
46
Chapter 6
Simulation Results
70
VSC-HVDC(Pdc=40)
VSC-HVDC(Pdc=0)
Basecase
0.9
VSC-HVDC(Pdc=40)
VSC-HVDC(Pdc=0)
Basecase
60
0.8
50
Critical power (MW)
Critical voltage (p.u.)
0.7
0.6
0.5
0.4
0.3
40
30
20
0.2
10
0.1
0
bus-2
bus-6
0
bus-7
(b)
(a)
Fig. 6.5 Critical voltage and power
A more direct way to illustrate the voltage support function of the VSC-HVDC could
be simulating a step change in the load at bus-2 at constant power factor, see Fig. 6.6.
VSC-HVDC stabilizes the voltage at bus-2 as the AC load is continuously changed by
adjusting it reactive power support accordingly while keeping the DC-power transport
constant with in its capability curve.
0.646
50.00
0.492
40.00
0.337
30.00
0.182
20.00
0.027
-0.127
-0.100
0.180
0.460
VSC-2: Reactive power in p.u.
VSC-1: Reactive power in p.u.
0.740
1.020
[s]
1.300
10.00
-0.100
0.900
0.980
0.880
0.970
0.860
0.960
0.840
0.950
0.820
0.940
0.800
-0.100
0.180
0.460
VSC-2: Active power in p.u.
0.740
1.020
[s]
0.930
-0.100
1.300
0.180
0.460
Load at bus-2: P in MW
0.740
1.020
[s]
1.300
0.180
0.460
bus-2: Voltage Magnitude in p.u.
0.740
1.020
[s]
1.300
Fig. 6.6 Dynamic voltage support of VSC-HVDC
The above discussions and conclusion are drawn based on static voltage stability
analysis of the studied grid at normal operation condition. A thorough investigation of
static voltage stability requires N-1 contingency analysis to get the exact PV curve.
But still the system will have better static voltage stability when operated with VSCHVDC ( Pdc = 40 MW) in each outage case. Thus still the drawn conclusions are valid.
6.3 Effect of VSC- HVDC on rotor angle
Rotor angle stability concerns the ability of interconnected synchronous machines in a
power system to remain in synchronism under normal operating conditions and after
being subjected to a disturbance. The transient rotor stability is influenced by initial
47
Chapter 6
Simulation Results
rotor angle, fault location and type, fault clearing time, and post-fault transmission
reactance.
Below detailed analysis of the effect of the VSC-HVDC connection on the steady
state rotor angle is investigated. Though the distribution grid, in Area 1, is modeled as
an aggregated single generator unit, it is still considered rational to compare the
steady state rotor angles of the equivalent generator, G 1, for the cases of with and
without the VSC-HVDC link. A simplified representation of studied system in steady
state can be represented as in Fig. 6.7. The DG unit and infinite bus are represented by
a constant voltage behind a sub-transient reactance.
v8
v1 1
v0
z2
z1
E dg ∠ δ dg
z HVDC
G1
E0 ∠ δ r
Fig. 6.7 Simplified network representation for rotor angle study
The electric power output of G1 is given by
Pe = Pc + Pmax sin(δ − γ )
(6.3)
where Pc = Edg 2 Re(Y11 ), Pmax = Edg E0 | Y12 | and δ = δ dg − δ r , with δ dg and δ r the
angular displacement of the rotor from the synchronously rotating reference axis
associated with the transient internal voltages Edg and E0 . γ = θ12 − π / 2 , with
θ12 = arg (Y12 ) , Y defined as the admittance matrix between Edg and E0 . Furthermore,
the network impedances are predominantly reactive (inductive) due to the presence of
T 3 and other transformers. Thus Eqn. (6.3) can be approximate by setting
Pc = 0 and γ = 0 as:
Pe = Pmax sin(δ ), Pmax =
Edg E0
X eq
(6.4)
The impendence of the DC-link z HVDC is a non-linear function of the transferred DCpower. Recalling that the DG units is modeled as an equivalent synchronous generator
operated at constant active power mode, Pe remains constant with/without VSCHVDC link, i.e. the rotor angle of the DG unit with respect to the local bus, δ dg , are
either equal or have insignificant difference for the case of with/without VSC-HVDC.
Therefore, a change in the transferred DC power is accompanied by a change in the
rotor angle of the DG unit reference to the reference bus voltage, δ to keep Pe
constant.
The equivalent impedance of parallel impedances is always less than the individual
impedances, therefore X eq with VSC-HVDC will be always less than or equal to that
48
Chapter 6
Simulation Results
of the case without the VSC-HVDC link. Thus, the rotor angle δ is always smaller
for the case with VSC-HVDC link. This is in agreement with the calculation results
shown in Fig. 6.8.
72
62
61
70
rotor angle [deg]
rotor angle [deg]
60
68
66
rotor angle with reference
to local bus voltage in deg
64
rotor angle with reference
to reference bus voltage in deg
62
rotor angle with reference
to local bus voltage in deg
59
58
rotor angle with reference
to reference bus voltage in deg
57
56
55
60
0
5
10
15
20
25
30
35
40
power transfered via VSC-HVDC link [MW]
54
0
5
10
15
20
25
30
35
40
power transfered via VSC-HVDC link [MW]
(a)
(b)
0.955
0.95
Pe * Xeq / Edg E0 [Ω]
0.945
0.94
0.935
0.93
0.925
0.92
0.915
0
5
10
15
20
25
30
35
40
power transfered via VSC-HVDC link [MW]
(c)
Fig. 6.8 Rotor angle versus VSC-HVDC active power set point (a) for G 1 (b) G 2/G 3 and (c)
equivalent impedance versus active power set point
The above results show that the steady state rotor angle of the DG unit depends on
the transferred DC power via the VSC-HVDC link but is always less than or equal to
the case without the VSC-HVDC link. Thus VSC-HVDC improves the rotor angle
stability of G1 by reducing the steady state rotor angle. Furthermore, Fig. 6.8 shows
that the rotor angles of G2 and G3 is not affect by the VSC-HVDC link, which is an
important result.
Fig. 6.7 cannot be used to analyze the rotor angle dynamics as VSC-HVDC link
behavior depends on the controller responses. Rotor angle swings as a result of
different disturbances will be discussed in subsequent sections.
6.4 Short-circuit contribution of VSC-HVDC
For the base case, i.e. without the VSC-HVDC connection, it is assumed that the
existing protection scheme has sufficient discrimination and selectivity. Most of the
50 kV lines are protected by distance and impedance relays with a clearing time of
100 ms.
In order to study the effect of VSC-HVDC link on the protection scheme and on the
circuit breaker requirements for the 50 kV equipment, the short-circuit contribution of
the VSC-HVDC is investigated.
The short-circuit ratio (SCR) is defined as the short circuit capacity of the AC system
observed at PCC divided by the power rating of the converter, Eqn. (6.5).
49
Chapter 6
Simulation Results
SCR =
S AC
Sconv
(6.5)
where S AC is the short-circuit capacity of the AC system at PCC and Sconv is the
rating of the converters.
The possible maximum relative short-circuit current increment ( ΔI max ) is determined
by the short-circuit ratio (SCR). ΔI max is inversely proportional to the SCR and can be
defined as
ΔI max =
I sc _ hvdc − I sc
(6.6)
I sc
where, I sc is the short-circuit current of the original AC system alone at a three-phase
fault and I sc _ hvdc is the short-circuit current of the AC system with the converter
station connected and in operation under the same fault conditions.
Different control modes and different operating points may change the short-circuit
current contribution from the VSC. However, this contribution will not be higher than
ΔI max .
