Power System Modelling and Study
Methodology for SRAS Generator Evaluation
PREPARED BY:
BABAK BADRZADEH
DOCUMENT REF:
SRAS MODELLING METHODOLOGY
VERSION:
5.0
DATE:
16/10/2014
NON-CONFIDENTIAL VERSION CREATED SEPTEMBER 2015
Power System Modelling and Study Methodology for SRAS Generator Evaluation
Important Notice
This is an AEMO internal operational document that has been made available to interested
stakeholders (with confidential or sensitive information removed) to provide transparency about
AEMO’s modelling and analysis methodology used to assess the capability of proposed system
restart ancillary services tendered for the period commencing 1 July 2015.
The original document was not created for publication and is not extensively referenced. When
reading the document you should keep this in mind.
This document or the information in it may be subsequently updated or amended. AEMO may not
apply the same modelling or methodology in any future evaluations. This document does not
constitute legal or business advice, and should not be relied on as a substitute for obtaining
detailed advice about the National Electricity Law, the National Electricity Rules, or any other
applicable laws, procedures, standards or policies. The document is incomplete due to the removal
of confidential or sensitive information, and AEMO does not guarantee its accuracy.
Accordingly, to the maximum extent permitted by law, AEMO and its officers, employees and
consultants involved in the preparation of this document:
 make no representation or warranty, express or implied, as to the currency, accuracy, reliability
or completeness of the information in this document; and
 are not liable (whether by reason of negligence or otherwise) for any statements or
representations in this document, or any omissions from it, or for any use or reliance on the
information in it.
Copyright © 2015 Australian Energy Market Operator Limited. The material in this publication may be used
in accordance with the copyright permissions on AEMO’s website.
16/10/2014
Page 2 of 56
Power System Modelling and Study Methodology for SRAS Generator Evaluation
Contents
Glossary........................................................................................................................ 4
Introduction ...................................................................................................... 5
Modelling .......................................................................................................... 6
2.1
Generating unit........................................................................................................... 6
2.1.1
2.1.2
2.1.3
2.1.4
2.1.5
2.1.6
2.1.7
Electrical generator ..................................................................................................................... 6
Generator unit transformer ........................................................................................................ 12
Generator excitation system ..................................................................................................... 17
Turbine-generator prime mover and governor .......................................................................... 18
Generator control system model verification ............................................................................. 22
Generator auxiliaries ................................................................................................................. 23
Generator protection ................................................................................................................. 26
2.2
Transmission network .............................................................................................. 33
2.2.1
2.2.2
2.2.1
2.2.2
2.2.3
2.2.4
2.2.5
Overhead transmission lines ..................................................................................................... 33
Surge arresters .......................................................................................................................... 37
Terminal stations ....................................................................................................................... 39
Reactors .................................................................................................................................... 39
Transmission network transformers .......................................................................................... 39
Network loads ............................................................................................................................ 41
Network relays ........................................................................................................................... 42
2.3
Auxiliary models ....................................................................................................... 44
2.3.1
2.3.1
Statistical switching model ........................................................................................................ 44
Frequency domain harmonic impedance plots ......................................................................... 45
2.4
Network under consideration for SRAS EOI assessment ......................................... 46
Assessment criteria .........................................................................................47
3.1
General .................................................................................................................... 47
3.2
Voltage criteria ......................................................................................................... 47
3.2.1
3.2.2
3.2.3
Types of overvoltages ............................................................................................................... 47
Continuous overvoltages ........................................................................................................... 48
Temporary and transient overvoltages...................................................................................... 48
3.3
Frequency criteria .................................................................................................... 50
3.4
Surge arrester withstand criteria ............................................................................... 51
Study methodology .........................................................................................52
4.1
Steady-state analysis ............................................................................................... 52
4.1.1
4.1.2
Study methodology.................................................................................................................... 52
Success criteria ......................................................................................................................... 52
4.2
Transient analysis .................................................................................................... 52
4.2.1
4.2.2
4.2.3
4.2.4
4.2.5
Transformer energisation .......................................................................................................... 52
Line energisation ....................................................................................................................... 53
Motor starting ............................................................................................................................ 54
Island synchronisation ............................................................................................................... 55
Load rejection ............................................................................................................................ 55
16/10/2014
Page 3 of 56
Power System Modelling and Study Methodology for SRAS Generator Evaluation
Glossary
Ra
xb
xd
armature (stator) resistance
air space reactance
direct-axis synchronous reactance
xd ,sat direct-axis saturated synchronous reactance
xq
quadrature-axis synchronous reactance
xd'
direct-axis transient reactance
xq'
quadrature-axis transient reactance
xd''
direct-axis sub-transient reactance
xq''
quadrature-axis sub-transient reactance
x
xp
armature leakage reactance
Potier reactance
xs
Ta
armature stray reactance
armature time constant
Tdo'
direct-axis transient open-circuit time constant
Tqo'
quadrature-axis transient open-circuit time constant
Tdo''
Tqo''
direct-axis sub-transient open-circuit time constant
quadrature-axis sub-transient open-circuit time constant
Td'
direct-axis transient short-circuit time constant
Tq'
quadrature-axis transient short-circuit time constant
Td''
Tq''
direct-axis sub-transient short-circuit time constant
quadrature-axis sub-transient short-circuit time constant
fn
Vt
rated frequency
open circuit terminal voltage
16/10/2014
Page 4 of 56
Power System Modelling and Study Methodology for SRAS Generator Evaluation
Introduction
This report serves as an addendum to the AEMO report ‘’ Power System Modelling and Analysis
for the 2015 System Restart Ancillary Services Procurement’’, and presents detailed power system
modelling and study methodology developed for assessment proposed SRAS sources for the 2015
SRAS procurement.
16/10/2014
Page 5 of 56
Power System Modelling and Study Methodology for SRAS Generator Evaluation
Modelling
2.1
2.1.1
Generating unit
Electrical generator
For synchronous generator modelling built-in PSCAD model with two q-axis damper windings, and
distribution of saturation on d-axis is used (no option is available for distribution of saturation on both
d- and q-axis). Shaft torsional characteristics do not have any impact from black start studies
perspective, and can be disabled in the model interface. A number of additional parameters need to
be entered in the PSCAD model compared to the parameters required for synchronous generator
modelling in PSS/e as discussed in section 2.1.1.4.
2.1.1.1 direct- and quadrature-axis reactances
To allow direct use of the PSS/e model into PSCAD the option of Generator Model in PSCAD is
used. When available R2 data should be used by the user. In the absence of R2 data R1 model
confirmed by the Generator can be used.
In some occasions no PSS/e model or reliable manufacturer datasheet exist. In these
circumstances typical data shown in Table 1 can be used. In this case it is also necessary to apply
a number of assumptions as discussed in remainder of this section.
Table 1 Typical range of synchronous generator parameters [1]
16/10/2014
Page 6 of 56
Power System Modelling and Study Methodology for SRAS Generator Evaluation
When using Table 1 one should regard to the natural relationship amongst various reactances, and
time constants as shown below.
xd  xq  xq'  xd'  xq''  xd''
1
1
1
(1)
xd
2
xq
x q'
x d'
x q''
x d''
(2)
4
(3)
 2.5
(4)
The lower range being related to the characteristics of the thermal units, and the upper range being
related to the characteristics of the hydro units.
In the absence of any information:
x q  0.9  x d
(5)
xq'  2  xd'
(6)
x q''  1.1  x d''
(7)
Note that even for practical round-rotor machines x q  x d .
2.1.1.2 Operational time constants
The following relationship exist between various operational time constants
Tdo'  Tqo'  Tqo''  Tdo''
(8)
xq''  xd''
1
Ta 
(
)
2f n  Ra
2
(9)
In the absence of any information:
Tqo'  0.1  Tdo'
(10)
16/10/2014
Page 7 of 56
Power System Modelling and Study Methodology for SRAS Generator Evaluation
Most of the analytical programs such as PSCAD and PSS/e require the operational time constants
to be entered in the form of open-circuit time constants. Manufacturer datasheets sometimes
specify the short-circuit time constants. The required conversions are stated below:
Tdo'  Td' 
xd
xd'
Tqo'  Tq' 
xq
Tdo''  Td'' 
Tqo''  Tq'' 
(11)
(12)
xq'
xd'
xd''
(13)
xq'
(14)
xq''
2.1.1.3 Generator saturation
When available the generator V-I curve can be entered in the PSCAD model with a set of (V,I)
points. In the absence of this curve, saturation characteristics used in the PSS/e model can be
converted into a set of (V,I) points as follows:
The saturation characteristic of the synchronous machines in PSS/e is provided at two values of
the flux linkage (field voltage) corresponding to 1.0 and 1.2 p.u. open-circuit voltage. The saturation
factor at 1.0 and 1.2 p.u flux linkage (field voltage) are defined as:
S (1.0) 
IB  I A
IA
(15)
S (1.2) 
I C  1.2 I A
1.2 I A
(16)
Where
IA= Current corresponding to Vt =1 p.u. on the air-gap line
IB= Current corresponding to Vt =1 p.u. on the saturation curve
IC= Current corresponding to Vt =1.2 p.u. on the saturation curve
The saturation function of the synchronous generator may be defined as follows for the saturated
region of the curve:
S  AG e BG (Vt 0.8)
(17)
where
AG 
S (1.0) 2
1.2  S (1.2)
BG  5Ln
(18)
1.2  S (1.2)
S (1.0)
(19)
16/10/2014
Page 8 of 56
Power System Modelling and Study Methodology for SRAS Generator Evaluation
An example of no-load open circuit saturation curve of the synchronous machine is shown in
Figure 1.
Figure 1 Open circuit, full load and short circuit characteristics of a synchronous generator
2.1.1.4 Additional parameters required in the PSCAD model
In the PSCAD model of the generator a number of additional parameters need to be entered as
discussed below
i)
Airgap factor
To calculate the airgap factor there is a need to calculate the Potier reactance, sometimes referred
to as the armature reactance of the stator without the rotor, first as follows:
x p  x  0.63( xd'  x )
(20)
where
x p  x
(21)
If x is not available, then
x p  0.8 xd'
(22)
If the type of the machine is known, then
For salient-pole machines:
16/10/2014
Page 9 of 56
Power System Modelling and Study Methodology for SRAS Generator Evaluation
x p  0.9 xd'
(23)
For round-rotor machines:
x p  0.7 xd'
(24)
The airgap factor can therefore be calculated as:
airgap factor 
ii)
xσ
xp
(25)
Stator resistance
PSCAD model of the generator allow entering stator resistance directly, or in the form of armature
time constant. This parameter is not required in the PSCAD model,
With known machine MVA and direct-axis sub-transient reactance the stator resistance can be
determined using Figure 2. In this case X  xd'' and R  Ra .
Figure 2 Variation of the synchronous machine X/R ratio against machine apparent power
2.1.1.5 Generator model verification
Before the generator model is used for power system studies, it is imperative to test the model to
ensure that it behaves as expected. To do so the test circuit shown in Figure 3 is implemented in
16/10/2014
Page 10 of 56
Power System Modelling and Study Methodology for SRAS Generator Evaluation
PSCAD where the under no-load condition generator is subjected to a three-phase short-circuit
fault without excitation system or governor being connected.
Figure 3 Test circuit used to confirm veracity of the generator model parameters
Now consider that the peak fault current waveform is that shown in Figure 4 where the asymmetric
peak current is 78.62 kA. This can be compared against the theoretical fault current calculated
based on the machine unsaturated d-axis sub-transient reactance as follows:
xd2  0.172 p.u.
Z base
V  15.75 kV S  137 MVA
V 2 15.75 2