The allowable short circuit current contribution of the VSC-HVDC can be determined
in the design phase. The actual current rating will depend on system conditions and
grid operator requirements as well as investment costs. Increased short-circuit current
capability will improve the voltage stability and minimize the reduction of bus voltage
due to faults. However, this means a higher converter rating for the same steady-state
power transfer, which results in higher investments. On the other hand, the reduction
of short circuit current may lead to voltage instability and voltage collapse during
faults. With AC-voltage control, which is in our case at bus-2, the reactive current
generation increase automatically when the AC-voltage decreases. The higher the
fault current, in AC-voltage control mode, the higher the bus voltage. This higher
fault current may have a negative impact on the required circuit breaker rating.
However, the resulted higher bus voltage during the fault may have positive impact on
the voltage and power stability of the AC system and the connected electricity
consumers may suffer less from the disturbance.
A three phase short circuit is applied at PCC 1, and PCC 2 at t=0.1 s. Usually the
circuit breakers do not react to over currents instantaneously due to the pick-up time
of protective relays and the applied time grading. Therefore, it is the short-circuit
current after the converter transients settle down that should be considered to calculate
ΔI max . Thus, the short-circuit currents are measured at t=0.2 s while the active power
set point is varied from -1 pu to 1 pu. Voltage-dependent current limiters are avoided
to get the maximum contribution from VSC. The relative increment in short-circuit
current, calculated using Eqn. (6.6), at different operating points of VSC-HVDC, is
shown in Fig. 6.9.
50
Chapter 6
Simulation Results
Three phase short circuit at PCC-2
Three phase short circuit at PCC-1
6
percentage change in short circuit current Δ I(%)
percentage change in short circuit current ΔI(%)
15
10
5
0
-5
-10
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
4
2
0
-2
-4
-6
-8
-1
1
-0.8
Active power set point in p.u
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1
Active power set point in p.u
(a)
(b)
Fig. 6.9 Percentage increase in short circuit current due to the VSC-HVDC link
Fig. 6.9a shows that the maximum ΔI max occurs when the VSC-HVDC is operating as
a SVC or STATCOM, i.e. very low active power transport. ΔI max decreases as the
active power transport increases from 0 to 1 pu. ΔI max is negative; the VSC-HVDC
link reduces the short-circuit current, when the active power transport is from VSC-2
to VSC-1. This is accounted to the current drawn from the fault-point through the
VSC-2 and change in impedance of the network.
The SCR at PCC 1 and PCC 2 are 14 and 7.3, respectively. The corresponding
maximum ΔI max are 6 % and 13 % as can be seen from Fig. 6.9. These results are in
accordance with the SCR versus ΔI max curve given in [24].
For grid operators an increase in short circuit current of 13 % maximum for shortcircuit current levels of in the range of 4 kA is acceptable, and in most cases do not
require revising the protection scheme.
We can conclude that there will be no need to change the rating of the 50 kV grid
circuit breakers. Moreover, the existing protection scheme is not needed to be revised
because of the addition of the VSC-HVDC link.
6.5 Loss of DG unit in the distribution grid
A typical daily operation of DG units, based on actual measurement data, is shown in
Fig. 6.10. The DG units are switched on daily around 9:00 AM and will be switched
off around 9:00 PM. The effect of switching off of the DG units on the operation of
VSC-HVDC link and Area 2 50 kV grid is investigated in this section. Though Fig.
6.10 shows a ramp-wise change in DG output power, instantaneous switching off is
assumed here which can be considered as the worst case.
51
Simulation Results
power ( MW)
Chapter 6
00:00:00
09:00:00
21:00:00
23:45:00
hours of a day
Fig. 6.10 Typical daily power production profile of DG
The aggregated equivalent generator model of the distribution grid of Area 1 is
replaced by an equivalent, based on a heuristic aggregation principle, consisting of
one 10 MVA and two 20 MVA generators in parallel as shown in Fig. 6.11. In this
heuristic model reduction, several generators with the same or similar dynamic
structure can be represented by a single equivalent generator model. The equivalent
inertia and rating is the sum of the inertia and rating of all the generators respectively
[3]. The effects of loss of 10 MW and 20 MW of power while keeping the VSCHVDC active power set point at 40 MW are investigated by switching off the
respective DG units.
Fig. 6.11 simplified network representation for loss of DG unit study
In this section, the response of the rotor angles of generators, G 1, G 2 and G 3 (with
reference to the local bus voltage and reference bus voltage) upon loss of one of the
DG units shown in Fig. 6.12 is investigated. Moreover, the dynamic response of the
VSC-HVDC to the same disturbances is studied.
Fig. 6.12 and Fig. 6.13 show the dynamic of the generators rotor angle due to the
applied disturbances, i.e. deviations from respective steady state values. The rotor
dynamics show similar patterns for the base case and with VSC-HVDC link. After the
disturbance the rotor angles reach a new steady state which is close to the pre-fault
value, especially for G2/G3. The steady state rotor angle deviations are within 1º for
G1 which is considered to be negligible for small generators like G1. We also see
almost negligible deviations in steady state rotor angles for G2 and G3.
52
Chapter 6
Simulation Results
Rotor angle of G1 reference to Local Bus voltage
Rotor angle of G1 reference to reference bus voltage
2
base case
with VSC-HVDC
2
Rotor angle in deg ( δ - δ 0 )
Rotor angle in deg ( δ - δ 0 )
3
1
0
-1
-2
0
1
2
3
4
5
time [s]
6
7
8
base case
with VSC-HVDC
0
-2
-4
-6
-8
9
0
1
Rotor angle of G1 reference to local bus voltage
4
5
time [s]
6
7
8
9
2
base case
with VSC-HVDC
2
1
0
-1
-2
-3
0
1
2
3
4
5
time [s]
6
7
base case
with VSC-HVDC
0
Rotor angle in deg ( δ - δ 0 )
3
Rotor angle in deg ( δ - δ 0 )
3
Rotor angle of G1 reference to reference bus voltage
4
-4
2
(a)
8
-2
-4
-6
-8
-10
-12
9
0
1
2
3
4
5
time [s]
6
7
8
9
(b)
Fig. 6.12 Rotor angle dynamics of G 1 (a) for loss of 10 MW near by DG (b) for loss of 20 MW
near by DG
Rotor angle of G2/G3 reference to Local Bus voltage
Rotor angle of G2/G3 reference to reference bus voltage
0.5
Rotor angle in deg ( δ - δ 0 )
Rotor angle in deg ( δ - δ 0 )
0.5
0
base case
with VSC-HVDC
-0.5
0
1
2
3
4
5
time [s]
6
7
8
0
base case
with VSC-HVDC
-0.5
9
0
Rotor angle of G2/G3 reference to local bus voltage
3
4
5
time [s]
6
7
8
9
0.5
Rotor angle in deg ( δ - δ 0 )
Rotor angle in deg ( δ - δ 0 )
2
Rotor angle of G2/G3 reference to reference bus voltage
0.5
0
base case
with VSC-HVDC
-0.5
1
(a)
0
1
2
3
4
5
time [s]
6
7
8
0
base case
with VSC-HVDC
-0.5
9
(b)
0
1
2
3
4
5
time [s]
6
7
8
9
Fig. 6.13 Rotor angle dynamics of G 2/G 3 (a) for loss of a 10 MW near by DG (b) for loss of a
20 MW near by DG
The transmitted DC power remained at its pre-fault value despite the relatively large
swings in rotor angle of the DG unit. This is accounted to the fact that the imbalance
is compensated by power drawn via the 150 kV winding of T 3. This change in power
flow direction resulted in new post-fault steady state rotor angles with respect to the
reference bus voltage. The difference between the pre-post and post-fault rotor angles
is small, shown in Fig. 6.14, as the change in power flows are comparatively low
besides the VSC-HVDC link being of low rating compared to the high-voltage part of
the studied system.
53
Chapter 6
Simulation Results
without disturbance
new steady state value after loss of 10MW DG unit
new steady state value after loss of 20MW DG unit
without disturbance
70
Rotor angle in deg of G2/G3 at steady state (δ 0 )
Rotor angle in deg of G1 at steady state (δ 0 )
70
60
50
40
30
20
10
0
new steady state value after loss of 10MW DG unit
new steady state value after Loss of 20MW DG unit
60
50
40
30
20
10
Base case
0
with VSC-HVDC
(a)
Base case
with VSC-HVDC
(b)
Fig. 6.14 Steady state rotor angles (a) for G 1 (b) for G 2/G 3
It is also observed that the disturbance did not result in any noticeable changes in the
dynamics of the VSC-HVDC link. Thus we can conclude that disturbances in the
distribution grid will not affect the operation of the VSC-HVDC link as far as it does
not result in considerable voltage dips, which is unlikely.
The grid operator can operate the system such that the active power transfer via the
VSC-HVDC link follows the available power from the DG units or even in its
STATCOM mode of operation after 9:00 PM to minimize the tie-line fees at T 3
150 kV winding from TenneT.
Fig. 6.13 shows that the damping in rotor angles is the same for the cases of with and
without the VSC-HVDC link. However, it is shown in [25, 26] that VSC-HVDC has a
possibility to significantly improve damping for certain grid configurations. The
additional control arrangements required for this, are not covered in this thesis.
6.6 Interaction with wind units
Recall that we did not consider the wind turbine units connected to Area 2, 50 kV
grid. In this section, the interaction of the largest (with respected to MVA-rating)
wind unit that is connected to line 3 and VSC-HVDC link is studied. A simplified but
sufficient model of the wind unit suitable for these simulations is proposed below.
The wind unit is modeled as a constant negative PQ load with a constant Q
corresponding to a typical lagging power factor of 0.9. In reality, the reactive power
consumption of the unit will change if the terminal voltage changes from nominal.
Then a shunt capacitor to fully compensate for the unit reactive power consumption
shall be added. In the past, some have made the coarse approximation of modeling
these units as active power injection only, at unity power factor. The rationale behind
this being that the unit is assumed reactive power neutral. However, this is only true at
1.0 pu system voltage [27, 28]. Reactive power limits can also be added to the model
to take into account the limited reactive power that can be supplied by wind unit’s
converter. Furthermore, when the injected current reaches the converter limits, the
model will switch to a constant current negative load model. Accordingly, the wind
unit model used in this thesis is as shown in Fig. 6.15 below.
54
Chapter 6
Simulation Results
I > I m ax
Fig. 6.15 Simplified model of wind unit
In variable speed turbines, wind speed fluctuations are not directly translated into
output power fluctuations. Only if the rotor speed varies, the active power will vary.
Due to the rotor inertia, the rotor acts as an energy buffer, i.e. rapid wind speed
variations hardly affect the rotor speed and therefore hardly observed in the output
power [29].
The dynamics of the wind units during operating in Region-1 & Region-2 of Fig. 6.16
under normal system operating conditions are slow compared to the time constants of
the VSC-HVDC. The VSC-HVDC has enough time to change its operating point
accordingly within its capability limit based on system requirements [29].
1.6
1.4
Active Power [p.u.]
Power in wind
1.2
Rated power
1
0.8
0.6
0.4
Region-2
Region-1
0.2
0
Rated speed
cut-in speed
cut-out speed
Wind speed [m/s]
Fig. 6.16 Typical power curves of wind unit
The wind unit reduces the transport of active power. This results in reduced voltage
drops in the long radial Area 2, 50 kV grid due to the reduced current flow. The effect
of cut-out of the wind unit, due to excessive wind speeds, on the VSC-HVDC link is
shown in Fig. 6.17. The wind unit is taken out of service instantaneous at t=0.18 s.
The reactive power supplied by VSC-2 increases as a result of the AC voltage
controller response.
55
Chapter 6
Simulation Results
1.01
bus-1
Voltage Magnitud [p.u]
Voltage Magnitud [p.u]
1.1
1.05
1
0.95
0
0.05
0.1
0.15
0.2
time [s]
0.25
0.3
0.35
Voltage M agnitude [p.u]
Active power [p.u.]
0
0.05
0.1
0.15
0.2
time [s]
0.25
0.3
0.35
0.4
VSC-1
VSC-2
0
-0.5
0
0.05
0.1
0.15
0.2
time [s]
0.25
0.3
0.35
bus-2
0.975
0.97
0.965
0.96
0.4
0
0.05
0.1
0.15
0.2
time [s]
0.25
0.3
0.35
0.4
0.9
Current Magnitude [p.u.]
0.3
Reactive Power [p.u.]
0.99
0.98
0.5
0.2
VSC-1
VSC-2
0.1
0
-0.1
1
0.98
0.4
1
-1
Dc bus
0
0.05
0.1
0.15
0.2
time [s]
0.25
0.3
0.35
0.4
0.88
VSC-1
VSC-2
0.86
0.84
0
0.05
0.1
0.15
0.2
time [s]
0.25
0.3
0.35
0.4
Fig. 6.17 Response to cut-out of wind unit
Furthermore, a three- phase short circuit lasting 100 ms is applied in Area 2, 50 kV
grid. The fault is applied at t=0.1 s and is cleared at t=0.2 s. Two cases are studied; in
the first case the under-voltage protection relay with setting V<0.8 pu, t>80 ms
disconnects the wind unit while in the second case the wind unit stays connected to
the grid. Fig. 6.18 shows the voltage at the terminals of the wind unit taking into
account the under-voltage relays.
1
Wind unit disconnected
Wind unit in service
Wind unit terminal: Voltage Magnitud [p.u]
0.8
0.6
0.4
0.2
0
0
0.05
0.1
0.15
0.2
time [s]
0.25
0.3
0.35
0.4
Fig. 6.18 Voltage at wind unit bus during short circuit
Fig. 6.19 shows the response of the VSC-HVDC upon the application of the shortcircuit fault described above. The response of the VSC-HVDC while the undervoltage relay is in operation is almost the same as the case where it is not in operation.
The responses are exactly the same for the two cases until t=0.18 s when the wind unit
is disconnected. There is a difference in the reactive power injected by VSC-2 for the
two cases, where it is higher for the case of loss of the wind unit for the reason
described in Fig. 6.17. The active power injection by VSC-2 has not affected the
operation of the under-voltage relay due to its active power controller. The direct
56
Chapter 6
Simulation Results
Dc bus: Voltage Magnitud [p.u]
voltage starts rising when the fault is applied. Due to control strategy of the outercontrollers as described in Section 5.7 the direct voltage will be controlled within
acceptable range.
1.1
Wind unit in service
bus-1: Voltage
Magnitud [p.u]
Wind unit disconnected
1.05
1
0.95
0
0.05
0.1
0.15
0.2
time [s]
0.25
0.3
0.35
1
0
0.05
0.1
0.15
0
-0.5
in service:VSC-1
in service:VSC-2
disconnected:VSC-1
disconnected:VSC-2
0.3
0.35
0.4
-1
0.9
0.8
Wind unit disconnected
0.7
Wind unit in service
0
0.05
0.1
0.15
0.2
time [s]
0.25
0.3
0.35
0.4
0
0.05
0.1
0.15
1
Wind unit in service:VSC-1
Wind unit disconnected:VSC-1
0.5
Wind unit in service:VSC-2
0
0
0.05
0.1
0.15
0.2
time [s]
0.25
0.3
0.35
0.4
1.2
Wind unit disconnected:VSC-2
Current Magnitude [p.u.]
Reactive Power [p.u.]
0.25
1
Wind unit
Wind unit
Wind unit
Wind unit
0.5
-0.5
0.2
time [s]
1.1
1
bus-2: Voltage
Magnitude [p.u]
Active power [p.u.]