 1.8106 
S
137
Ifault_three - phase_peak 
V2 2
15.75 2  2

 82.58kA
3  Zbase  x2d
3 1.8106 0.172
Calculated value is marginally higher than the PSCAD model as it neglects the resistive
component.
16/10/2014
Page 11 of 56
Power System Modelling and Study Methodology for SRAS Generator Evaluation
Main : Graphs
Igen_a
100
Igen_b
Igen_c
No load fault current (kA)
80
60
40
20
0
-20
-40
-60
-80
1.920
1.940
1.960
1.980
2.000
2.020
2.040
2.060
2.080
...
...
...
2.100
Figure 4 Synchronous generator no load three-phase peak fault current
Referring to Figure 5 the steady-state fault current is 3.4 kA. This is of the same order as obtained
using the machines unsaturated xd value as follows.
xd  2.09 p.u.
I fault_three-phase_rms 
V 2
3  Z base  xd

15.75  2
3  1.8106  2.09
 3.398 kA
Main : Graphs
5.0
Igen_a
Igen_b
Igen_c
No load fault current (kA)
4.0
3.0
2.0
1.0
0.0
-1.0
-2.0
-3.0
-4.0
-5.0
9.650
9.700
9.750
9.800
9.850
9.900
9.950
...
...
...
Figure 5 Synchronous generator no load three-phase rms fault current
2.1.2
Generator unit transformer
2.1.2.1 Components of importance for black start studies
A transformer model can be divided into two parts: representation of windings and representation of
the iron core. The first part is linear, the second one is nonlinear, and both of them are frequency
dependent. Each part plays a different role, depending on the study for which the transformer model
is required. For black start studies the transformer winding can be represented with the same level
16/10/2014
Page 12 of 56
Power System Modelling and Study Methodology for SRAS Generator Evaluation
of details as the positive-sequence PSS/e model where the frequency dependency of the winding is
neglected.
The three main features associated with the iron core are the saturation, hysteresis and eddy current
losses. In general, hysteresis loops of modern transformers have a negligible influence on the
magnitude of the magnetising current. The necessary information for modelling hysteresis
characteristic is not generally provided, and therefore excluded for the transformer modelling for
black start studies. Eddy current losses can be neglected for frequencies of up 2-3 kHz, hence
excluded from the transformer model required for black start studies. The transformer saturation
characteristic play a significant part in black start studies.
Saturation characteristic can be incorporated from test data/manufacturer’s curves, or estimating the
key parameters from transformer geometry. The former approach is a simple and convenient way of
determining worst case inrush currents at the design stage. This is because with this approach
frequency dependency of the losses is neglected. The latter approach provides marginal accuracy
gain, however, the information required for this type of modelling is often only available to the
transformer manufacturer.
2.1.2.2 Transformer inrush current characteristic
A typical transformer energisation inrush current for a Delta-Star connected transformer when
energized from delta side is shown in Figure 6. In this figure each vertical division represents 1 pu
peak rated current. Note that the main parameter of interest in an inrush current waveform is the
peak instantaneous inrush current.
Figure 6 Typical inrush current waveshape
For energisation inrush studies residual magnetism in the core needs to be accounted for. This
phenomenon is commonly known as “remanence”, or residual flux. The degree of magnetizing inrush
current during energizing is a function of:
1. The position on the supply voltage wave shape that each phase of the closing circuit breaker
actually closes on.
2. The remanence existing in each of the main legs of the transformer core. The level of remanence
in the core is determined by the conditions associated with the de-energizing event of the
transformer. Even though such conditions are usually unknown, it is useful to anticipate the worst
scenario that might be expected on any random energisation.
The remanent flux may be as much as 80 % of the nominal magnetic flux density Bn. The saturation
magnetic flux density typically occurs at values greater than 1.3 Bn. This means that when the
16/10/2014
Page 13 of 56
Power System Modelling and Study Methodology for SRAS Generator Evaluation
threshold level of 1.3 Bn is exceeded, the saturation current will start to flow. The flux curve is a
function of the actual remanent flux and the instant of switching as shown in Figure 7. At the values
mentioned above and on switching at voltage zero for example, the magnetic flux density may attain
a maximum value of (0.8 + 2.0) Bn = 2.8 Bn. The maximum peak inrush current occurs when the flux
is pushed towards 2.8 pu.
Figure 7 Inrush current as a function of remanence and instant of switching in transformer
The following section discusses three different models developed for the investigation of the
transformer energisation inrush studies.
2.1.2.3 Estimation of transformer parameters required for energisation studies:
PSCAD-EMTDC employs a model which requires the user to input data from the well-known per
phase transformer equivalent circuit. Additional parameters required are the core losses at rated
voltage, the knee point of the no load voltage/magnetising current curve, the RMS magnetizing
current at rated voltage, and the air core (saturated) reactance of the transformer windings.

Aircore reactance [2]:
o
Step-up transformer (outer winding): 𝐿𝑆 ≈ 2.5 𝐿𝐾
o
Step-down transformer (inner winding): 𝐿𝑆 ≈ 1.5 𝐿𝐾
o
Auto-transformer (high-voltage side): ): 𝐿𝑆 ≈ 4 𝐿𝐾
Where Lk is the positive-sequence short circuit reactance of the transformer.

Knee point voltage:
When transformer V-I curve is available, the knee point voltage is the point where a 10%
increase in the voltage would results in 50% increase in the magnetisation current. This
concept is illustrated in Figure 8.
16/10/2014
Page 14 of 56
Power System Modelling and Study Methodology for SRAS Generator Evaluation
Figure 8 Calculation of knee point voltage when the transformer V-I saturation curve is given
In the absence of transformer saturation curve a typical knee point of 1.2 pu is assumed.


Magnetisation current:
The percentage of primary winding current that flows through the transformer magnetizing
branch. This value can be calculated based on the open-circuit test results. Core losses:
generally provided in the transformer datasheets
All other parameters in the PSCAD model can be left unchanged irrespective of the
transformer under consideration.
2.1.2.4 Approaches for transformer saturation modelling
2.1.2.4.1 PSCAD/EMTDC V4.5 and earlier versions
i)
Star energised winding
In the PSCAD model, prediction of the transformer inrush current relies upon specifying the fluxlinkage/magnetising current characteristic. In the program, this is specified using three inter-related
variables as shown in Figure 9. The air core or saturated winding inductance LA is represented by
the straight line characteristic which bisects the flux axis at Φ K . The sharpness of the knee point is
defined by Φ M and I M which represent the peak magnetising flux and magnetising current at rated
voltage. An asymptotic function for current in the non-linear saturating region LS can be defined if LA,
Φ K , Φ M and I M are known. This asymptotic function is programmed internally within the
PSCAD/EMTDC program.
16/10/2014
Page 15 of 56
Power System Modelling and Study Methodology for SRAS Generator Evaluation
Figure 9 Saturation characteristics of a transformer
The core’s magnetic non-linearity in PSCAD standard transformer model is represented by a single
valued curve; the core’s magnetic hysteretic behaviour is not therefore directly represented. Because
of this, it is necessary to resort to alternative methods in order to simulate the effects of remnant flux
in the core.
The most commonly used approach to simulate remanence in a de-energized transformer is by
introducing controlled DC current sources. This current provides the desired level of residual flux
linkage within the transformer core. During steady state operation, the magnitude of the flux linkage
in each limb of a three-phase, three-limb transformer will vary in a sinusoidal manner. The peak
magnitude of the flux linkage in each limb will be identical and they will be displaced in time/phase
by 120 degrees. When the transformer is de-energised, the flux linkage will be “frozen” at the instant
the transformer is disconnected from the supply. To represent this “frozen” or remanence state it is
normally assumed that one limb of the transformer has a residual flux linkage of +80% of the steady
state peak value, the second limb has -80%, and the third limb has no residual flux linkage. This is
generally considered to be the “worst case” that might be expected on any random transformer
energisation.
The maximum remanence that might exist in any leg of the core is around 80% of the peak flux
generated at rated voltage. This is determined from the rated rms voltage Vr of the winding that is
being referenced for remanence. Peak flux linkage M referenced to winding at rated Vr is:
M
vr
4.44f r
(26)
where fr is rated power frequency in Hz.
The magnitude of injected dc current is determined such that the residual flux linkage in two of the
three phases corresponds to 80% of rated flux linkage as calculated by (26).
ii)
Delta energised winding
In a transformer with its primary circuit connected in delta, there is no direct access to the terminals
of each winding, therefore making it difficult if not impossible to control how much of the DC injected
current goes to each of the winding phases. In this case a second method can be used for simulating
inrush currents. In this method, two sources 180 degrees apart are connected in parallel through a
couple of breakers. One of the sources is used to obtain the required remanence flux in the core
during the pre-energisation period, while the other one represents the system from which the
transformer is going to be energized. A schematic diagram representing the above approach to
create the required flux remanences is shown in Figure 10.
16/10/2014
Page 16 of 56
Power System Modelling and Study Methodology for SRAS Generator Evaluation
Timed
Breaker
Logic
Closed@t0
BRK1
I_HV
BRK2
LV
HV
TV
100% voltage
Ea1
BRK1
80% voltage at
180 degrees
B1a
BRK2
B1b
B1c
Figure 10: Representation of DC current injections to create the required flux remanence.
The methodology described here is applicable to both conventional two winding transformers and
auto-transformers. In each case at high levels of magnetic flux linkage the core will saturate requiring
a substantial increase in the magnetising current.
2.1.2.4.2 PSCAD/EMTDC V4.6
The most recent version of PSCAD EMTDC includes a detailed Jiles- Atherton model of transformer
magnetization and hysteresis characteristics. This obviates the need for the use of fictitious dc
current sources, or an additional voltage source model which would be otherwise needed in the
earlier versions of PSCAD. This version of PSCAD was released in the middle of power system
modelling and studies carried out for evaluating proposed SRAS generators, and applied for some
of the transformer energisation studies. Note that the theoretical ’’worst-case’’ flux remanences
would still need to be set manually.
2.1.3
Generator excitation system
Depending on the data availability two different approaches can be used for modelling excitation
systems:

When transfer function block diagram and model parameters are known.
o Model all components of the excitation system including AVR, exciter (if applicable),
PSS, and all limiters such as over- and under-excitation limiters, stator current
limiter, V/Hz limiter, etc.
o Note that the design of most PSS is such that the PSS is only enabled when the
load increases above 20-25% of the nominal load.
16/10/2014
Page 17 of 56
Power System Modelling and Study Methodology for SRAS Generator Evaluation

2.1.4
When transfer function block diagram for some elements, e.g. the block diagram for one or
all of the limiters is not provided, or when model parameters are incomplete.
1. Refer to the information provided by the manufacturer for a description of operating
principles of the control system in question.
2. Nominate a generic model based on the same type of excitation system using
standard models in IEEE Std 141 or the IEEE Task Force paper on ’’Computer
Models for Representation of Digital-Based Excitation Systems’’.
3. If one or several parameters cannot be accurately determined, a sensitivity study
should be conducted. Results derived from such an approach will show whether
these parameters are of concern or their influence is negligible.
Turbine-generator prime mover and governor
In multi-machine power systems it is a standard practice to operate all speed governors on a droop
control mode. However, because of the proportional characteristic of the droop governor control, a
steady-state frequency error will remain in the system. In normal operating condition this error is
eliminated by the slower acting automatic generation control (AGC). However, this control is
inactive during very start of the black start sequence. To address this problem an isochronous or
constant frequency control mode is adopted in practice for the black start generators. This means
that the speed control loop of the machine uses an integral component as opposed to the
proportional component used for droop control mode. Additionally, with isochronous mode the
active power control loop of the governor (if applicable) is disabled.
Only one generating unit in any isolated system may be selected to isochronous control. If there
are two generators in isochronous control and the controllers are well-behaved, one may reduce its
output to zero while the other picks up the whole load. However, if the two isochronous controllers
are not well-behaved, there could be large sustained power oscillations between the two machines.
When synchronising two island systems, one must ensure that the isochronous control in one
system is switched off immediately prior to or immediately after synchronisation occurs. Generally
the larger generator should remain in isochronous control and the smaller generator should be
switched to normal droop control. When switching over to the droop control mode, the active power
control loop of the governor (if applicable) is regained.
When additional units are added, the preferred control mode for speed governors is droop control
mode. In some cases, it may be preferable to keep one large unit in isochronous control mode.
Units should not be operated in parallel with more than one unit in isochronous control mode.
Most modern turbines are furnished with digital speed governors with which a selection of either a
droop or isochronous operating mode can be carried out by means of a simple change in
command.
Further details for modelling each type of speed governor is discussed below. Note that if one or
several parameters has not been provided, a sensitivity study should be conducted. Results
derived from such an approach will show whether these parameters are of concern or their
influence is negligible.
2.1.4.1 Gas turbine
Information provided by SRAS tenderers primarily focuses on droop mode of operation. To study the
response of SRAS generators during initial period of black start a model of isochronous mode of
operation is necessary. Standard IEEE models GAST2A and GGOV1 allow for simulation of
isochronous frequency control. The level of details included in both models is similar. The
main improvement of the GGOV1 model over the GAST2A is the flexibility of the model to
provide for various governor control options (PID, PI, and P) and feedback signals (speed, speed
16/10/2014
Page 18 of 56
Power System Modelling and Study Methodology for SRAS Generator Evaluation
reference, and droop). Additionally, GGOV1 model has already been supplied by a number of SRAS
generators for representation of normal droop control mode.
GGOV1 model allows for switchover between the isochronous and droop modes of operation
which is a critical feature for the black start generators. The switch over between the isochronous
and droop modes is typically done at around 20% loading. (For precise information for each
generator refer to the particular SRAS submission).
A model of GGOV1 with switchover between the isochronous and droop modes of operation is
developed in PSCAD/EMTDC. The user need to enter the following parameters based on plant
specific information provided by each SRAS tenderer. All other parameters can be assumed
constant for all gas turbine models.


R, Permanent droop, pu
Kpgov, Governor proportional gain

Kigov, Governor integral gain

Kdgov, Governor derivative gain (often set to zero)

Tdgov, Governor derivative controller time constant (s) (often set to zero)

Kturb, Turbine gain

Wfnl, No load fuel flow (pu)

db, Speed governor deadband (Hz)

Roepn, maximum valve opening rate (pu/sec)

Rclose, maximum valve closing rate (pu/sec)
Note that the model itself decides as to which parameters apply to each of the isochronous and
droop control modes.
When tests results are provided by the SRAS tender in a machine readable format, e.g. .csv, which
clearly demonstrates the response of speed governor in the isochronous mode, the GGOV1 model
should be validated against the tests results, and model parameters should be tuned if necessary.
One of the shortcoming of the GGOV1 is that it does not account for the guide vane dynamics
which can have a significant impact during black start conditions and load pick-up. For this reason
when the information provided by the SRAS tenderer includes a transfer function block diagram of
the guide vane dynamics, it is recommended to add this part to the GGOV1 model.
The GGOV1 model includes simplified model of slow acting components of the gas turbine. These
components often have a more pronounced impact when the turbine is operating near full load
conditions. This is unlikely to occur during system restoration. The impact of simplifications applied
is therefore negligible.
In addition to the GGOV1 model, a detailed vendor specific heavy duty gas turbine model based on
Alstom approach is developed [3], and can be used for relevant turbine-governor systems.
2.1.4.2 Trip to house load
i)
Governor
For the trip to house load generators no standard or generic model exists for isochronous mode of
operation. To allow for simulation of frequency control during black start conditions the following
modifications are applied to the normal droop control mode:

The normal droop or proportional controller is disabled during the initial period of black start
conditions. For different governors this can be done in different ways including:
16/10/2014
Page 19 of 56
Power System Modelling and Study Methodology for SRAS Generator Evaluation
o
Changing the appropriate parameter (droop or proportional gain). Note that for the
droop control mode the proportional gain is the inverse of the droop. For example,
to achieve a 5% droop a gain value of 20 is needed. A gain value of 500 or above
shall be used to replicate the isochronous mode of operation. Alternatively, when
the droop is directly used in the model, it can be set to zero to emulate the
isochronous mode of operation.
o
Using a significantly large deadband which effectively disables the normal droop
control mode for small frequency changes.

Power control loop is disabled initially and enabled after the switch over to droop control
mode is made.

A switchover control mode is implemented to allow for regaining the droop control once the
load increases above a pre-defined threshold.
In the absence of suitable transfer function block diagram, the standard IEEE models [4,5] should
be used. These models include sufficiently accurate representation of all major types of the steam
governor which are available in PSCAD as built-in models. When using such models the minimum
information needs to be provided by the SRAS tender include:

Gate time constant, s

Gate opening rate, pu/s

Gate closing rate, pu/s

Deadband value for islanding mode (if applicable), Hz
ii)
Turbine
For black start studies slow dynamics of the steam turbine in terms of configuration (reheat/nonreheat, tandem-compound/cross-compound), steam chest and high pressure piping, and intercept
valve should be taken into account. In the absence of suitable transfer function block diagram, the
standard IEEE models should be used. These models include sufficiently accurate representation
of all major components of the steam turbine system which are available in PSCAD as built-in
models. The minimum information needs be provided include:

HP turbine contribution, pu

LP turbine contribution, pu

Steam chest time constant, s

Cross-over time constant, s

Reheater time constant (if applicable), s
2.1.4.3 Conventional hydro
The procedure discussed in 2.1.4.2 can be used to emulate the response of conventional hydro
governor under isochronous mode of operation.
Information provided by SRAS tenders for hydro turbines and governors are of varied standard. In
some case the transfer function block diagram is for the speed governor only whereas in some
other cases detailed representation of the prime mover dynamics is also included. For black start
studies it is necessary to account for slow dynamics associated with the turbine-governor and
penstock. These dynamics are listed below.

Penstock dynamics

Surge tank chamber dynamics

Tunnel dynamics
16/10/2014
Page 20 of 56
Power System Modelling and Study Methodology for SRAS Generator Evaluation

Penstock, tunnel, and surge tank chamber orifice losses

Surge tank chamber level beyond maximum or minimum alarm
IEEE publication ’’Hydraulic Turbine and Turbine Control Models for System Dynamic Studies’’
presents suitable models for these slow dynamics. These models are currently available in PSCAD
as built-in models. Additionally, this publication proposes suitable models for different types of
hydro governor which have been implemented in the PSCAD as built-in models.
In the absence of any information on the governor control the use of enhanced governor control
(GOV3) in the PSCAD master library is recommended. This model is suitable for the analysis of
phenomena pertaining to load pick-up and load rejection during black start studies.
Alternately, when a validated model of hydro governor is available, but no information is provided
on the turbine and penstock dynamics, a combination of the validated governor model, and the
PSCAD built-in model of the hydro turbine can be used. In this case the former model will need to
be implemented as a user-written model in PSCAD using CSMF blocks.
2.1.4.1

Pumped storage hydro
Conventional pumped storage
Conventional pumped storage hydro units have many similarities to conventional hydro plants. The
major difference is that the flow is bidirectional. Usually the same equipment is used for both
generation and pumping; thus, the synchronous generator also operates as a motor, and the hydro
turbine also operates as a pump. Both components are therefore reversible in their functionality.
Black start studies should be performed with units operating as pump and generator at the light
load and high load conditions, respectively.
In the NEM the following units are pumped storage hydro all of which are also SRAS tenderers.

Tumut 3

Wivenhoe

Kangaroo Valley
When modelling conventional pumped storage units the following should be taken into account:
 The model of generator and excitation system would be the same irrespective of the prime
mover operating mode.
 For the generating mode of operation, the approaches discussed for modelling turbinegovernor for conventional hydro units will remain valid.
 To represent the hydro units in pumping mode, the turbine-governor model for the motor
should be different from the governor model used when generating. Generally there is no
speed regulation. The operator opens and closes the gates under manual control, and the
gate position remains fixed. A model of hydro governor in pumping/ mode is not therefore
necessary.
 The models and parameters for both the generating mode and the pumping mode of
operation should be verified. It is strongly recommended to request electronic copies of
recorded measurements in particular for the pumping mode of operation, and confirm the
model performance against the recorded measurements.

Ternary pumped storage
The major difference between a ternary plant and other types of pumped storage plants is that the
ternary plant turbine and pump are on the same shaft. The plant can therefore simultaneously
operate as both the pump and turbine. All other pumped storage plant designs operate either in a
generating mode or a pumping mode, and the shaft rotates in opposite directions in these two
modes.
16/10/2014
Page 21 of 56
Power System Modelling and Study Methodology for SRAS Generator Evaluation
Because the pump and turbine are able to operate simultaneously, the hydraulic flow is more
complex than in a conventional pumped storage unit. The flow in the penstock is bi-directional as it
goes from the head water to the tail water when the unit is generating electrical power, and from
the tail water to the head water when the unit is absorbing electrical power (motoring).
The model for the ternary unit should incorporate all three modes (turbine operation only, pump
operation only, or both turbine and pump operation) in a single model.