Wind unit disconnected
1.05
0.95
0.4
1.5
-1.5
Wind unit in service
1.1
0.2
time [s]
0.25
0.3
0.35
1
0.8
Wind unit in service:VSC-2
0.6
Wind unit disconnected:VSC-1
Wind unit disconnected:VSC-2
0.4
0.4
Wind unit in service:VSC-1
0
0.05
0.1
0.15
0.2
time [s]
0.25
0.3
0.35
0.4
Fig. 6.19 VSC-HVDC link dynamics up on three phase short circuit
The penetration level of the wind unit installed at line 3 in Area 2 is 15 % which is
regarded as small in practice. The penetration level is calculated as
% penetration level =
Pwind
×100
PArea 2
(6.7)
where Pw in d is the total active power generated by the wind unit installed at line 3
and PA r e a 2 is the total load in Area 2.
The above results in conjugation with the penetration level of the wind unit being
15 % show that the influence of the dynamics of the wind unit on the VSC-HVDC in
this particular case is negligible, and needs no further detailed consideration in the
planning studies presented in this thesis.
6.7 STATCOM mode of operation
The VSC-HVDC scheme is capable of working as a STATCOM in case one of the
converters is either blocking or disconnected. The effect of losing one VSC due to
internal faults and external faults is discussed in this section through several case
studies.
VSC-1 is disconnected together with its AC-filter banks by opening the main circuit
breakers due to an internal fault at t=0.1 s. Fig. 6.20 & Fig. 6.21 shows the dynamics
of the converters and generators.
57
Chapter 6
Simulation Results
The rotor angle of G 1 with reference to its local bus voltage reaches a new steady
state after the fault is cleared. The new steady state value differs from the pre-fault
value only marginally. This is accounted to the small change in the magnitude of
current flows. However the rotor angle reference to the reference bus voltage changes
by noticeable value for the same reasons discussed in Section 6.3. The rotor angles of
G 2 & G 3 hardly change due to such disturbances as the converters rating and DCpower transfer are relatively small compared to that of G 2 and G 3. Moreover, the
network topology as seen from G 2 and G 3 is not that much affected by the VSCHVDC link. This is verified via simulation results as shown in Fig. 6.20.
The VSC-HVDC operates stably after disconnection of the VSC-1 from bus 1 as
shown in Fig. 6.21. The reactive power supplied by VSC-2 increases to control the
voltage to its pre-fault set point. Thus VSC-2 will work as continuous reactive voltage
support equipment within its capability curve. The overvoltage at bus 1 is due to the
maximum modulation index being set to 1. The overvoltage can be avoided by
dynamically controlling the modulation index maximum value as a function of the AC
voltage.
62.11
60.53
58.96
57.38
55.81
54.23
0.00
1.25
G2: Rotor angle with reference to reference bus voltage in deg
G2: Rotor angle with reference to local bus voltage in deg
2.50
3.75
[s]
5.00
0.00
1.25
CHP DG unit: Rotor angle with reference to local bus voltage in deg
CHP DG unit: Rotor angle with reference to reference bus voltage in deg
2.50
3.75
[s]
5.00
72.08
68.72
65.36
61.99
58.63
55.27
Fig. 6.20 Generator rotor angle dynamics due to loss of VSC-1
1.06
1.020
1.04
1.012
1.02
1.004
1.00
0.996
0.98
0.988
0.96
0.94
0.00
0.980
0.10
0.20
bus-1: Voltage Magnitude in p.u.
0.30
[s]
0.00
0.40
0.957
0.10
Dc Bus: Vdc
0.20
0.30
[s]
0.40
0.969
0.574
0.964
0.191
0.959
-0.191
0.954
-0.574
0.949
-0.957
0.00
0.10
0.20
VSC-2: Active Power in p.u.
VSC-1: Active Power in p.u.
0.30
[s]
0.40
0.944
0.40
1.000
0.30
0.780
0.20
0.560
0.10
0.340
0.00
0.120
-0.10
-0.100
0.00
0.10
0.20
VSC-1: Reactive Power/Terminal AC in p.u.
VSC-2: Reactive Power/Terminal AC in p.u.
0.30
[s]
0.40
0.00
0.10
0.20
bus-2: Voltage Magnitude in p.u.
0.30
[s]
0.40
0.00
0.10
0.20
VSC-1: Current Magnitude in p.u.
VSC-2: Current Magnitude in p.u.
0.30
[s]
0.40
Fig. 6.21 VSC-HVDC dynamics due to loss of VSC-1
58
Chapter 6
Simulation Results
A three phase short-circuit fault is applied on line 9 in Fig. 6.1 at t=0.1 s and is
cleared at t=0.2 s by opening the corresponding line circuit breakers. The response of
the converter controllers and the generators rotor angles is shown in Fig. 6.22 & Fig.
6.23 below. Comparing Fig. 6.20 & Fig. 6.21 with Fig. 6.22 & Fig. 6.23, we can see
that the VSC-HVDC link works stability after sever system disturbances which result
in loss or blocking of VSC-1. Moreover, it supplies reactive power to keep flat
voltage profile in the Area 2, 50 kV grid.
1.25
1.050
1.00
1.036
0.75
1.022
0.50
1.008
0.25
0.994
0.00
-0.25
0.00
0.10
0.20
bus-1: Voltage Magnitude in p.u.
0.30
[s]
0.40
0.980
0.00
0.957
0.980
0.574
0.971
0.191
0.961
-0.191
0.952
-0.574
-0.957
0.00
0.20
0.30
[s]
0.40
0.10
0.20
bus-2: Voltage Magnitude in p.u.
0.30
[s]
0.40
0.10
0.20
VSC-1: Current Magnitude in p.u.
VSC-2: Current Magnitude in p.u.
0.30
[s]
0.40
0.943
0.10
0.20
VSC-2: Active Power in p.u.
0.30
[s]
0.40
0.933
0.00
0.60
1.210
0.40
0.968
0.20
0.726
0.00
0.484
-0.20
0.242
-0.40
0.00
0.10
Dc Bus: Vdc
0.10
0.20
VSC-1: Reactive Power/Terminal AC in p.u.
VSC-2: Reactive Power/Terminal AC in p.u.
0.30
[s]
0.40
0.000
0.00
Fig. 6.22 Converter dynamics upon three phase to ground fault on line 9
62.09
60.40
58.70
57.01
55.32
53.63
0.00
1.00
G2: Rotor angle with reference to reference bus voltage in deg
G2: Rotor angle with reference to local bus voltage in deg
2.00
3.00
4.00
[s]
5.00
0.00
1.00
2.00
CHP DG unit: Rotor angle with reference to local bus voltage in deg
CHP DG unit: Rotor angle with reference to reference bus voltage in deg
3.00
4.00
[s]
5.00
82.48
75.64
68.81
61.98
55.14
48.31
Fig. 6.23 Rotor angle dynamics upon three phase to ground fault on line 9
A third case is studied to scrutinize the impact of losing reactive power and active
power supplied to Area 2, 50 kV grid due to blocking or operation of AC-circuit
breakers at the PCC of VSC-2. A three-phase short circuit is applied on line 10 in Fig.
59
Chapter 6
Simulation Results
6.1 at t=0.1 s and is cleared at t=0.2 s by opening the corresponding circuit breakers,
which result in loss of the DC-power and the reactive power support of VSC-2 to bus2. Thus a new steady-state voltage is reached after the fault is cleared that is less than
the pre-fault voltage level. Moreover, the rotor angle swing of G1 is less than that of
Fig. 6.23 due to the DC-link as shown in Fig. 6.24 below. Fig. 6.25 shows the
response of the VSC-HVDC link due to the applied disturbance.
1.044
0.881
0.717
0.554
0.391
0.228
0.00
0.10
bus-2: Voltage, Magnitude in p.u.
C\bus-4: Voltage, Magnitude in p.u.
A\bus-5: Voltage, Magnitude in p.u.
B\bus-6: Voltage, Magnitude in p.u.
bus-7: Voltage, Magnitude in p.u.
0.20
0.30
[s]
0.40
62.02
60.33
58.64
56.95
55.26
53.57
0.00
1.00
2.00
G2: Rotor angle with reference to reference bus voltage in deg
G2: Rotor angle with reference to local bus voltage in deg
3.00
4.00
[s]
5.00
1.00
2.00
CHP DG unit: Rotor angle with reference to local bus voltage in deg
CHP DG unit: Rotor angle with reference to reference bus voltage in deg
3.00
4.00
[s]
5.00
72.67
69.17
65.67
62.18
58.68
55.18
0.00
Fig. 6.24 Effect of loss of VSC-2 on rotor angles and voltage levels
1.08
1.020
1.04
1.012
1.00
1.004
0.96
0.996
0.92
0.988
0.88
0.84
0.00
0.10
0.20
bus-1: Voltage Magnitude in p.u.
0.30
[s]
0.40
0.957
0.980
0.00
0.10
Dc Bus: Vdc
0.20
0.30
[s]
0.40
0.10
0.20
bus-2: Voltage Magnitude in p.u.
0.30
[s]
0.40
0.10
0.20
VSC-1: Current Magnitude in p.u.
VSC-2: Current Magnitude in p.u.
0.30
[s]
0.40
1.003
0.574
0.849
0.191
0.694
-0.191
0.540
-0.574
-0.957
0.00
0.385
0.10
0.20
VSC-2: Active Power in p.u.
VSC-1: Active Power in p.u.
0.30
[s]
0.40
0.231
0.00
0.375
1.042
0.250
0.861
0.125
0.681
0.000
0.501
-0.125
0.321
-0.250
-0.375
0.00
0.10
0.20
0.30
VSC-1: Total Reactive Power/Terminal AC in p.u.
VSC-2: Total Reactive Power/Terminal AC in p.u.
[s]
0.40
0.140
0.00
Fig. 6.25 VSC-HVDC link dynamics upon loss of VSC-2
60
Chapter 6
Simulation Results
6.8 Voltage dip
Voltage dips are caused by power system disturbances such as short circuits. The
impact of voltage dips on system operation strongly depends on the location and the
type of disturbance. Voltage dips caused by disturbances in the transmission grid will
cover large area while disturbances in the low voltage network are hardly noticed in
the local MV-grid.
Voltage dips are characterized by the depth, duration and frequency of occurrence.
The depth of the dip is determined by the fault location, feeder impedance and fault
level. The duration of a dip mainly depends on the setting of the protection system.
Severe faults are caused by balanced three-phase faults. However, most faults are
unbalanced and less severe.
Most of the faults in the 380 kV are of single phase to ground due to the large
physical distance between the phases on overhead transmission lines. An overview of
voltage dip depth and durations in the various voltage level in the studied grid can be
found in [30], which will also be used in this section. Voltage dips in the 380 kV grid
can be considered to originate mainly from unbalanced faults on the 380 kV lines or
distant short-circuits in the grid. To simulate distant short-circuits, a voltage dip is
created through impedance division at bus 0, i.e. adding external impedance at bus 0
with R/X ratio the same as the infinite bus representation of the external grid to avoid
phase jumps as shown in Fig. 6.26. The values of the shunt impedance are given as,
Xd =
k1 VLL 2
1
1 − k1 S 1 + k2 2
(6.8)
Rd = k2 X d
where X d and Rd are as shown in Fig. 6.26, k1 is voltage dip in pu, k2 is the R/X
ration which is kept constant, VLL is the line to line voltage of infinite grid and S is
the short circuit power of the infinite grid.
v0
v0
Xi
Rd
Ri
1∠0 p.u.
Xd
Fig. 6.26 Voltage dip propagation
A voltage dip in the 380 kV will propagate in the whole grid including PCCs of VSCHVDC, bus 1 and bus 2, as depicted dotted lines in Fig. 6.26. This reduces the
61
Chapter 6
Simulation Results
capability of the VSC-HVDC in transferring active power, and also requires special
attention during the design of the control system. Short-circuits in the 50 kV grid of
Area 2 will not cause considerable voltage dips in Area 1 due to the large impedances
of transformers, T 1/T 2 and T 4.
A Voltage dip of 0.8 pu (0.8 pu residual voltage) with 100 ms duration is created on
bus 0 at t=0.1 s through impedance division as described above. The response of
VSC-HVDC and the rest of the grid for the base case (without VSC-HVDC), and with
VSC-HVDC link is shown in Fig. 6.27 and Fig. 6.28, where the dotted lines represent
the response for the base case.
Fig. 6.27 shows that voltage dip at bus 0 is 0.8 pu, and the phase jump at bus 0 due to
the impedance division deviates from zero only marginally as the R/X ratio is kept
constant. It also shows that the dip propagates to bus 1 and bus 2. The remaining
voltage at bus 2 is larger for the case of VSC-HVDC link than the base case due to its
reactive power support as can be seen from Fig. 6.28. The remaining voltage at bus 1
is larger for the base case than the case with VSC-HVDC link due to the reactive
power set point of VSC-1 controller being set to zero. The rotor angle G 1 has the
same dynamics for both cases, thus the VSC-HVDC link which results in ring
network topology does not worsen the system response upon 380 kV voltage dip as
the two AC-sides are isolated via its DC-link.
1.055
1.013
0.982
0.968
0.909
0.923
0.835
0.878
0.762
0.833
0.689
0.00
0.10
0.20
bus-1: Voltage, Magnitude in p.u.
bus-2: Voltage, Magnitude in p.u.
bus-1: Voltage, Magnitude in p.u.
bus-2: Voltage, Magnitude in p.u.
0.30
[s]
0.40
0.788
79.62
62.98
75.77
61.96
71.92
60.95
68.07
59.94
64.22
58.92
60.37
0.00
0.00
0.10
0.20
.\bus 0: Voltage, Magnitude in p.u.
.\bus 0: Voltage, Magnitude in p.u.
[s]
0.40
0.00
2.00
4.00
6.00
[s]
G 2: Rotor angle with reference to reference bus voltage [deg]
G 2: Rotor angle with reference to reference bus voltage [deg]
8.00
57.91
2.00
4.00
6.00
[s]
G 1: Rotor angle with reference to reference bus voltage [deg]
G 1: Rotor angle with reference to reference bus voltage [deg]
8.00
1.118
0.796
0.474
0.152
-0.170
-0.492
0.00
0.30
2.00
4.00
.\bus 0: Voltage, Angle in deg
.\bus 0: Voltage, Angle in deg
6.00
[s]
8.00
Fig. 6.27 Dynamics of studied grid upon Voltage dip
62
Chapter 6
Simulation Results
The voltage dip experienced at bus 1 is smaller than at bus 2 for both cases. Because
the current limit is the same for both VSC-1 and VSC-2, a rise in the direct voltage
will result due to the imbalance in power exchange. The direct voltage is kept within
an acceptable range by adopting the control strategy that has been introduced in
section 5.7, which results in a temporary higher power exchange between AC-side
and VSC-2 than between AC-side and VSC-1 as can be seen from Fig. 6.28.
1.50
1.104
1.00
1.081
0.50
1.059
0.00
1.036
-0.50
-1.00
-1.50
0.00
1.014
0.10
0.20
VSC- 1: Active power [p.u.]
VSC-2: Active power [p.u.]