2.1.5
In generating mode with only the turbine in operation, the model of the ternary unit is similar
to that of a conventional hydro unit.
In pumping mode with only the pump in operation, the model of the ternary unit is again
similar to that of a conventional hydro unit.
The turbine and governor will adjust the gate position to participate in the usual governor
speed control..
The main inlet valve for the pump is adjusted to obtain maximum pump efficiency. The
pump controls do not participate in the speed (frequency) control.
Generator control system model verification
After the models of excitation system and governor are developed, it is necessary to ensure that the
model behave as expected. An example of the test circuit used to test the model is shown in Figure
11. The machine is initially operated with no load. A load equal to 20% of nominal load was switched
in at t=10 s and switched out at t=50 s. The size of the load being connected and rejected can be in
the range of 20-30% of the generator nominal power.
Steam Turbine Generator
ENAB
ENAB
S2M
S2M
TAP
1.0
BRK
BRK01
Timed
Breaker
Logic
Open@t0
23.29 [MW]
BRK_GTG
Set to 1 to open breaker
1.619e-007 [MW]
2.253e-011 [MVAR]
Set to 0 to close breaker
0.0
BRK01
BRK_GTG
1.618e-007 [MW]
4.937e-016 [MVAR]
3 Phase
RMS
Supergrid voltage
0.0
14.2 [MVAR]
Figure 11 Model used for the excitation system and governor testing
The second stage of generator control system model verification, comprises connection of the
generator model to a voltage source, and application of voltage set-point step response tests.
Several operating points within the generator capability chart in terms of different levels of active
and reactive power are tested. Steps applied would need to demonstrate correct operation of both
non-limited and limited excitation system responses. This will ensure correct operation of the over16/10/2014
Page 22 of 56
Power System Modelling and Study Methodology for SRAS Generator Evaluation
and under-excitation limiters. Results should be compared with the site measurements when
available. Correct operation of V/Hz limiter can be tested by increasing the generator terminal
voltage and applying a step (down) into the frequency.
Subsequent to testing the excitation system model, the turbine-governor model of the generator
under consideration is tested against another generator with excitation system and turbinegovernor models enabled. The correct load sharing under both isochronous and droop modes of
operation is then assessed.
2.1.6
Generator auxiliaries
2.1.6.1 Motor Model Development
Auxiliary loads of gas and hydro units are generally less than 2 MW collectively, and do not cause
any concerns with respect to inrush current and voltage dip during restoration. For black start
studies involving auxiliary loads of gas and hydro units load dynamics can be neglected. These
loads can therefore be represented with resistive loads.
Auxiliary loads of thermal plants (both trip to house load black starter, and the generators which
need to be restarted subsequent to the black start unit) can be as large as 60 MW. In vast majority
of applications the largest auxiliary loads comprise direct connected induction motor loads. The
method used for starting up these large motors is often a hard start, which applies full line-to-line
voltage across the motor terminals. In practice, the auxiliary loads are not generally energised
simultaneously. To study the impact of induction motor starting, it is appropriate to focus on the
largest auxiliary loads only. In most applications the three largest auxiliary loads generally include:

Boiler feed water pump

Induced draft fan

Forced draft fan
Very occasionally other types of auxiliary loads may constitute the largest auxiliary loads. In this
circumstances such auxiliary loads should also be represented explicitly.
The information provided by the Generator is limited to the nameplate data of the induction motors
used in plant auxiliaries. This is inadequate to study power system dynamic issues caused by
motor starting. For SRAS EOI assessment a spreadsheet has been developed which automatically
converts the induction motor nameplate data into the equivalent circuit format as required in the
PSCAD model of the induction motor. The conversion is based on the formulae indicated below.
INPUT PARAMETERS
================
la = Locked Rotor Current
 = Efficiency
Sr = Rated Slip
PF_Rated = Rated Power Factor
PF_St = Locked Rotor Power Factor
Mk = Breakdown Torque
Ma = Locked RotorTorque
Xmag = Xh - (magnetising reactance)
Rs = R1 - (stator resistance)
Xs = X1 - (stator reactance)
Rr_st = R2 - (rotor starting resistance)
Xr_st = X2 - (rotor starting reactance)
Rr_fl = R2_fl - (rotor full load resistance)
Xr_fl = X2_fl - (rotor full load reactance)
16/10/2014
Page 23 of 56
Power System Modelling and Study Methodology for SRAS Generator Evaluation
PARAMETER ESTIMATION (BASED ON NEPLAN EQUATIONS)
===============================================
(27)
X1n
k
lr
lan
(28)
1
Xh n
cos  n 
sin  n
ha n
lr 
cos _st n
lan
R1n
ha n
2
lr

Ma n
lan
Mr
Mk n
Mk n
Mr
Mr
 
n
2
1.0
(29)
cos  n
(30)
1.0 Srn
(31)
hb n
ha n
R1n
2
lr 
sin _st n
lan
2
X1n
Xh n
(32)
R2n
hc n
ha n
2
Xh n 
R1n
hb n
lr 
lr 
sin _st n 
sin _st n
lan
lan
X1n
X1n
Xh n
R1n
lr 
cos _st n
lan
2
(33)
(34)
X2n
hd n
R2_fln
X2_fln
Xh n 
hc n
hb n
cos  n
R1n
2
2
Xh n  cos  n
Xh n
hd n
 X1
n
sin  n
R1n 
X1n
Xh n
(35)
2
Srn
(36)
hd n
sin  n  sin  n
X1n
Xh n
R1n
16/10/2014
cos  n
2
(37)
Page 24 of 56
Power System Modelling and Study Methodology for SRAS Generator Evaluation
In the absence of any information the overall output power of all auxiliary loads can be estimated
as follows:






Gas turbine power station 1 %
Hydro power station 1 %
Combined cycle power station 2 %
Thermal power station, gas fired 4 %
Thermal power station, oil fired 5 %
Coal fired power station 7 to 10 % (lower range for black coal and upper range for brown
coal)
With known total output power of all auxiliary loads one can estimate suitable nameplate data
based on information provided for similar size induction motors, and apply the formulae above to
convert the nameplate data into the equivalent circuit model.
To study power system dynamic issues during system restoration, aside from the need to develop
an equivalent circuit model of the induction motor, it is necessary to include the torque-speed
relationship. In vast majority of cases the information is not provided. However, it is well
understood that centrifugal pumps and fans used in power plants follow a quadratic torque-speed
characteristic. A quadratic torque-speed characteristic is therefore assumed for all induction motors
used for black start studies.
2.1.6.2 Connection to the generating unit
Unit connected generators require a second source of power to the auxiliary load bus, with the
auxiliary load served from either the auxiliary transformer or the start-up supply as shown in Figure
12. Many unit connected plants do not have a generator breaker.
When restarting generator auxiliaries during black start conditions the auxiliary loads are supplied
from the auxiliary transformer (Breaker A opened and Breaker B closed in Figure 12. Motor bus
transfer is required to safely transfer the loads from one source to another such that as soon as the
generator has started the auxiliary load is supplied from the auxiliary transformer (Breaker A closed
and Breaker B open in Figure 12).
Figure 12 Schematic diagram of a unit connected generator and connection to auxliairy loads
16/10/2014
Page 25 of 56
Power System Modelling and Study Methodology for SRAS Generator Evaluation
For black start studies it is necessary to represent the auxiliary transformer and start-up supply
transformer. The information provided often only includes the transformer MVA. Table 2 is excerpts
from IEC 60076 which indicates the positive short-circuit impedance and X/R ratio of transformers
of various ratings.
Table 2 Standard transformer impedances and X/R ratio
2.1.7
Generator protection
Performance of protective systems may be measured by the relative percentages of (a) correct and
desired relay operations, (b) correct and undesired operations, (c) wrong tripping, and (d) failure to
trip. The primary reason for the (b) and (d) categories are changes in the power system topology.
During restoration, the power system undergoes continual changes and therefore is subject to (b)
and or (d) relay operations.
It is important that the performance of relays and relay schemes be evaluated under restoration
conditions for both switching actions and fault conditions. The foreknowledge that certain system
operating conditions could cause (b) or (d) relay operation makes it possible to avoid such operating
conditions during system restoration.
To allow for investigation of relay’s performance during system restoration it is necessary to develop
transient relay models for the relays of significance. Transient relay models are most detailed types
of relay models, which take into account not only the steady-state but also the time transients of
voltages and currents. These models extract their input phasors from instantaneous samples
obtained from the PSCAD time-domain simulation which accurately represents the power system
dynamics during system restoration.
When developing transient models of the relay, the focus is only on those relays which can Ardsley
interact during system restoration. Relay without the potential to impact the restoration are not
represented. The following sections discuss model development of generator and generator
transformer relays. Refer to section 2.2.5 for relay modelling for transmission network elements.
To accurately assess the performance of relays during system restoration it is imperative to
represent the manufacturer specific relay algorithms. However, the following modules associated
with the signal processing parts of the numerical relays can be considered generic and therefore
apply from all types of generator and network relays:
 Signal conditioning filter
16/10/2014
Page 26 of 56
Power System Modelling and Study Methodology for SRAS Generator Evaluation
 Sampler and A/D
 Phasor calculations
The remainder of this sub-section discusses generator relay model development.
Note: Example models presented in the sub-sections below for each relay function is valid
for a specific relay make only. Different vendors sometimes use different algorithms to
implement seemingly the same function. Detailed vendor specific relay models were
developed for each different relay make. For brevity one example is presented in this report
for each relay function.
2.1.7.1 Loss of excitation
The relay is usually designed to trip the generator when the relationship between the alternating
voltages and currents, measured at the generator terminals, indicates that a loss-of-field condition
has occurred. ANSI has assigned device number 40 to loss-of-field relays.
One or two offset mho under impedance elements are often used for providing loss of excitation
protection. The characteristics of a typical two-stage loss of excitation protection scheme are
illustrated in Figure 13.
Figure 13 typical two-stage loss of excitation protection scheme
The following four settings are required for each zone of protection:
 impedance element diameter, Xb1
 impedance element offset Xa1
 time delay on pick-up, td1 = 0
 time delay on drop-off, tdo1
Pick-up and drop-off time delays td1 and tdo1 are associated with this impedance element. Timer
td1 is used to prevent operation during stable power swings that may cause the impedance locus
of the generator to transiently enter the locus of operation set by Xb1.
Figure 14 shows a screenshot indicating the parameters need to be entered by the user for each
zone of protection for loss of excitation relay.
16/10/2014
Page 27 of 56
Power System Modelling and Study Methodology for SRAS Generator Evaluation
Figure 14 Parameters need to be entered by the user in the PSCAD model of loss of excitation
relay
2.1.7.2
Overflux (V/Hz)
Either excessive voltage, or low frequency, or a combination of both can result in overfluxing. This
protection will protect the generator and transformer from overfluxing which may occur if the AVR
fails to reduce generator voltage when machine speed is less than rated.
A voltage to frequency ratio in excess of 1.05 pu can normally be used as an indicative of this
condition. Excessive flux arising during transient stability events are not generally a problem for the
generator. This is because such an excessive flux lasts for a few hundred milliseconds only. For
example, a generator can be subjected to a transiently high power frequency voltage immediately
after full load rejection. Sustained overfluxing can arise during black start conditions, in particular
during unit start-up and shut-down, which may last for several seconds.
In practical over-flux relays trip settings can be applied in several stages including an alarm stage,
and one or more trip stages based either of definite time or inverse time characteristics. The three
common protection schemes that are used for volts per hertz protection include:
 Definite-Time (Single-level or Dual-level)
 Inverse-Time
 Combination of definite-time and inverse time
(a) definite time
16/10/2014
Page 28 of 56
Power System Modelling and Study Methodology for SRAS Generator Evaluation
(b) Inverse time
(c) Combined characteristic
Figure 15 Common V/Hz protection relay schemes
A PSCAD model of V/Hz relay has been implemented based on this principle. The parameters
need to be entered by the user are shown in Figure 16. Some V/Hz relays includes more than
three stages of trip settings. For this reason the relay model developed allow up to nine stages of
settings.
16/10/2014
Page 29 of 56
Power System Modelling and Study Methodology for SRAS Generator Evaluation
Figure 16 Parameters required for the PSCAD model of V/Hz relay
2.1.7.3 Negative phase sequence
This protection is applied to prevent overheating due to negative sequence (second harmonic)
currents. The algorithm commonly used for negative phase sequence is similar to that of the
overflux relays where either of the define time or inverse time characteristic is used.
Compared to V/Hz relay two additional settings as follows:


To ensure tripping in the event of negative phase sequence currents only slightly in excess
of the trip setting a maximum tripping time (tmax) is introduced. This is typically in the order
of several hundred seconds.
To prevent tripping of the negative phase sequence element during unbalanced faults in
equipment adjacent to the generator, a minimum tripping time (tmin) is introduced. This is
selected so as to allow sufficient time for clearance of transformer faults by the appropriate
unit protection.
Figure 17 Parameters required for the PSCAD model of negative phase sequnce relay
2.1.7.4
Generator/transformer differential protection
Figure 18 depicts operating principles of a biased differential protection. The principles discussed
are applicable to both generator and transformer differential protection. To determine whether relay
operation is permitted or it should be restrained, two current calculations are performed. The first is
the differential current which is sum of the currents entering the equipment under consideration. The
second current is the biased or restraint current. Different manufacturers apply different calculation
methods for determination of the biased current as will be discussed in this section.
16/10/2014
Page 30 of 56
Power System Modelling and Study Methodology for SRAS Generator Evaluation
Figure 18 Operating principles of biased differential protection
Note that a differential protection can be designed based on comparison of currents or voltages. The
former is referred to as biased current differential. Excessive inrush currents caused during
transformer energisation may result in mal-operation of this type of differential relays. For this reason
it is necessary to include a sufficiently accurate model of this relay in the integrated models used for
transformer energisation studies, and confirm the correct operation of the relay.
High impedance voltage based differential protections are also available, and widely used when
there is a concern on saturation of current transformers during severe faults. As these relays are
voltage based rather than current based, they do not show any sensitivity to the level of inrush current
present during transformer energisation. For this reason these relays are not represented for black
start studies.
The differential relay characteristic is designed such that at low currents, the bias is small, thus
enabling the relay to be made sensitive. At higher currents, for example during inrush or fault
conditions, the bias used is higher, and thus the spill current required to cause operation is higher.
To achieve these characteristics four basic settings are needed as a minimum as follows:
IS1= Basic differential setting
IS2= Knee of tripping characteristic
K1 = slope 1
K2 = slope 2
Additionally, depending on the manufacturer the following settings may also be used as described
in Figure 19:
I_diff (low) = minimum differential current required for operation
I_diff (high) = diff hi-set (when I_diff > I_diff-hi operation occurs irrespective of restraint current)
16/10/2014
Page 31 of 56
Power System Modelling and Study Methodology for SRAS Generator Evaluation
Figure 19 Common settings applied to a biased differential relay
Sometimes an additional parameter referred to as first range factor is used as identified by K3 in
Figure 19. A PSCAD model of differential relay has been developed based on these settings.
Figure 20 shows a screenshot of the parameters need to be entered in the PSCAD model.
Figure 20 Parameters required for the PSCAD model of differential protection relay
Note that the additional settings discussed above are more widely used for transformer differential
protection. For generator differential protection the use of four basic settings is often adequate. In
any case the user should refer to the settings sheet provided for each particular SRAS tenderer. To
disable I_diff (low) and I_diff (high), zero and high values are entered, respectively, in the PSCAD
model of the relay.
An additional parameter entered in the PSCAD model interface is the current base. With current
bases set to 1 the values of IS1, IS2, Idiff-low and Idiff-hi are entered in Ampere as seen from CT
secondary. Alternatively, if the current based is entered in Ampere (as seen from CT secondary)
the values of IS1, IS2, Idiff-low and Idiff-hi will be entered in pu.
Depending on the manufacturer restraint current can be calculated as one of the following:
16/10/2014
Page 32 of 56
Power System Modelling and Study Methodology for SRAS Generator Evaluation
I_rest= K [abs(Ia)+abs(Ib)]
(38)
I_rest= K [Ia-Ib]
(39)
In both equation K can be either 0.5 or 1 depending on the manufacturer.
The relay settings discussed above are primarily for response of the differential relay to fault
conditions. An additional harmonic restraint logic is often applied to transformer differential
protection so as to avoid relay operation for high inrush current caused by transformer
energisation. This design feature blocks a trip output from the transformer differential relay and is
intended to prevent operation of this relay on transformer excitation current. Typical designs use
second harmonic for restraining operation, as it is the major harmonic component of transformer
inrush current. An additional feature for blocking the fifth harmonic current is sometimes provided.
Figure 21 shows additional parameters need to be entered in the PSCAD model to allow for
inclusion of inrush current restrain function. In this particular case the second harmonic restraint
was used only. To disable the fifth harmonic restraint high values are entered for the set-point and
blocking time for the n-th harmonic.
Figure 21 PSCAD model parameters for representation of inrush current restraint function
2.2
2.2.1
Transmission network
Overhead transmission lines
Transmission lines are nonlinear in nature due to frequency dependency in conductors (skin effect)
and the ground or earth return path. Two main approaches exist to represent the transmission lines
in electromagnetic transient programs. The simplest method is to use PI sections. The second and
more detailed method is to use a distributed transmission line. The PI section model is a lumped
parameter model based on series RL elements and parallel CG elements. This model can be
adopted to study the transient behaviour when the end-to-end length of the line is shorter than a
couple of km, or when studies need to be run for several tens of seconds. This model can also be
used when tower geometry is unknown.
The distributed transmission line models are based on the principle of traveling waves. Relative to
the PI models, distributed parameter models are more accurate but more computationally
intensive.
There are three types of distributed line models which can be selected in PSCAD to represent the
transmission line:
 the Bergeron model;
 the frequency-dependent (Mode) model;
 the frequency-dependent (Phase) model
A Bergeron model is a constant distributed-parameter model at the specified frequency whereas
the frequency-dependent distributed parameter model (both Mode and Phase model) are fitted for
a given frequency range, hence more accurate. The parameters of the Bergeron model are
constant and thus may be used when no frequency dependency is to be represented. However,
this model may be quite sensitive to the model frequency specified by the user.
With the extent of data available for network planning studies and PSS/e load flow cases, a PI
model or a Bergeron model can be readily developed (the latter model would need to be converted
16/10/2014
Page 33 of 56
Power System Modelling and Study Methodology for SRAS Generator Evaluation
from the PI model via a PSS/e to PSCAD conversion tool). These models are generally appropriate
for short lines so long as harmonic resonance is not a matter of concern. With a PI or Bergeron
model the line impedances need to be explicitly entered in the model. Note that for black start
studies, the negative- and zero-sequence impedances, and mutual impedances of the double
circuit lines are critical, and need to be entered in the PSCAD model should either of the PI or
Bergeron models be used.
Of three distributed line models the frequency-dependent (phase) model is the most accurate, and
should be used whenever harmonic resonance need to be investigated. The use of frequencydependent (phase) model is envisaged for all transmission lines involved in the system restoration
provided that tower geometry is given.
With frequency-dependent models of the line, the sequence impedances are not entered in the
model but calculated automatically through a line constant routine based on given tower geometry
and conduct data.
The following information is generally sought:









Transmission line conductor diameter and resistance per unit length (can also be selected
from a user-defined list for standard/commonly used conductors).
Total length of each transmission line.
Phase transformation data and distances between phase transformations.
Spacing between conductors in a phase bundle.
Spacing between phases.
Shield wire diameter and resistance per unit length.
Height of each conductor and shield wire at the tower and sag to midspan, or average
height of each conductor and shield wire above ground.
Tower dimensions.
Ground resistivity.
An example demonstrating how to enter the geometrical line data into the built-in PSCAD model is
shown in Figure 22.
16/10/2014
Page 34 of 56
Power System Modelling and Study Methodology for SRAS Generator Evaluation
Figure 22 Example demonstrating how to populate the PSCAD line parameters based on
geometrical line data
Ground resistivity is not often provided. In the absence of any data approximate values of and can
be used. The typical tower grounding resistance is between 10 and 100 Ohms. Ground resistivity
can be assumed 100 Ohms.m.
16/10/2014
Page 35 of 56
Power System Modelling and Study Methodology for SRAS Generator Evaluation
Note that all of the above information, except conductor and ground resistances and ground
resistivity, is from geometric line dimensions. The resistances of conductor and ground are either
provided in separate sheets, or can be readily obtained from the line manufacturer website.
When modelling complex conductor configurations, such as cross-phased lines and segmented
ground wires, interim sections of the line will only be represented if their length is greater than 5
km.
Care should be taken when modelling transposed transmission lines. There are essentially two
methods for modelling transposed lines in PSCAD. Both methods require that each section of the
total line to be represented as an individual line. The transpositions can be implemented in one of
two ways for this case. The first is to physically transpose the circuit interconnections between the
line sections on the PSCAD Canvas. An alternative method is to alter the conductor XY coordinates
within the properties editor of each line section. An example demonstrating the latter method is
shown in Figure 23. One benefit of this method is that phase A could remain conductor #1 throughout
the length of the transmission line.
Figure 23 Recommended approach for line transposition
Black start studies involving motor starting, and investigation of the long-term reactive power
compensation capability of the generating unit would sometimes necessitate running the studies
for several tens of seconds. The use of frequency dependent model is impracticable for such
studies. Cascaded PI models can be used instead which is effectively a series connection of a
number of PI models. Modelling the overhead line with cascaded pi-sections neglects the
frequency dependency of the line, resulting in slightly more pessimistic results as there is reduced
damping in the model. The number of PI-sections required for the correct representation of the line
depends mainly on the expected frequency of the transient oscillation. The highest frequency that
16/10/2014
Page 36 of 56
Power System Modelling and Study Methodology for SRAS Generator Evaluation
can be attained by a PI-section representation is the natural frequency of one individual element
representing a fraction of the total line length. This is given by:
f max 
1
C'
2 s L' 
2