0.30
[s]
0.40
0.991
0.00
0.717
0.10
Dc Bus: Vdc
0.20
0.30
[s]
0.40
0.30
[s]
0.40
1.05
1.00
0.525
0.95
0.333
0.90
0.142
0.85
-0.050
-0.241
0.00
0.80
0.10
0.20
VSC- 1: Reactive power [p.u.]
VSC-2: Reactive power [p.u.]
0.30
[s]
0.75
0.00
0.40
0.10
0.20
VSC- 1: Current Magnitude [p.u.]
VSC-2: Current Magnitude [p.u.]
Fig. 6.28 Dynamics of VSC-HVDC upon voltage dip
A three-phase short-circuit having a fault impedance of 0 + j 1 Ω is applied on overhead line 7 at t=0.1 s and is cleared after 100 ms by opening the line circuit breakers.
This results in overloading of line 8 accompanied by large voltage drops along the
radial 50 kV Area 2 grid. Fig. 6.29 shows the response of the system upon such a fault
for the base case ( dotted lines ) and with VSC-HVDC link ( solid lines). The base
case shows a large voltage drop in the 50 kV Area 2 after the fault is cleared. The grid
with the VSC-HVDC system is able to maintain a pre-fault voltage level in the system
after the fault is cleared. This is due to the continuous reactive power support from the
VSC-2 as shown in Fig. 6.30. the slight but acceptable voltage drop compared to the
base case at bus 2 for the case with the VSC-HVDC link is caused by an effort to
control the reactive power to use T 3 effectively for active power transport without
overloading it.
The above simulation result shows that the rating of the VSC’s is enough to provide
the reactive power demand needed to keep flat voltage profiles in the Area 2 50 kV
grid even under outage of line 7.
63
Chapter 6
Simulation Results
1.076
1.004
0.917
0.994
0.759
0.985
0.600
0.975
0.442
0.965
0.283
0.00
0.10
0.20
bus-1: Voltage, Magnitude in p.u.
bus-2: Voltage, Magnitude in p.u.
bus-1: Voltage, Magnitude in p.u.
bus-2: Voltage, Magnitude in p.u.
0.30
[s]
0.40
0.956
0.00
74.87
62.05
72.35
61.67
69.84
61.29
67.33
60.91
64.81
60.53
62.30
0.00
2.00
4.00
6.00
[s]
8.00
G 1: Rotor angle with reference to reference bus voltage [deg]
G 1: Rotor angle with reference to reference bus voltage [deg]
60.14
0.00
0.10
0.20
.\bus 0: Voltage, Magnitude in p.u.
.\bus 0: Voltage, Magnitude in p.u.
0.30
[s]
0.40
2.00
4.00
6.00
[s]
8.00
G 2: Rotor angle with reference to reference bus voltage [deg]
G 2: Rotor angle with reference to reference bus voltage [deg]
Fig. 6.29 Response of studied grid upon three phase short-circuit on line 7
1.20
1.106
0.80
1.081
0.40
1.055
0.00
1.030
-0.40
-0.80
-1.20
0.00
1.005
0.10
0.20
VSC- 1: Active power [p.u.]
VSC-2: Active power [p.u.]
0.30
[s]
0.40
0.979
0.00
0.800
0.10
Dc Bus: Vdc
0.20
0.30
[s]
0.40
0.30
[s]
0.40
1.20
1.00
0.560
0.80
0.320
0.60
0.080
0.40
-0.160
-0.400
0.00
0.20
0.10
0.20
VSC- 1: Reactive power [p.u.]
VSC-2: Reactive power [p.u.]
0.30
[s]
0.00
0.00
0.40
0.10
0.20
VSC- 1: Current Magnitude [p.u.]
VSC-2: Current Magnitude [p.u.]
Fig. 6.30 Dynamics of VSC-HVDC link due to three phase short-circuit at line
64
Chapter 7
Conclusions and Further Research
7.1 Conclusions and Recommendations
The recent development in semiconductor and control equipment has made HVDC
transmission based on voltage-source converter (VSC-HVDC) feasible. VSC-HVDC
has numerous advantages such as a controllable short-circuit current contribution, the
rapid and independent control of active and reactive power, and good power quality.
With these advantages VSC-HVDC can likely be used to solve constraints in subtransmission and distribution networks.
The technical characteristics of a coupling between two sub-transmission systems on
50 kV level through VSC-HVDC are investigated. The main objective of the VSCHVDC link is to mitigate network constraints in one of these networks. Area 1 is a
greenhouse area having a large penetration level of CHP-plants, and Area 2 feeds a
50 kV grid that is expected to be constrained in the nearby future. The dynamic
reactive power support capability of the VSC-HVDC and the available distributed
generation in Area 1 are used to solve the constraints in Area 2.
An adequate model of a VSC-HVDC taking into account the gird constraints,
identified through extensive load flow and N-1 outage calculations, is presented. The
vector control system of the converters valid for balanced RMS simulations has been
established and the performance of the VSC-HVDC under idealized network
conditions is investigated. From simulation results, it can be concluded that the
system response is fast, and the active and reactive power can be controlled
independently and bi-directionally.
Comprehensive scenarios have been investigated to assess the performance of the
VSC-HVDC in mitigating the network constraints and postpone major reinforcements
of the network infrastructure. The impact of the VSC-HVDC on major power system
planning issues; interaction with other installed units (e.g. wind generators),
protection scheme, impact on network topology are investigated.
Based on the work performed for this thesis, the following conclusions can be drawn;
•
The involved physical constraints, static voltage sensitivity analysis and
load growth trends revealed that the VSC-HVDC model shall be of back to
back topology placed in Area 2 with PCCs at bus 2 and bus 1.
•
The rating of the VSC-HVDC converters, 45 MVA is found to be
sufficient to transfer the planned 40 MW, and have enough reactive power
capacity together with the AC filter banks to deal with the worst case N-1
outage voltage drops.
65
Chapter 7
Conclusions and Further Research
•
The performance of the converter controllers under steady state and
various disturbances has been found to be in agreement with other research
works.
•
It is not required to model the distribution grid in Area 1 in detail, i.e. the
aggregated model of the distribution grid is valid for the purpose of the
planning studies presented in this thesis.
•
The steady-state and dynamics of rotor angles of the large synchronous
generators G 2/G 3 is affected only marginally due to the introduction of
VSC-HVDC. Moreover, the transmission grid power flows change only by
a relatively small proportion, 3 % because of the VSC-HVDC instigate
power flow changes.
•
The ability of the distribution grid to remain in synchronism with the rest
of the grid is improved by the VSC-HVDC link to some extent.
•
The interaction of the largest wind unit installed at line 2/line 3 with the
VSC-HVDC is not significant. This may be attributed to the low i.e. 15 %
penetration level of the wind power in the 50 kV Area 2 grid.
•
The maximum relative increase in short-circuit current in the 50 kV Area 2
grid due to the VSC-HVDC is found to be 13 %. For grid operators, an
increases of 13 % for short-circuit current levels in the range of 4 kA is
acceptable. Thus, there is no need to upgrade existing circuit breakers and
revise the protection scheme of the existing grid.
•
The response of the network to deep voltage dips in the 380 kV network is
not significantly affected by the VSC-HVDC link. The performance of the
VSC-HVDC deteriorates under such deep voltage dip. However, the
probability of occurrence of such dips is rare.
•
Technically the VSC-HVDC is found to be an innovative and sufficient
solution to mitigate the network constraints described in this thesis.
The author believes that this is the first attempt made to give a comprehensive
investigation of the use of VSC-HVDC to mitigate network constraints in a meshed
network than on idealized network conditions.
7.2 Further Research
To further evaluate the application of a VSC-HVDC for alleviating network
constraints in the studied grid, a more detailed assessment using an appropriate EMT
model, taking into account all possible grid operating conditions, need to be made.
Some of the possible future works could be listed as follows:
•
Improvement in the control system to deal with unbalanced faults and
harmonic elimination. This can be achieved by dealing with the negative and
positive sequence currents separately. This will help to supply generate
66
Chapter 7
Conclusions and Further Research
balanced currents and voltages in the case of unbalanced faults which may
avoid unnecessary tripping of the DC-link.
•
Further optimization of the controller parameters with respect to control
system stability and dynamic responses. This will help to increase the
availability of the VSC-HVDC link.
•
A little insight to the associated investments is given in this thesis. A more
detailed analysis and comparison with other solution may make management
decision easier.
•
Extending the load growth and other studied scenarios to a more detail level
may reveal other benefits or disadvantage of the VSC-HVDC solution.
67
Appendix A
Space vectors
In three phase systems, it can often be assumed that the instantaneous sum of the
components e.g. voltages add up to zero
va + vb + vc = 0∀t
(A.1)
That is, a zero sequence component is assumed not to exist. This removes one degree
of freedom, since on of the phase voltages, e.g., vc always can be expressed in the
other two: vc = −(va + vb ) . Therefore it is possible to describe the three phase system
as an equivalent two phase system, with two perpendicular axes, denoted as α and β .
It is convenient to consider these axes as the real and imaginary axes in a complex
plane. With the complex two phase representation vαβ = vα + jvβ , the three phase/two
phase transformation is given by normalized Clark transformation also called power
invariant Clark transformation as:
v αβ = vα + jvβ =
2
(va + e j 2π /3vb + e j 4π /3vc )
3
(A.2)
where vαβ indicated the stationary reference frame. A set of positive sequence
∧
components, with the α -phase quantity given by va = V cos(θ1 + φ ), dθ1 dt = ωe ,
∧
where the peak value V and the phase angle φ can be assumed constant in steady
state, resulting in the space vector
∧
v αβ = V e j (θ1 +φ )
(A.3)
This vector rotates counterclockwise with the angular synchronous frequency ωe . A
negative sequence waveform gives a similar expression, but with a minus sign in the
exponent. Other three phase quantities particularly currents can be transformed in
similar way. Fig.A.1 shows voltage and currents components in αβ -frame and dq frame.
It is often reasonable to express the space vector in some other coordinate system than
the stationary αβ reference frame. The transformation from the stationary to a
general coordinate system is given by
v g = v αβ e-jθg
(A.4)
where θ g is the phase angle of the general reference frame with respect to the real
axis of the stationary frame. If the general coordinate system rotates with the same
68
Space vectors
angular frequency ωe as the three phase system, a coordinate system called
synchronous reference frame, synchronous coordinate or dq -frame is obtained. This
coordinate system is particularly useful in the case of PWM VSC because in steady
state space vector quantities become constant.
Fixed β-axis
ta
is
-ax
ω
e
Ro
gq
tin
e
ω
iβ
vd
vβ
ta
Ro
g
tin
x
d-a
is
id
iq
θ
1
vα
iα
Fixed α-axis
Figure A.1 Transformation of axes for vector control
The transformation from stationary, αβ -frame to dq -frame is given by Park
transformation described as
∧
v dq = e − jθ1 v αβ = V e jφ
(A.5)
where θ1 is the phase angle of the dq -frame with respect to the real axis of the
stationary αβ -reference frame. This removes the rotation of the vector, which
becomes constant in steady state. The component form of the synchronous coordinate
space vector is expressed as
v dq = vd + jvq .
(A.6)
where subscripts d and q refer to the direct and quadrature axis of the synchronous
reference frame respectively.
Splitting Eqn. (4.2) into its real and imaginary parts yield the matrix relation
⎡ vd ⎤ ⎡ cos θ1 sin θ1 ⎤ ⎡ vα ⎤
⎢v ⎥ = ⎢
⎥⎢ ⎥
⎣ q ⎦ ⎣ − sin θ1 cos θ1 ⎦ ⎣ vβ ⎦
69
(A.7)
Space vectors
The reverse transformation from the synchronous coordinate to the stationary
coordinate is calculated as
v αβ = e jθ1 v dq
(A.8)
An important property of the dq transformation is that the time derivative of a space
vector, according to the chain rule, is dq transformed as
d v αβ
dt
=
d (e jθ1 v dq )
dt
= e jθ1 ( jwe v dq +
d v dq
dt
) = ( s + jwe ) v dq
(A.9)
That is “s” in αβ -frame is replaced by “ s + jwe ” in dq -frame. This can be
interpreted as transient changes plus constant rotation.
The instantaneous power p (t ) of a three phase system is equal to the sum of the
instantaneous powers produced by each of the three phases
p (t ) = va ia + vbib + vc ic
(A.10)
Transforming the current and voltages using Eqn. (A.2), the instantaneous power can
be given in terms of the components of the stationary reference frame as
p (t ) = vα iα + vβ iβ + voio
(A.11)
Because the instantaneous power is independent of the coordinate system, using
Eqn. (A.8) and Eqn. (A.11) the instantaneous power can be given in synchronous
reference frame as
p(t ) = vd id + vq iq + voio
(A.12)
The instantaneous apparent power s (t ) , which is also called complex power, is
defined in terms of the voltages and current space vectors as
s (t ) = v αβ i αβ = p (t ) + jq (t )
(A.13)
where p (t ) and q (t ) are the instantaneous active and reactive powers respectively.
The instantaneous reactive power is thus defined as the imaginary component of s (t ) .
q (t ) = vβ iα − vα iβ
(A.14)
Eqn. (A.14) can also be expressed in terms of components of synchronous reference
frame as
q(t ) = vq id − vd iq
(A.
70
Appendix B
Generator, Governor,
Excitation system data
Power
System
Stabilizer
and
Table B.1: Data for synchronous generator G 1
Parameter
S
VLL
H
xq
xd
xl
x d ''
0.19
Value
11.5 kV
51 MVA 2.1 s
2.92
2.48
0.15
parameter
xd '
x q ''
x rl
Td '
T d ''
T q ''
value
0.27
0.2
0
6s
0.036 s
0.15 s
Table B.2: Data for synchronous generators G 2/G 3
VLL
S
H
Pf
xq
xl
x d ''
21 kV
600 MVA
5.89
0.9
1.91
0.181
0.268
xd '
x q ''
x rl
Td '
T d ''
T q ''
xd
0.338
0.268
0
1.07
0.0143
0.005
2.2
Parameter
Value
parameter
value
V max
1 + T2 s
1 + T3 s
1
1 + T1s
1
R
+
−
Pref
+
−
Pmech
V min
Δω
Dt
Fig. B.1: TGVO1 governor model block diagram
V max
Δω
1
R
−
+
Pref
1
1+ T1s
1
1 + T2 s
Pmech
V min
Fig. B.2: Pcu-GAST governor model block diagram
71
Appendix B
Generator, Governor, Power System Stabilizer and Excitation system data
Se (e fd )
E1, Se1, E2, Se2
V max
U
1
1 + Tr s
−
+
+
−
Ka
1 + Ta s
−
Ke
1 + Te s
+
V min
U ref
sK f
1 + Tf s
Fig. B.3: IEET1 excitation system model block diagram
Accel. power
1+ sT3
1+T4s
1+ sT1
1+T2s
Ls max
Vcu
KsT5
1+T6s
limiter
Ls min
V cl
u
Fig. B.4: IEEEST type power system stabilizer model block diagram
Table B.3: IEEET1 excitation system data for G 1
Parameter
Value
parameter
value
Tr
Ka
Ta
Ke
Te
Kf
Tf
0.028
300
0.03
1
0.266
0.0025
1.5
E1
S e1
E2
Se2
Vr min
Vr max
6
1.5
8
2.46
-12
12
Table B.4: IEEET1 excitation system data for G 2/G 2
Parameter
Tr
Ka
Ta
Ke
Te
Kf
Tf
0.028
75
0.03
1
0.08
0.0025
1.5
parameter
E1
S e1
E2
Se2
Vr min
Vr max
value
4.5
1.5
6
2.46
-12
12
Value
72
Appendix B
Generator, Governor, Power System Stabilizer and Excitation system data
Table B.5: Pcu-GAST type governor system data
Parameter
Value
R
T1
T2
AT
Kt
V max
V min
0.047
0.4
0.1
1
2
1
0
Table B.6: TGVO1 type governor system data
Parameter
Value
R
T1
T2
T3
Dt
At
V max
V min
0.05
0.1
0.2
0.2
0
1
1
0.51
Table B.7: IEEEST type power system stabilizer data
Parameter
T1
T2
T3
T4
T5
T6
Tdu
Value
0.04
0.667
1
1
3
3
1
Parameter
T1 u
L s m in
V cl
L s m ax
Vcu
Ks
Value
0.1
-0.1
0.8
0.1
1.1
-1.7
73
Appendix C
VSC design values and controller gain parameters
Table C.1: Steady state data of VSC, phase reactor and DC capacitor
Parameter
Value
S
VLL
U dc
52 kV
100 kV
L
R
45 MVA 31.91 mH
Cdc
0.83 Ω
37.5 µF
Table C.2: AC filter data
Parameter
L filter
C filter
R filter
Qf
Q filter
Value
1.04 mH
7.94 µF
0.34 Ω
3%
15 %
Table C.3: Data of simplified external grid representation of studied network of Error! Reference
source not found.