v
4.44  s
(40)
For simplicity this is sometimes written as:
f max 
v
5 s
(41)
where:
s = individual element length
L’ = line inductance per unit length
C’ = line capacitance per unit length
v = propagation speed of electromagnetic wave ≈ 0.3 km/µs
For switching transients investigated in black start studies a maximum frequency of 30 kHz can be
expected theoretically.
2.2.2
Surge arresters
The primary reason for inclusion of surge arrester in black start studies is that system restoration
can impose excessive temporary overvoltages lasting for several seconds. Such TOVs can
compromise insulation of the surge arrester. Modelling surge arresters for voltages in the range 1.01.3 pu, and assessment of TOVs resulting from switching transients is critical.
Surge arresters exhibit a nonlinear voltage versus current (V-I) characteristic such that they have an
extremely high resistance during normal system operation and a relatively low resistance during
transient overvoltages. Examples of V-I curves for Surge arresters are shown in Figure 24 (a) and
(b), respectively.
(a) Sample 1 surge arrester V-I curve
16/10/2014
Page 37 of 56
Power System Modelling and Study Methodology for SRAS Generator Evaluation
(b) Sample 2 surge arrester V-I curve
Figure 24 V-I curve for [sample manufacturer] surge arresters
The commonly used frequency-independent surge arrester model is appropriate for simulations
involving low frequency transients and most switching frequencies. This is because the frequency
dependency will only become relevant at very high frequency overvoltages associated with lightning
strikes or transients associated with GIS substations.
In the PSCAD model the volt-amp characteristic for the Surge Arrester Component is entered in per
unit for voltage for the vertical [Y] axis and kA for the horizontal [X] axis under “I-V Characteristic“.
Manufacturers’ datasheets represent behaviour of the arrester with a number of (V,I) points in the
kA range part of the surge arrester operation. These currents generally correspond to voltages in
the range of 1.5-3.0 pu.
When referring to the manufactures’ the following voltages are to be noted:

Um: maximum system voltage. This is typically 5-10% higher than the equipment nominal
voltage.

Ur: surge arrester rated voltage: This is smaller than the Um. Note that for a given Um
manufacturers often offer surge arresters with different Ur.

Uc: continuous operating voltage. This is typically 10-20% less than the Ur.

TOV capability: expressed in both 1s and 10s. It can be used as an indication of withstand
capability of the surge arrester during black start conditions.
Care has to be taken in interpreting the data provided in the manufacturers’ datasheets. One supplier
may define their per-unit characteristic such that 1.0 pu is the crest value of rated voltage Ur. Another
supplier may define their per-unit characteristic such that 1.0 pu is the crest value of continuous
operating voltage Uc.
Under normal operating conditions a surge arrester typically conducts a very small current, typically
in the range of 1-5 mA. For the [sample manufacturer] surge arresters the characteristics shown in
Figure 24 can be used to replicate the behaviour of the surge arrester for low current TOV region.
Output for arrester absorbed energy and current is entered as “Internal Output Variables” in the
PSCAD model by designating appropriate signal names. Energy is in kJoules and current in kA if
the arrester is operating under conditions of kV and kA. Often, absorbed energy is desired in
kJoules/kV based on the “Arrester Voltage Rating“.
16/10/2014
Page 38 of 56
Power System Modelling and Study Methodology for SRAS Generator Evaluation
2.2.1
Terminal stations
For black start studies correct representation of the exact arrangement of circuit breakers and
isolators is critical. Refer to ’’High Voltage Network Main System Diagram’’ in the link below:
http://www.aemo.com.au/Electricity/Planning/Related-Information/Maps-and-Diagrams
2.2.2
Reactors
Should reactors be exposed to high temporary overvoltages (TOV), the reactor may be subject to
saturation in the same fashion as transformers. As the magnitude of overvoltages is not known
beforehand, it is prudent to represent the saturation characteristic for all shunt reactors. In most
cases the V-I saturation curve of reactors is not provided. Assumptions stated for modelling
transformer saturation can be applied. However, it is not often a common practice to replicate the
flux remanence in the model of shunt reactor.
2.2.3
Transmission network transformers
i)
Two-winding transformers
Modelling methodology discussed for generator transformer remains valid.
ii)
Three-phase three-winding autotransformer model development
A limitation of the PSCADEMTDC library which does not contain a model of an autotransformer with
a Delta-Wye-Wye configuration. The difficulties modelling three-phase autotransformers with delta
connected windings is discussed in detail in [7] which also explain how to convert the three-winding
transformer short-circuit data into leakage reactance terms between pairs of windings I, II, and III as
shown in Figure 25.
a
I
H
b
III
II
T
L
c
Figure 25: One-phase of three-phase auto transformer
The voltage ratings of each winding I, II, and III are:
VI  VH  VL
VII  VL
(42)
VIII  VT
where VH, VL and VT are per phase values.
The short circuit test between H and L is already the correct test between I and II, since II is shorted
and the voltage is applied across I with b and c being at the same potential through the short-circuit
connection. Assuming that all the impedances are given on a common MVA base, the short circuit
impedance ZHL is simply converted to the new voltage base VI:
16/10/2014
Page 39 of 56
Power System Modelling and Study Methodology for SRAS Generator Evaluation
Z I, II
 VH
 Z HL 
 VH  VL



2
(43)
in pu values.
No modifications are needed to the test between II and III:
Z II, III  Z LT
(44)
For the short-circuit test between H and T, the modification can be explained in terms of the
equivalent star-circuit of Figure 26 with the impedances being ZI, ZII, and ZIII, based on VI, VII, and
VIII. With III short circuited, 1 pu current (based on VIII = VT) will flow through ZIII. This current will
also flow through I and II as 1 pu based on VH, or converted to bases VI, and VII, II = (VH – VL)/VH
and III = VL/VH.
With these currents, the pu voltages become:
VI  ZI
VH  VL
 ZIII
VH
(45)
in pu values.
VII  ZII
VL
 ZIII
VH
(46)
in pu values.
Converting VI and VII to physical units by multiplying (45) by (VH – VL) and (46) by VL, adding them
up, and converting the sum back to a pu value based on VH produces the measured pu value
 H  VL
Z HT  Z I  H
 VH
2

V
  Z II  L

 VH
2

  Z III

(47)
in pu values.
Equations 43, 44 and 47 can be solved for ZI, ZII, ZIII since ZI,II = ZI + ZII and ZII,III = ZII + ZIII, which
produces
ZI, III  ZHL
VH VL
VH  VL 2
 ZHT
VH
VL
 ZLT
VH  VL
VH  VL
(48)
ZII
ZI
ZIII
L
H
T
Figure 26: One-phase of three-phase auto transformer
16/10/2014
Page 40 of 56
Power System Modelling and Study Methodology for SRAS Generator Evaluation
The autotransformer can therefore be modelled as a transformer with three windings I, II, III by
simply re-defining the short-circuit input impedances using equations 43, 44, and 48. This is
implemented in PSCAD as shown in Figure 27. A spreadsheet has been developed for the SRAS
technical assessment project which automatically converts the required impedances based on the
above equations.
#2
#1
#3
V = 0,9994
LV
ELV
A
V
EHV
#2
#1
V = 35
A
V
HV
#3
#2
#1
A
V
V = 0,4001
ETV
#3
TV
Figure 27 PSCAD implementation of auto-transformer based on three single-phase transformers
2.2.4
Network loads
To represent the network load models (excluding generator auxiliary loads) a static representation
of the load is adopted using the expressions below which account for voltage dependency of the
active and reactive components of the loads:
𝑉
𝑃 = 𝑃𝑜 (𝑉 )𝑁𝑝
(49)
0
𝑉
𝑄 = 𝑄𝑜 (𝑉 )𝑁𝑞
(50)
0
Po
Qo
Vo
V
Equivalent load real power
Rated real power per phase
Load voltage
Rated load voltage (rms, line-to-ground)
16/10/2014
Page 41 of 56
Power System Modelling and Study Methodology for SRAS Generator Evaluation
Figure 28 Load indices for network loads
NSW
Smelting load
Other loads
2.2.5
QLD-N
QLD-S
SA
TAS
VIC
Np=1.37, Np = 1.7,
Nq = 2.8
Nq=3.17
NA
NA
Np=1,
Nq=3
Np=1.69,
Np=1,
Nq=3
Np = 1.2,
Np = 1.3,
Nq = 3.0
Nq = 2.8
Np=1,
Nq=3
Np=1,
Nq=3
Np=1,
Nq=3
Nq=1.97
Network relays
2.2.5.1 Synchrocheck
To ensure successful breaker closure a simulation model of the synchrocheck relay has been
developed in PSCAD. The model has the following settings:

frequency difference between the two systems being synchronised must be within 0.11 Hz,
and

Phase angle difference across the breaker must be within +/- 40º.

Voltage magnitude difference of 15% when synchronising a dead system against a live
system

Voltage magnitude difference of 30% when synchronising two live systems (two viable
islands)

Delay time for frequency, voltage magnitude, and voltage phase angle is set to 1 s.
Figure 29 Relay settings for PSCAD model of synchrocheck relay
2.2.5.1 Distance relays
Faults on transmission lines are commonly detected by distance relays that measure and respond
to one or another form of the ratio of voltage to current. This ratio is impedance or a component of
impedance. A comparison of the measured impedance with the line impedance provides an
indication whether the fault is in the protected zone of the relay or not. This type of relay is also
assigned an ANSI device number of 21.
Basic distance protection will comprise instantaneous directional Zone 1 protection and one or
more time-delayed zones. Typical reach and time settings for a 3-zone distance protection are
shown in Figure 30. Digital and numerical distance relays may have up to five or six zones, some
set to measure in the reverse direction. Typical settings for three forward-looking zones of basic
distance protection are given in the following sub-sections.


Zone 1 = 80-85% of protected line impedance
Zone 2 (minimum) = 120% of protected line
16/10/2014
Page 42 of 56
Power System Modelling and Study Methodology for SRAS Generator Evaluation



Zone 2 (maximum) < Protected line + 50% of shortest second line
Zone 3 forward = 1.2 (protected line + longest second line)
Zone 3 reverse= 20% of protected line
Figure 30 Conventional Mho distance relay with three zones of protection
The characteristic shown in Figure 30 is for a Mho distance relay. This is one of the most widely
used types of distance relays. A less widely used shape sometimes used for distance relays is the
quadrilateral or lens characteristic as shown in Figure 31.
Figure 31 Quadrilateral distance relay with three zones of protection
Distance relays are commonly used for both phase and ground protection. The same characteristic
is generally used for both phase and ground protection albeit with different parameter settings.
Note that distance relay settings can be provided either in rectangular (R,X), or Polar (Z, Theta)
format. The PSCAD model of the Mho and Lens relays allow user to choose the format for data
entry.
A model of distance relay with three zones of protection is developed in PSCAD. In this example
relay has a Mho design, and settings are entered in (R,X) format. A screenshot of the settings
need to be entered by the user are shown in Figure 32.
16/10/2014
Page 43 of 56
Power System Modelling and Study Methodology for SRAS Generator Evaluation
Figure 32 Relay settings needs to be entered in the PSCAD model of Mho distance relay for the
three zones of protection
2.2.5.2 Transformer differential
See section 2.1.7.4.
2.3
2.3.1
Auxiliary models
Statistical switching model
Transient voltage and current magnitudes depend upon the instant on the voltage waveform at which
the circuit breaker contacts close electrically. To determine the most onerous switching instant and
the resulting peak voltage, it is necessary to conduct statistical switching studies. A statistical
switching case typically consists of 100 or more separate simulations, each using a different set of
circuit breaker closing times.
To account for the closing times of the three phases of the breaker two parameters are considered.
Firstly, the closing command which can happen anywhere on the waveform. Secondly, the non-ideal
behaviour of the three phases. In practice the three phases respond to the closing command with a
random delay due to mechanical dispersions and prestrikes. The most commonly used approach to
account for the two factors above is to assume the closing command is uniformly distributed on the
waveform and the actual closing times are around it, with a Gaussian probability. For the Gaussian
distributions, the standard deviation of the inter-phase time distribution is not always known. This is
typically in the range of 0.8 ms to 2 ms. For SRAS studies a typical value of 1 ms is used.
A model of the statistical circuit breaker has been developed in PSCAD/EMTDC as shown in Figure
33. The model can be applied for any transformer or line switching studies. The model requires four
input parameters as shown in Figure 34. The only parameter needs to be changed is the breaker
rated voltage (example below shows the line energisation for a 500 kV line). To use this model the
multiple-run feature in PSCAD needs to be enabled.
16/10/2014
Page 44 of 56
Power System Modelling and Study Methodology for SRAS Generator Evaluation
Figure 33 Snapshot of the PSCAD model for statistical switching
Figure 34 Parameters need to be entered into the statistical breaker model
2.3.1
Frequency domain harmonic impedance plots
In PSCAD, there is a component called “Interface to Harmonic Impedance Solution”. This
component will perform an impedance scan on the connected electrical system. It requires just a
run of two simulation steps and all power electronic devices are assumed to be in their off-state.
Further, all transformers and arresters are assumed to be in their unsaturated region.
Note that this is a static component, and performs the frequency scan based on system
configuration at t=0 s. It is therefore unable to recognise any switching which will happen during
course of dynamic simulation. For this reason it is necessary for the user to make sure that all
appropriate circuit breakers needed to establish a certain configuration are manually closed and
open such the desired network configuration is achieved at t=0 s. An example of the output of
harmonic impedance scan is shown in Figure 35.
16/10/2014
Page 45 of 56
Power System Modelling and Study Methodology for SRAS Generator Evaluation
|Z+|(ohms)
12000
10000
8000
6000
4000
2000
0
0
100
200
300
400
500
600
700
800
900
1000
1100
1200
Figure 35 Example of frequency domain harmonic impedance plot
Results obtained from frequency domain harmonic impedance should corroborated against the
time-domain individual and total harmonic distortion levels. For example, if the harmonic
impedance plot shows a sharp peak in the vicinity of the 4th harmonic, one would expect a high
level of 4th harmonic voltage distortion.
2.4
Network under consideration for SRAS EOI assessment
For the purpose of system restoration each of the five states in the National Electric Market
(NEM) are treated as one electrical sub-network except Queensland which is divided into Southern
and Northern sub-networks. From modelling standpoint each sub-network would therefore be treated
as in island without connection to adjoining sub-networks.
16/10/2014
Page 46 of 56
Power System Modelling and Study Methodology for SRAS Generator Evaluation
Assessment criteria
3.1
General
The following criteria are to be applied for system restoration studies discussed in Section 4:
(1) Power system security criteria
(a)
A secure operating state as defined in S4.2.4 of the National Electricity Rules is
considered impracticable to achieve under all operating conditions during system
restoration. The criteria defined in S4.2.2 of the National Electricity Rules for a
satisfactory operating state will be applied to all states of restoration process except
the criteria for permissible frequency and voltage range defined below.
(b)
A credible contingency event as defined in S4.2.3 (b) of the National Electricity Rules
will apply to system restoration studies. Partial or full loss of the restoration path may
happen during non-credible contingency events as defined in S4.2.3 (e) of the
National Electricity Rules. For both credible and non-credible contingency events it is
expected that all equipment are secure against damage.
(2) Interaction with generator technical performance standards:
(a)
3.2
3.2.1
SRAS Generators are expected to adhere to the registered generator performance
standards (GPS) as agreed at the time of the generator connection. Relaxation may
be granted with respect to the following aspects:
(i)
S5.2.5.1: “compatibility levels” set out in Table 1 of Australian Standard AS/NZS
61000.3.7:2001 will be applied as opposed to “planning levels”
(ii)
S5.2.5.6: “compatibility levels” set out in Table 1 of Australian Standard AS/NZS
61000.3.7:2001 will be applied as opposed to “planning levels”
(iii)
S5.2.5.13: Slightly slower response characteristics, e.g. longer settling time,
may be permitted due to possible changes required to be applied to the settings
of excitation system control and associated limiters
Voltage criteria
Types of overvoltages
Standard insulation withstand voltage levels of equipment and system components are defined in
international standards such as IEC 60071-1, and IEEE Std 1313.1-1996. These standards are
similar and thus information in this guideline will refer to the IEC standard.
Voltages/overvoltages are classified and defined in IEC 60071-1 and are reproduced in Table 3. The
following types of overvoltages are relevant for black start studies:
 Continuous overvoltages
 Temporary overvoltages
 Slow-front overvoltages (sometimes referred to as transient overvoltages)
16/10/2014
Page 47 of 56
Power System Modelling and Study Methodology for SRAS Generator Evaluation
Table 3 Classes and shapes of overvoltages, standard voltage shapes and standard withstand
voltages from IEC 60071-1
The following criteria should be used when assessing acceptability of each type of overvoltage:
3.2.2
Continuous overvoltages
The supply voltage should remain between 90 percent and 110 percent of the normal voltage for
majority of the time during the four-hour restoration system. A voltage range between 85 percent
and 115 percent of the normal voltage may be accepted momentarily so long as it does not
compromise any other criteria defined for satisfactory operating state.
3.2.3
Temporary and transient overvoltages
Manufacturers of power system components test the equipment to withstand overvoltages defined
in the standards as follows:
 Standard short-duration power-frequency voltage: sinusoidal voltage with frequency
between 48 Hz and 62 Hz for the duration of 60 s. The voltage withstand level against
standard short-duration power-frequency voltage is called “Alternating Current Withstand
Voltage” (ACWV).
 Standard switching impulse: impulse voltage having a time to peak of 250 μs and a time
to half-value of 2 500 μs. The voltage withstand level against a standard switching impulse
is called “Switching Impulse Withstand Voltage” (SIWV).
 Standard lightning impulse: impulse voltage having a front time of 1.2 μs and a time to
half-value of 50 μs. The voltage withstand level against a standard lightning impulse is called
“Lightning Impulse Withstand Voltage” (LIWV).
16/10/2014
Page 48 of 56
Power System Modelling and Study Methodology for SRAS Generator Evaluation
Equipment with the “highest voltage of equipment “ (Um) in the range of 1 kV < Um ≤ 245 kV is tested
for the standard rated short duration power frequency withstand voltage and standard rated lightning
impulse withstand voltage according to Table 4.
Table 4 Standard insulation levels with highest voltage of equipment 1 kV < Um ≤ 245 kV.
Highest
Standard rated short-duration power frequency withstand Standard rated
voltage
for voltage (60 s) (ACWV)
lightning impulse
[kV] (RMS value)
equipment
withstand
(Um)
voltage (1.2/50
[kV](RMS value)
µs) (LIWV)
[kV]
(peak
value)
60
12
28
75
95
95
24
50
125
145
145
36
70
170
(185)
(450)
123
230
550
(185)
(450)
145
230
550
275
650
(275)
(650)
(325)
(750)
245
360
850
395
950
460
1050
NOTE: If values in brackets are considered insufficient to prove that the required
phase-to-phase withstand voltages are met, additional phase-to-phase withstand
voltage tests are needed.
Manufacturers are required to test with the standard switching impulse if equipment Um > 245 kV,
according to Table 5. All equipment should be tested with the standard lightning impulse.
Table 5 Standard insulation levels with highest voltage of equipment Um > 245 kV
Highest
Standard rated switching impulse withstand Standard rated
voltage
for voltage
lightning impulse
equipment
withstand voltageb
Longitudinal
Phase-toPhase-to(Um)
insulationa
earth
phase
[kV]
(RMS
(ratio to the [kV] (peak value)
value)
kV
(peak [kV] (peak phase-tovalue)
value)
earth peak
value)
1050
850
850
1.6
1175
1175
420
950
950
1.5
1300
1300
950
1050
1.5
1425
1175
950
950
1.7
1300
550
950
1050
1.6
1300
16/10/2014
Page 49 of 56
Power System Modelling and Study Methodology for SRAS Generator Evaluation
1425
950
1425
1175
1.5
1050
1550
Value of the impulse component of the relevant combined test while
the peak value of the power frequency component of opposite
polarity is
a
Um  2 / 3
b
These values apply as for phase-to-earth and phase-to-phase
insulation as well; for longitudinal insulation they apply as the
standard rated lightning impulse component of the combined
standard rated withstand voltage, while the peak value of the
power-frequency
component
of
opposite
polarity
is
0.7  U m  2 / 3
As observed in Table 4 and Table 5, different voltage withstand levels may be selected for some
types of overvoltages such as lightning impulse and for longitudinal insulation.
Note that for TOV studies, rms values are of interest. Conversion of instantaneous voltage
waveform to rms value is therefore necessary.
Note that both ACWV and LIWV specifies the phase-to-ground voltages. Care should be taken to
plot the phase-to-ground voltage in PSCAD case.
The peak phase-to-ground value of the nominal voltage should be used as base voltage when
converting from kV to pu:
U base 
3.3
2
Un
3
(49)
Frequency criteria
Frequency operating standards: Frequency standards for island systems as defined in the AEMC
reliability panel document ’’Application of Frequency Operating Standards During Periods of
Supply Scarcity’’ will apply during the restoration process until 40% of annual peak demand is
restored. A summary of the requirement is indicated in Table 6.
Table 6 Operating standards during periods of supply scarcity
16/10/2014
Page 50 of 56
Power System Modelling and Study Methodology for SRAS Generator Evaluation
3.4
Surge arrester withstand criteria
To ensure the integrity of surge arresters two criteria need to be assessed:

Energy dissipation across the surge arrester (this parameter is provided in the
manufacturers’ datasheet). As an example energy capability of [Sample manufacturer]
surge arresters is given in
Table 7.
Table 7 Energy capability for various [manufacturer] surge arresters
Energy
Line
Normal
Arrester
capability
discharge
application
Type
(2 impulses)
class
range (Um)
kJ/kV (Ur)
EXLIM R
2
5.0
≤ 170 kV
PEXLIM R
2
5.1
≤ 170 kV
EXLIM Q
3
7.8
170 - 420 kV
PEXLIM Q
3
7.8
170 - 420 kV
EXLIM P
4
10.8
362 - 550 kV
PEXLIM P
4
12
362 - 550 kV
HS PEXLIM P 4
10.5
362 - 550 kV
EXLIM T
5
15.4
420 - 800 kV
HS PEXLIM T 5
15.4
420 - 800 kV

Surge arrester 1s and 10s TOV withstand capability (these parameters are also provided in
the manufacturers’ datasheet). As an example TOV withstand capability of [manufacturer]
surge arrester is given in Figure 36.
Figure 36 TOV withstand curve for [manufacturer] surge arrester
16/10/2014
Page 51 of 56
Power System Modelling and Study Methodology for SRAS Generator Evaluation
Study methodology
4.1
Steady-state analysis
4.1.1
Study methodology
PSS/e load flow studies are conducted to:

Confirm appropriateness of initial operating point including required transformer tap settings

Establish correct initial conditions for dynamic analysis in terms of P,Q, and V

Confirm thermal and voltage limit are maintained within the operational limits for all steps of
the restoration plan
4.1.2
Success criteria

All voltages are within 90% to 110% of the nominal values (a safety margin of 5% is
allowed for any data inaccuracy)

Thermal limits are within 80% of the equipment rating (this is unlikely to be a problem)
4.2
Transient analysis
4.2.1
Transformer energisation
4.2.1.1 Study methodology

Apply correct transformer energisation methodology based on whether the winding to be
energised is star or delta connected (the residual flux needs to be replicated on the
transformer to be energised as well as any other transformer connected to the same
busbar).

Determine the most onerous voltage phase angle by performing multiple energisation
studies using statistical breaker model.

If the transformer has the soft-start capability1, ensure that the synchronous generator
excitation field voltage is ramped up gradually over a defined period, e.g.10 s, as specified
by the Generator.

Integrate model of the following generator relays into the PSCAD case:
o
Generator differential,
o
Unit transformer differential relay including inrush current restraint element (if
applicable),
o
Generator over-flux (V/Hz) relay.

Calculate positive, negative and zero-sequence driving point frequency-domain harmonic
impedances at both high and low side of the transformer to be energised.

Calculate individual harmonic voltage distortion up to 13th harmonic, and total harmonic
voltage distortion

Run the study for 10 s, and observe if any relay has operated.

Re-run the study for 30 s if any generator limiters is activated.
1
This feature applies to unit transformers only. With a soft start, first the connection between generator and
transformers is established. The generator is then started without excitation voltage. With the generator
running at no load full speed, the excitation current of the generator is gradually increased.
16/10/2014
Page 52 of 56
Power System Modelling and Study Methodology for SRAS Generator Evaluation
4.2.1.2 Success criteria

Maximum energy dissipation across surge arrester is within 80% of the arrester energy
capability.

1s and 10s TOV levels across the surge arrester are within 80% of the arrester TOV
capability which can be found in manufacturers’ datasheets.

Transformer TOV level is not greater than 80% of the respective overvoltage indicated in
Table 4 and Table 5.

Transformer differential relay blocks targeted harmonic frequencies.

Generator or unit transformer differential relays have not operated.

The maximum steady-state total harmonic distortion does not exceed 8%.

Generator over-excitation and over-flux relays are not activated.

All system voltages are maintained within 0.85-1.15 pu.

Island frequency is maintained within 49.5 Hz to 50.5 Hz.
4.2.2
Line energisation
4.2.2.1 Study methodology

Energise one line at a time (allow 5s if a second parallel line is to be energised).

Determine the most onerous voltage phase angle by performing multiple energisation
studies using statistical breaker model.

Integrate model of sending end generator under-excitation limiter models into the PSCAD
case.

Observe whether the under-excitation limiter is activated, and compare the reactive power
consumption of the generator against the reactive power capability chart provided by the
Generator.

Integrate model of the following sending end generator relays into the PSCAD case:
o
Loss of excitation,
o
Over-flux, and
o
Negative phase sequence.

Calculate positive, negative and zero-sequence driving point frequency-domain harmonic
impedances at both sending and receiving ends of the line.

Calculate individual harmonic voltage distortion up to 13th harmonic, and total harmonic
voltage distortion

Run the study for 10 s, and observe if any relay has operated.

Re-run the study for 30 s if any of the limiters are activated.

Pick-up loads if specified in the System Restart Procedure to maintain the system
frequency, and control any potential overvoltages.
4.2.2.2 Success criteria

Maximum energy dissipation across surge arrester is within 80% of the arrester energy
capability.

1s and 10s TOV levels across the surge arrester are within 80% of the arrester TOV
capability which can be found in manufacturers’ datasheets.
16/10/2014
Page 53 of 56
Power System Modelling and Study Methodology for SRAS Generator Evaluation

Transformer TOV level is not greater than 80% of the respective overvoltage indicated in
Table 4 and Table 5.

All generators have maintained their stability.

Sending end generator relays have not operated unless in the case of generator selfexcitation.

Maximum steady-state total harmonic distortion does not exceed 8%.

All system voltages are maintained within 0.85-1.15 pu.

Island frequency is maintained within 49.5 Hz to 50.5 Hz.
4.2.3
Motor starting
4.2.3.1 Study methodology

Determine the three largest auxiliary loads for each generating unit, and include them in the
model.

Determine the sequence by which the auxiliary loads need to be energised:
o
Assess whether to start the motors from largest to smallest, or vice versa.
o
The sequence must also accommodate the start-up requirements of the plant, which
may require certain motors to be started before others. This information should be
provided by the Generator.
o
To save the simulation time all auxiliaries of a generating units may be restarted
simultaneously so long as the impact is acceptable on the network (refer to success
criteria)

Energise the motors through the generating unit auxiliary transformer as shown in Figure
12.

Monitor motor terminal voltage, reactive power, electric torque, and slip

If motor restart successful, restart a second set of generator auxiliaries if it is required as
per System Restart Procedure. (Allow 5s when restarting the auxiliaries of the next
generating unit)

Restart additional motors if it is required as per System Restart Procedure.

Transfer the motors to the generator bus after successful motor restart such that the motors
are directly connected to the generator unit transformer.

Run the case for 120 s. Check the system frequency.

If the system frequency is outside the 49.5-50.5 Hz band, run the case for longer duration
and observe whether the frequency error declines.
4.2.3.2 Success criteria

All system voltages are maintained within 0.85-1.15 pu.

Motor voltage does not drop below 0.8 pu (as per IEEE Std 399-1997)

Island frequency is maintained within 49.5 Hz to 50.5 Hz.

No relays has operated.

All generators have reached new steady-state conditions.
16/10/2014
Page 54 of 56
Power System Modelling and Study Methodology for SRAS Generator Evaluation
4.2.4
Island synchronisation
4.2.4.1 Study methodology

Start and run each island separately with switching proceeding in each Island until the two
islands could be synchronised.

Ensure no more than one generator is running on isochronous mode in each island before
synchronisation occurs.

Apply synchrocheck relays at all appropriate locations.

Ensure that the isochronous control in one system is switched off immediately prior to or
immediately after synchronisation occurs. Generally the larger generator should remain in
isochronous control and the smaller generator should be switched to normal droop control.

Run the study for 30 s.
4.2.4.2 Success criteria

4.2.5
Synchronisation has successfully taken place.
Load rejection
4.2.5.1 Study methodology

This study is only conducted on busbars to which substantial amount of load is connected.

Integrate model of the following generator relays for all on-line generators into the PSCAD
case:
o
Loss of excitation,
o
Over-flux,
o
Negative phase sequence, and
o
Out-of-step relay

Trip all connected loads simultaneously on the given busbar

Run the case for 120 s. Check the system frequency.

If the system frequency is outside the 49.5-50.5 Hz band, run the case for longer duration
and observe whether the frequency error declines.
4.2.5.2 Assessment criteria

Maximum energy dissipation across surge arrester is within 80% of the arrester energy
capability.

1s and 10s TOV levels across the surge arrester are within 80% of the arrester TOV
capability which can be found in manufacturers’ datasheets.

Transformer TOV level is not greater than 80% of the respective overvoltage indicated in
Table 4 and Table 5.

All system voltages are maintained within 0.85-1.15 pu.

Island frequency is maintained within 49.5 Hz to 50.5 Hz.

No relays has operated.

All generators have reached new steady-state conditions.
16/10/2014
Page 55 of 56
Power System Modelling and Study Methodology for SRAS Generator Evaluation
Selected references
1. P. Kundur, Power System Stability and Control, McGraw-Hill, 1994.
2. CIGRE WG 33.02, Guidelines for representation of network elements when calculating
transients, CIGRE Brochure 39, 1990.
3. D. Povh and W. Schultz, “Analysis of overvoltages caused by transformer magnetizing
inrush current,” IEEE Trans. Power App. Syst., vol. PAS-97, no. 4, pp. 1355–1365,
Jul./Aug. 1982.
4. IEEE Power and Energy Society, Dynamic Models for Turbine-Governors in Power System
Studies, January 2013.
5. IEEE Working Group on Prime Mover and Energy Supply Models for System Dynamic
Performance Studies, ’’Dynamic Models for Fossil Fuelled steam units in power system
studies’’, IEEE Transactions on Power Systems, Vol.6, No. 2, May 1991.
6. Argone National Laboratory, Review of exisiting hydroelectric turbine-governor models,
2013.
7. V. Brandwajn, H. W. Dommel, and I. I. Dommel, “Matrix representation of three phase n
winding transformers for steady state and transient studies,” IEEE Trans. Power App. Syst.,
vol. PAS-101, no. 6, pp. 1369–1378, Jun. 1982.
8. Alstom, Protective Relays Application Guide. Stafford:2014.
9. CIGRE WG B51.7, “Relay software models for use in electromagnetic transient analysis
program’’, June 2006.
10. IEC, 60071-1 International Standard., Insulation co-ordination - Part 1: Definitions,
principles, rules, 2006.
11. IEEE., IEEE Standard for Insulation Coordination—Definitions, Principles, and Rules, 1996.
IEEE Std 1313.1-1996.
12. IEC., IEC 60071-2 Insulation co-ordination - Part 2: Application guide, 1996-12. IEC 600712.
13. IEEE., IEEE Standard for Metal-Oxide Surge Arresters for AC Power Circuits (> 1 kV), New
York : IEEE, 2006. C62.11-2005.
14. Applications of PSCAD/EMTDC 2008, Manitoba HVDC Res. Centre Inc., Winnipeg, MB,
Canada, 2008.
16/10/2014
Page 2 of 56