Grid side 1
Sk
Ik ''
Parameter
Value
''
630 MVA
7 kA
R/X
Sk
0.041
360 MVA
''
Grid side 2
Ik ''
R/X
4 kA
0.21
Table C.4: Current controller data of VSC-HVDC
VSC-1
K pq
Parameter
K pd
Tid
Value
0.6
2 ms
0.6
VSC-2
K pq
Tiq
K pd
Tid
2 ms
15 %
2 ms
0.6
Tiq
2 ms
Table C.5: Data for outer controllers of VSC-HVDC
Parameter
Value
Active power
control
Kp
KI
0.015
70
Dc-voltage
control
Kp
KI
8
0.8
74
Reactive power
control
Kp
KI
0.01
50
Ac-voltage
control
Kp
KI
1
600
References
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
"DIgSILENT
PowerFactory/Reference
document/Synchronous
machine
modelling", [Available from: http://www.digsilent.de/], Germany, 2007
Reza, M., "Stability analysis of transmission systems with high penetration of
distributed generation ", Ph.D. dissertation, Delft University of Technology,
Delft, The Netherlands, 2006
Ishchenko, A., "Dynamics and stability of Distribution Networks with Dispersed
Generation", Ph.D. dissertation, Techinical University of Eindhoven, Eindhoven,
The Netherlands, 2008
Shire, T.W. "Power flow controller as an innovative solution to network
constraints ", Internal document, STEDIN, Rotterdam, The Netherlands, 2008
M.H.Haque, "Power flow control and voltage stability limit:regulating
transformer versus UPFC", IEE Proc-Gener.Transm.Distrib, Vol. 151, may 2004
Fischer de Toledo, P., "Feasibility of HVDC for city infeed", Licentiate Thesis,
Royal institute of Technology, Stockholm, 2003
Verboomen, J., "Optimisation of transmission systems by use of phase shifting
transformers", Ph.D. dissertation, Delft University of Technology, Delft, The
Netherlands, 2008
Bajracharaya, C., "Control of VSC-HVDC for wind power", M.Sc. thesis,
Norwegian University of Science and Technology, June 2008
Du, C., "The control of VSC-HVDC and its use for large industrial power system",
Licentiate thesis, Chalmers University of Technology, Göteborg, Sweden, April
2003
Pan, J., et al. VSC-HVDC control and application in meshed networks in Proc.
2008 IEEE PES general meeting 2008. Pittsburgh, Pennsylvania.
Mohan, N., T.M. Underland, and W.P. Robbins, eds. Power Electronics-converter,
design and application. 2 ed. 1995, John Wiley.
Xu, Z., "Advanced semiconductor device and topology for high power current
source converters", Ph.D. dissertation, Virginia Polytechnic Institute and State
University, Blackburg, Virginia, 2003
Norrga, S., et al., Voltage source converters in transmission application, in Proc.
2008 IEEE Power Electronics Specialists Conference. 2008: Rhodes, Greece.
D.Hansen, A., et al. "Dynamic wind turbine models in power system simulation
tool DIgSILENT", Riso National Labratory, Technical University of Denmark
Roskilde, Denmark, 2007
Pollanen, R., "Converter-flux-based current control of voltage source PWM
rectifier analysis and implementation", Ph.D. dissertation, Acta University,
Lappeenrantaensis, Finland, December 2003
Abbay, C. "Protection coordination planning with distributed generation", CETC
Varennes-Energy technology and program sector, Canada, June 2007
Morren, J., "Grid support by power electronic converters of distributed
generation units", Ph.D. dissertation, Delft University of Technology, Delft, The
Netherlands, 2006
75
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
Du, C. and E. Agneholm. Investigation of Frequency/Ac voltage control for
inverter station of VSC-HVDC. in Proc. 32nd IEEE Annual Conference on
Industrial Electronics. 2007. Paris, France.
Bohn, C. and D.P. Atherton, "Analysis package comparing PID ani-windup
strategies", IEEE Control Systems Magazine, Vol. 15(2): p. 34-40, April 1995
Harnefors, L. Control of VSC-HVDC Transmissions. in Proc. 2008 IEEE Power
Electronics Specialists Conference. 2008. Rhodes, Greece.
Xu, L., B.R. Anderson, and P. Cartwright, "VSC-Transmission operating under
unbalanced AC conditions-Analysis and Control Design", IEEE Transactions on
Power Delivery, Vol. 20(1), January 2005
Bjorklund, P.-E., K. Srivastava, and W. Quaintance. HVDC Light modeling for
dynamic performance analysis. in Proc. IEEE Power Systems Conference and
Exposition. 2006. Atlanta, Georgia.
Khan, A.A., "A simple method for tracing PV curves of a radial transmission line
", Proceedings of World Academy of Science, Engineering and Technology
(PWASET), Vol. 27, February 2008
Jiang-Häfner, Y., M. Hyttinen, and B. Pääjärvi. On the short circuit current
contribution of HVDC light. in Proc. IEEE/PES Transmission and Distribution
Conference and Exhibition; Asia Pacific. 2002. Yokohama, Japan.
"HVDC Light system interactive tutorial", [Available from: www.abb.com/hvdc],
Johansson, S.G., et al. Power system stability benefits with VSC dc-transmission
systems in Proc. CIGRE Session 2004. 2004. Paris, France.
"Wind Farm integration in British Columbia - Stages 1 & 2: Planning and
interconnection Criteria ", Report number 2005-10988-2.R01.4, issued by ABB
Electric Systems Consulting, March 2005
Lasseter, R., K. Tomsovic, and P. Piagi. "Scenarios for distributed technology
applications with Steady state and dynamic models of loads and micro-sources",
Consortium for Electrical Reliability Technology Solutions, Power system
Engineering Research center, University of Wisconsin, Madison, April 2000
Slootweg, J.G., "Wind Power Modelling and Impact on Power System Dynamics",
Ph.D. dissertation, Delft University of Technology, Delft, The Netherlands, 2003
Coster, E.J., J.M.A. Myrzik, and W.L. Kling. Grid Interaction of MV-connected
CHP-Plants during Disturbances. in IEEE PES General Meeting. 2009. Calgary,
Canada
76
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertisement