Guidelines Transient and Voltage Stability

10/09/2015
G-TVS

Table of contents
Table of contents............................................................................................................................ 2
1 INTRODUCTION..................................................................................................................... 4
1.1 Rough summary of AhT findings......................................................................................4
1.2 Purpose............................................................................................................................ 4
1.3 TSOs practice and guidelines for system representation (Jonathan & Dmitrij).................6
1.3.1 State of the art and best practices.............................................................................6
1.4 Proposal for common model with harmonized requirements for both transient and
voltage stability .......................................................................................................................... 8
1.4.1 Static model requirements as base for dynamic model.............................................8
1.5 Definitions...................................................................................................................... 10
2 Min and max Model requirements for influence area.............................................................13
2.1 RES................................................................................................................................ 13
2.2 HVDC............................................................................................................................. 15
2.3 Conventional generators................................................................................................16
2.4 Transformers..................................................................................................................17
2.4.1 Generation step-up transformer..............................................................................17
2.4.2 Transformer with on-load tap changer.....................................................................18
2.5 Loads............................................................................................................................. 18
2.6 Lowest voltage levels.....................................................................................................20
2.7 Other components, controls and automata.....................................................................22
3 Min and max Model requirements for remote dynamic model...............................................23
3.1 RES................................................................................................................................ 23
3.2 HVDC............................................................................................................................. 23
3.3 Conventional generators................................................................................................23
3.4 Transformers (HOLD).....................................................................................................24
3.5 Loads............................................................................................................................. 24
4 Methods................................................................................................................................ 25
4.1 Introduction.................................................................................................................... 25
4.2 Steps to be conducted for the sizing of the influence area in the context of transient and
short-term voltage stability........................................................................................................25
4.3 Steps to be conducted for the sizing of the influence area in the context of long-term
voltage stability......................................................................................................................... 28
4.4 Example of application...................................................................................................28
4.4.1 Determination of the size of the influence area based on static criteria...................29
4.4.2 Creation of the reference dynamic model................................................................29
4.4.3 Validation of the size of the influence area and of the minimum modeling
requirements of the influence and remote area.....................................................................31
5 Implementation issues........................................................................................................... 32

2
G-TVS

5.1 Proposal for approach before CGMES...........................................................................32

5.2 Target approach with CGMES.......................................................................................32
5.3 Model reduction guidelines.............................................................................................32
5.4 Model maintenance / data exchange triggering..............................................................33
5.5 Model exchange/ legal requirements..............................................................................35
6 Appendix............................................................................................................................... 36
6.1.1 Abbreviations and definitions..................................................................................36
6.1.2 Difference between transient, short term and long term voltage stability.................36
6.1.3 References.............................................................................................................. 39
6.1.4 Existing CGMES controllers in development and testing.........................................39

3
G-TVS

1 INTRODUCTION

1.1 Rough summary of AhT findings

This document shall cover the actions 1.1. and 1.2 which were described by the Cross committee
Adhoc team on dynamic studies. The actions contain two main items:

Task 1 - Guideline on modelling for TSOs with regard to transient and voltage stability.
 Provide best practice support for TSOs in modelling their own network.
 Provide support and guidelines for the adequate modelling of new technologies, including:
o Power Park Modules;
o HVDC links capabilities and configurations and
o Demand Side Response.
 Provide support and guidelines for the adequate representation of distribution networks
and its embedded generation and demand performance.
o Propose a common understanding on that subject in order to facilitate TSOs requests
when addressing DSOs on modelling data, as the adequate modelling of distribution
networks is considered a topic of increasing relevance.
 Technical capabilities that are required in network codes also need to be correctly
modelled and incorporated in the models. Recommendations for TSOs when requesting
simulation data as established in the Network Codes should also be provided.

Task 2 - Description of conditions for the exchange of sufficient representation of

neighbouring systems and on network reduction
 Define the conditions and requirements for the exchange of sufficient representation of
neighbouring systems. This should take into account the observability area concept1
and should use the ENTSO-E study model as the basis for the neighbouring system
dynamic model
 Provide guidance on key principles for consistent network reduction
 Propose how and when standard models could be used in order to facilitate having
comparable results and more efficient and practical collection of neighbouring system data

NOTE 1: the term used in this document shall be influence area to avoid confusion with the
observability term in state estimation.

1.2 Purpose

When considering classical stability phenomena, the target of dynamic studies, as can be
exemplified by Figure 1 below, is not only the assessment of the pre and post contingency
operating point but also the full transition period between two steady state operating points.

4
G-TVS

Figure 1: Example of the necessity to assess operating points and transition period in classical
stability phenomena.

The final report of the AhT on dynamic studies distinguishes between four possible phenomena of
interest and described the level of this studies. Transient and voltage stability studies will be
performed on TSO or Multi-TSO Level. Following table gives an overview of the different kind of
stability studies.

Level of studies
Stability
Coordination
Phenomena Multi
TSO SA*
TSO
Studies should involve all the TSOs of the
synchronous area, as the model has to
Frequency
represent the overall system inertia with X
stability
adequate representation of the frequency
sensitivity of the connected devices.
Studies should involve all the TSOs of the
Small Signal synchronous area, as the oscillations
X
Stability identification methodology requires a detailed
model for the whole system.
The phenomena are local or regional, so the
studies can be performed at TSO level.
Voltage
Nevertheless, neighbouring TSOs should be X X
Stability
represented if their reactive power sources
have a significant influence on the study area.
The phenomena are local or regional, so the
studies can be performed at TSO level.
Transient
Nevertheless, neighbouring TSOs system X X
Stability
should be represented to an extent with a
significant influence on the study area.

* SA = Synchronous Area
As for the transient and voltage stability phenomena only elements within a limited impedance-
radius area are of relevance the representation of grid regions can be of a different level of detail.
The inner circle requiring the highest model quality is called observability area. E34

5
G-TVS

Within the observability area “detailed” models depending on the study objective are
recommended to be used.
Outside the observability area the system is modelled to represent the mutual impact and
interaction of the areas with respect to each other and with respect to the phenomenon under
investigation (long distance effect). See Figure 2.

The extent of the observability areas will vary from one TSO zone to another, depending on the
topology and characteristics of TSO’s local network and neighbouring networks. For those
reasons both detailed models and generic models can be used in the same study and are needed
for achieving highest accuracy in the dynamic investigations. Determining the size of the
observability area is a relevant issue which requires an important degree of expertise.

1.3 TSOs practice and guidelines for system representation (Jonathan & Dmitrij)

1.3.1 State of the art and best practices

As an introduction to the subject, the state of the art of best practices used by TSOs of the
ENTSO-E area in terms of dynamic model for voltage and transient stability have been collected
and are described in this section Internal grid model

In the context of transient stability studies, the internal transmission grid model is represented in
details while the distribution grids are partially reduced. The reduction of the distribution grid is
dependent on the influence of the distribution connected generation and loads to the transient
stability phenomenon (e.g. penetration level of distributed generation, the stability of a particular
synchronous generator or group of generators, stability of particular loads or loads areas,). As an
example, the French and German dynamic models fully represents their internal grid down to the
220kV voltage level, Spanish and Serbian dynamic models go down to 110kV while Belgian
model goes down to the 36kV. The lower voltage levels of the internal grid are reduced.

The internal generators are modeled in details based on data collected from users. For the larger
generating units, standard model are rarely used. No standard parameters are used but each
generator has its own setting to match setting of the physical unit.

In the context of transient and short-term voltage stability, the model contains at least the step-up-
transformer, synchronous machine or converter model, AVR, PSS, GOV. Under and Over-

6
 G-TVS

excitation limiters are not always represented for system representation upto 10-20s after the
triggering event.

In the context of long term votlage stability, different best practices are to be highlighted. Some
TSO rely on quasi-steady state approaches and therefore different models compared to the one
used for the transient stability analyses while other TSOs are relies on the same tool as the one
used for transient stability analysis (with variable step sizes in order to reduce the computational
burden). Therefore, either the same model, supplemented by excitation limiters, automatic
transformer tap changers and automatic capacitor bank/inductor bank switching are respresented
or a simplified model with excitation limiters, automatic transformer tap changers and automatic
capacitor bank/inductor bank switching is used. Manual actions are not represented.

Wind and PV farms are aggregated and represented by a robust model based on manufacturer
information or aggregated by type of technology. For local studies of individual farm performance,
a model provided by the manufacturers is used. The step-up transformer is always explicitly
modelled.

HVDC links are represented using a model provided by the manufacturer or representing
sufficiently the performance of the HVDC-technology.

In the context of long term planning, future generators, yet to be build and connected to the grid,
are modeled using standard models with standard parameter depending on technology and
minimum grid code requirements.

In general, the load is aggregated. One possibility is an 'IZ' model with frequency depend
parameter and frequency damping (e.g. Model library IEEL), continuous time evolving model
variating from constant impedance model during large transients to constant power model when
system is stabilized. Other models representing step-down tap changer, transmission line and a
share of induction and constant impedance load may also be used for dedicated studies. In most
of the cases models used to represent industrial load and distribution grids are different.

There are different slow dynamics that might be not represented, for instance, transformer tap
changers and capacitor and inductor bank switching.

In the context of transient stability studies, external grids are often split in two parts,
 one physical grid model, where lines, transformers, load and generators correspond to real
elements. This part of the grid will be referred to as “influence area” in the context of this report.
 one remote grid model to correctly represent inertia and equivalent country load and generation
and short-circuit power. This part of the grid will be referred to as “remote area” in the context of
this report.

It needs to be noted that the size of the influence area, depends on the mathematical accuracy of
the remote grid model (the more precise is this remote grid model, the smaller could be the
influence area). This choice also influence the computational burden of the dynamic studies (the
higher is the computing effort, the smaller must be the total represented model). Best practices
ranges from:
 a large influence area (typically a large share of the synchronous area) with detailed EHV-Grid
and connected generators and a very simple model of the remote grid (typically a thevenin
equivalent to match steady state load flows)

7
 G-TVS

 or a limited influence area (typically the whole or a share of the neighbouring TSOs EHV model)
and a elaborated remote area model obtained by application of dynamic network reduction
techniques such as coherent model aggregation and regulator parameter tuning for transient
voltage and frequency curve fitting.

1.4 Proposal for common model with harmonized requirements for both transient and voltage stability
This document is addressing transient stability, short term and long term voltage stability. For
different phenomena the relevant model may be different (more or less complex). Transient
stability and short term voltage stability are tightly related and require the same models except for
the load model which is more critical for voltage stability than rotor angle stability.

For short term voltage and transient stability the machines, FACTS, HVDC and load models
should describe the phenomena occurring during a time frame up to 20s after the fault clearing.
Phenomena or controls with a response time of ms should not be modelled.

For long term voltage stability, the models should describe the phenomena occurring at least 10
minutes after the outage. The models do not need to represent phenomena or controls with a
response time lower than 1 second.

In the rest of the document, two types of model requirements will be described. The “minimal
requirements” are the characteristics that must be included in the models exchanged in order to
guarantee that the phenomena which are studied are correctly simulated. If such features are not
modelled the simulations could miss or over estimate the studied phenomena and lead to a false
assessment of the grid security or planning quality criteria.

The “maximum requirements” define the features and characteristics of the models that should
not be exchanged for they are not necessary to the relevance of the simulations and are a burden
to the data process and the simulation tools. For instance, the detailed modelling of IGBT
switching in a HVDC converter station is not necessary for the types of simulations addressed in
this document nor the representation of the full distribution network.

In order to be an adequate base for building up a dynamic model, static models should contain
enough data to be able to at least simulate three phase short-circuits. Moreover, a minimum
quality of the steady state data (generators/HVDC/FACTS power outputs and bus voltages) is
required to properly initialize the dynamic models..

1.4.1 Static model requirements as base for dynamic model

As a consequence of the differences between transient, short and long term voltage stability, the
model requirements are different also as described below. Nevertheless, the model for short term
voltage stability and transient stability are identical except for the load model which is of greater
importance for voltage instability. In the rest of the document, the requirements will be the same
for both transient and short term voltage stability.
The models required for long term voltage stability are different in two ways : it requires less data
for the short term dynamics but demands more data for the long term behaviour of the load

8
G-TVS

(distribution tap changers), generators (excitation limitations) and the grid (transformer tap
changer dynamics, automatic shunt capacitor bank switching, secondary voltage control…).
Most of the requirements for short term dynamics match the needs of long term voltage stability
studies, however there are specific requirements to be considered as an addition for the short
term model. only when it adds an additional requirements to the short term model.

Large disturbance
Rotor angle
stability and Large
long term Voltage stability
disturbance short
term Voltage
stability
Voltage control steady-state
Synchronous Voltage control,
behaviour (droop, step-up transformer
Generator model governor control,
compound), overexcitation limiters,
Power Plant Model PSS, turbine model
other limits (thermal
On-load tap changers dynamic,
Load model Induction motors voltage dependency of distribution
load
HVDC Dynamic model steady state model
Switchable shunts, Generators/HVDC
Automata and slow undervoltage protection, phase shifter
No need
protections tap changers. Automatic Generation
Control (AGC)
FACTS High contribution
High contribution (steady state model)
(SVC/TSC/STATCOM…) (dynamic model)

9
G-TVS

1.5 Definitions

In order to conduct meaningful voltage and transient stability studies in a realistic time, it is
proposed to divide the model in several areas

 The “study area” is the part of the power system that is under investigation and is modeled
at least with the same level of detail as the influence area. This is typically the region of a
TSO in which the stability of the generating units or the voltage stability is to be analyzed
but it should also include the triggering event such as a short-circuit or a load increase.
The study area is always located within the responsibility of the TSO(s) conducting the
study.

 The “influence area” is the part of the power system in which the grid structure and the
dynamic behavior of the generating units and of the loads significantly influence the
results on the study area. Oscillations or redistributions of the active and reactive power
flows within the study area and the influence area should be precisely represented. The
size of the “influence area” depends on the aim of the study (transient stability study,
short-term or long term voltage stability). The influence area typically covers a part of the
system under the responsibility of the TSO(s) conducting the study and a part of the
system under the responsibility of other neighbouring TSOs.

 The “remote area” is the part of the power system which does not influence significantly
the quality of the result on the study area for transient stability and voltage stability
analyses. It however, influence quality of the results in the “influence area” but the
cascading influence on the results in the study area remains negligible. The main
objective of modelling this area in a simplified way is to reach numerically robust
simulations while reducing the computing time as much as possible.

The definition of study area, influence area and external area are technical electrical zones, which
are independent of the area of responsibility of the different TSO. However, it is of common sense
that the study area is located within the area of responsibility of the party conducting a study
(which can typically be one TSO or adjacent TSOs working together)

10
G-TVS

It needs to be recalled that this document defines minimum and maximum requirements for the
dynamic data exchanges to be received by the party conducting a study (in the study area) from
the external parties. The party conducting the study is typically either one TSO or adjacent TSOs
working together on a study and external parties are typically neighbouring TSOs from the party
conducting the study.

This documents also aims are providing guidelines/recommendation for minimum requirement of
the study and influence area located within the area of responsibility of the parties conducting the
study.

In the context of Spain, the influence area is the network of the neighboring countries (France,
Portugal and Morocco). With this approach, the results are reliable for both transient stability and
distribution of power flow through the lines interconnecting neighbouring countries. It is
recommended that the size of the influence area should not disregard the dominant dynamics
provided by relevant generators and their control systems for maintain voltage and frequency
stable. The remote area would be the rest of interconnected systems, which is represented by
means of large loads and generators with equivalent inertia and dynamic standard models in
order to correctly represent power flows in the influence area and the frequency behavior of
Continental Europe.

In the context of Serbia, for rotor angle stability of single generating unit (study area is a
substation), the influence area consists of the whole Serbian power system with additional 400 kV
and 220 kV neighboring power systems. Within the whole influence area (internal Serbia and
external), the same level detail is used. For transient voltage stability studies, the influence area
includes 400 kV and 220 kV network of the 11 countries of south-east Europe. The study area
(Serbia) is modelled down to 110 kV and in some parts down to 35 kV.

In the context of Germany, the influence area for rotor angle stability in large-scale studies
contains whole network of Germany. The network in Germany is modelled as follows: Lines of
voltage level 380 kV and 220 kV are modelled. All coupling transformers which connect the EHV
(380/220 kV) network with the 110 kV network are modelled while the 110 kV network is reduced.
The conventional generating units above 100 MW are considered explicitly with a voltage
controller and a governor.
Wind infeed is aggregated at the 110-kV level and modelled with a dynamic model. PV generation
is modelled as fixed P-Q-infeed.
The influenced area outside of Germany and also the remote area are modelled by using the
Dynamic Study Model (DSM) which was developed by the System Protection and Dynamics
(SPD) group under the ENTSO-E. Wind and PV generation is modelled as fixed P-Q-infeed. The
conventional generation outside of Germany is modelled with standard machine data and
standard controller models.

In the context of France, for rotor angle stability, the influence area is usually limited to the
generation sites surrounding the site where the short-circuit is simulated, but in some
circumstances (notably high power flows across large areas) the influence of a short circuit may
reach remote zones (voltages dips or power swings on the other part of the French system). Then
for transient studies, the whole French EHV grid (225 and 400kV levels) is modeled plus a part of
neighboring TSOs. For long term voltage stability, about a third up to half of the French system is
detailed (EHV + 63 and 90kV) the rest being modeled only at the EHV level. The detailed area is
defined based on a priori administrative boundaries and also reach a number of 63 or 90 kV
substation beyond that (ex. Eastern France + 10 HV nodes beyond this zone). Such large areas

11
G-TVS

are used because the voltage collapse may involve a large part of the system as the generators
reach their excitation limit further and further from the initiating contingency in a kind of cascading
effect (some generators may even trip by under-voltage protection).

In the context of Belgium, the study area is always the full power system of Belgium down to the
36kV grid. The size of the influence area is defined every few years on a typical snapshot for
which full dynamic model of Continental Europe is available (e.g. EWIS study, UPS-IPS study,
…). The influence area is designed to hold elements from the external power system which have
a significant impact on boundary internal elements. This impact is computed at each instant of a
transient simulation by the product of, the voltage influence of the border node (Xnodes) or transit
influence on the interconnection lines, times the contribution of the element (variation of the active
or reactive power, variation of the voltage of generator output) of the influence area to this
influence. In the context of Belgium, the latest computation of this influence area holds the
Nederlands 380kV and 150kV grid, west of Germany 380kV, 220kV and 110kV, Luxembourg and
the north of France (down to Paris) 400kV and 225kV. The same influence area is used for both
short-term (angular and voltage stability) and for long-term voltage stability even if it is believed
that for transient angular stability, this influence area could be slightly reduced.
The remote area model is obtained by reducing the external CE power system by static
REI(Radial Equivalent Independent) reduction, followed by identification of coherent generators
using waveform correlation analysis and aggregation of these coherent generators and then
tuning of parameters of controller of the equivalent generator.

The current TSO best practices will be used, in the next section to propose a method that can be
used to size the influence area for a given study.

12
 G-TVS

2 Min and max Model requirements for influence area

2.1 RES
-Model detail; representation of the 110 kV bus bars. Representation of all the RES sources
connected directly to that voltage level. Connected RES at 110 kV should be represented as
follows

Figure 2 WP Aggregated representation

13
G-TVS

Figure 3 PV aggregated generator representation

- Aggregation level, aggregated generation sources by technology using with a maximum

aggregation level of 50-500 MW or maximum capacity of the node where the aggregated
generation is connected (distribution or transmission) as long as it is aggregated by technology
type.
-Model content, standard models based in IEC 61400-27 should be used. The model content
should cover the requirements indicated in the NC –RfG for voltage support, reactive support,
fault right through capability and frequency response.
The model content for WP should contain Aerodynamic model. Mechanical model, Generator
system(1), control system, Protection system.
The model for PV should contain, PV array model, converter model, reactive power or voltage
regulator model.
Although for the purpose of the transient and voltage stability analysis the outputs from the
aerodynamic model and the PV array model are kept constant.

NOTE(1): As defined in IEC 61400-27, includes the converter model for type 3, type 4 classes.

- Model initialization
When adding non-conventional generation controllers to an existing model, it is recommended to
verify the interaction of the conventional generation controllers with the added controllers. A 100
ms non-disturbance simulation can be performed to this effect.

The following parameters should be measured to check initialization:

• Conventional generators; Active and Reactive Power Output, terminal Voltage, machine
rotor angle
• Non conventional generators; Active and Reactive Power Output, terminal Voltage,
system frequency deviation,

After initialization a single disturbance event should be performed. Typically a 3-phase short
circuit at the PCC, without tripping. During these tests the following should be monitored:
• Non conventional generators response; Active and Reactive Power Output, terminal
Voltage, system frequency and voltage deviation at PCC.

More detailed information for modelling RES, please use the document G-RES.

14
G-TVS

2.2 HVDC

This chapter shortly describes the requirements for HVDC system models which are necessary
for dynamic studies in an influenced area of a synchronous system. In more specific terms, this
means that three parts of HVDC systems are affected:
 embedded HVDC system for power transportation as a Backbone system to the under laid
AC system (1)
 HVDC systems connecting two synchronous areas (2) including B2B
 HVDC connected generation, i.e. Offshore windfarms (3)

For further and more detailed information refer to [G-HVDC]. All models should fulfill the relevant
aspects of NC HVDC.

Depending on the subject matter of HVDC systems, different tasks should primarily be fulfilled by
the HVDC and therefore different aspects of model requirements have to be considered. While
the main intention of (1) is for power transportation within an electricity system and therefore
enhancing system stability also by providing additional system services e.g. reactive power
support, (2) connects two synchronous systems by DC only. As to type (3), where only generation
is connected, one end of the HVDC system is in the influenced area located and therefore
considered only.
Focusing on transient (rotor angle stability) and voltage stability of an AC system the main impact
of HVDC transmission (1) or unique converter stations (2), (3) has to do with small signal stability
(long term AC voltage decrease) and large signal stability (voltage stability issues, generator
stability) in the connected AC system due to the occurrence of disturbances in the AC power
system itself or in the HVDC system. Additionally and not covered in this document, especially (2)
and (3) can have impacts to frequency stability of a synchronous AC system.
The effect on transient and voltage stability of an AC power system differs in the system type of
HVDC as mentioned above, technology of the converter, the fault location, the fault clearing time,
the faulted system AC or DC and the fault type.

Therefore additionally to the above-mentioned points the minimum requirements for embedded
HVDC transmission systems (1) or unique converter stations (2), (3) for influenced areas are
shortly listed:
 type of HVDC system as mentioned above
 number and main technical parameters of converter transformers.
 HVDC technology (LCC, VSC)
 number and technology of converters (half-bridge, full bridge,…)
 structure of the transmission line (monopole, bipole)
 network structure if existing
 main additional equipment if existing (reactive supply, DC breakers,…)
 minimum requirements to PCC if necessary, i.e. Sk`` for LCC
For more information, refer to G-HVDC.

Additionally to the main parameters of an entire HVDC system, the transmission line and
additional equipment HVDC models and/or converter models should cover the main aspects of
behaviour in time ranges transient and voltage stability studies are normally done.

15
G-TVS

The minimum requirements of a dynamic HVDC model for transient and voltage stability studies
respectively converter model are listed below and differ dependent to the already above-
mentioned aspects of the converter technology or system configurations:
 P/Q-diagram or
 maximum active and reactive current in steady state
 maximum change rates of P and Q in steady state
 voltage ranges in normal operation AC and DC
 reactive power control mode (see also § 20 of NC HVDC)
 frequency ranges the model can be used if necessary (see also chap.2 of NC HVDC)
 FRT capability
 maximum fault clearance time of DC faults if possible
 maximum system recovery parameters
 additional reactive power supply (characteristics, timing and accuracy)
 maximum delay within adjusting a new operation point (i.e. change of reactive power control
mode or STATCOM mode)
 additional relevant control functions (e.g. POD as it may affect the transient stability and
short-term voltage stability)
 main converter protection settings (overvoltage, undervoltage,…)

For more and detailed information, refer to G-HVDC.

Additionally to the listed requirements, the model can contain all specifications mentioned in the
NC HVDC.

The detailed power electronics models can be neglected for the purpose of the effect on transient
stability. However the behaviour of a HVDC system (1-3) during and after AC and/or DC faults is
mainly dependent on the converter technology, on the converter and system configuration and
the implemented control functions of the power electronics. Therefore, at least four different
HVDC models have to be considered in order to represent the full range of possible
configurations ranging from LCC to various VSC technologies.

The adjustment of the models of active elements (HVDC) affects the results of considered studies
mainly and has to be clearly and exactly made before starting analysis. Otherwise, no
authoritative results can be accomplished.
Following all this, the proposal is to use reduced (RMS), vendor independent models, but to
ensure a high compliance to all the requirements indicated before a reduced model will neglect
the effect of the DC dynamics, the power electronics, but will include all necessary behaviour of
HVDC systems and/or unique converter stations for transient and voltage stability studies.

2.3 Conventional generators

Static model requirements to ensure sufficient data quality for dynamic simulations:
 Generators are modelled at their terminal voltage level.
 Generator step-up transformers should be modelled explicitly.
 If auxiliary consumption exists at generator terminals it should be at least modelled as
constant PQ load (4 percent of nominal/maximum active power).

16
G-TVS

 In case of more than one generator connected to a particular node, no aggregation of

generation should be made. Generating units greater than 100 MVA should be modelled
explicitly.
 Generator active power limits: The generation capability limits specified for generators
(Pmin and Pmax) should represent realistic unit output capability at the node on which the
generator is modelled. The generator maximum active power output (Pmax) should be
consistent with the auxiliary consumption. Pmax should always be greater than or equal to
Pmin.
 Generator reactive power limits: The reactive power limits specified for generators (Qmin
and Qmax) should represent realistic unit output capability at the node on which the
generator is modelled. Qmax should always be greater than or equal to Qmin.
 Small generators (e.g. 10 MVA) have limited reactive capability and cannot effectively
regulate transmission node voltage. Consideration should be given to modelling them as
non-regulating by specifying equal values for Qmin and Qmax.

dynamic model requirements:

Conventional generators have a distinct interaction to the entire network. Due to this it is very
important to use exact generator models as far as possible. In addition to the generator model, it
is necessary to model all corresponding controllers. This includes the voltage controller with the
PSS function and both the over and under excitation limiter. Furthermore, the speed controller is
required. According to the scope of investigation, the requirements to the generator and controller
models are different. In order to study local effects in detail e.g. connect a power plant to the grid,
it is necessary to model the generator and the above-mentioned controllers with models given by
the manufacturer. In other cases, it could be sufficient to use standard models and parameters.
The requirements on the generators and controllers are described below:

 Recommendation to use Park's generator model with parameters provided by the

manufacturer or a model library with the set of owner parameters that characterize the
dynamic behaviour of the manufacturer model.
 Attention should be paid to the acceleration time constant T A (TA=2H). It is necessary to
use TA of the whole shaft system given by the manufacturer (not only TA of the generator).
 Structure and parameters of controllers mentioned above have to be modelled by
manufacturer information or by a model library with the set of owner parameters that
characterize the dynamic behaviour of the manufacturer model.
 Protections should be modeled if they are described for internal studies and necessary for
the investigation.

2.4 Transformers

2.4.1 Generation step-up transformer

Unit transformer models shall represent the tap changer characteristics as number of steps,
change in voltage and uk per step and neutral step information even if no automatic voltage
regulation is applied in normal operation.
If the transformer is operates in voltage regulation mode further information of the time-
domain tap changer action is recommended to be given to assess interaction with
excitation limiter. If no relevant interaction is expected the time domain performance can be
represented by a simplified controller action to limit model and simulation effort
For investigation of transient stability phenomena tap changer dime domain action can be
neglected.

17
G-TVS

2.4.2 Transformer with on-load tap changer

Transformers connecting grids of different voltage levels from EHV to HV level shall
represent the tap changer characteristics as number of steps, change in voltage and uk per
step and neutral step information even if no automatic voltage regulation is applied in
normal operation.
If the transformer is operated in voltage regulation mode further information of the time-
domain tap changer action is not required. To assess voltage stability phenomena a
simplified approach can be chosen to keep modeling effort and simulation time in an
acceptable range. If tap changer action is e.g. 1 step per 5 seconds, the load response is
represented sufficiently for long term voltage stability studies.
For investigation of transient stability phenomena, tap changer dime domain action can be
neglected.

2.5 Loads
This chapter describes the requirements for load models in the influenced area for dynamic
studies of the network. Some recommendations are provided in order to address the dynamic
phenomena in a more accurate way.

Minimum requirements:

Embedded generation (RES or other types of generation) should be separated from the load. For
short term phenomena the load may be aggregated at the EHV level (150 kV and above), the
dependency of the active and reactive power to the voltage must be modeled.
For long term phenomena, the dynamics of all modelled tap changers should be modelled (if it
exists see chapter 2.4) as well as the active and reactive power dependency to the voltage.

The dependency of the load to voltage may be expressed at least by classic static laws (without
time-dependent behavior). For example using the static exponential model or the ZIP model. Pn,
Q n, and U n are the steady state values of Pload Qload and U at the load connection node.

Static exponential model:

k pu
U
Pload =Pn ( )
Un

k qu
U
Qload =Qn ( )
Un

Polynomial model (ZIP model) (static model):

3
U 2 U
Pload =Pn p1
[( ) ( ) ]
Un
+ p2
Un
+ p 3 , ∑ p i=1
3
i =1

U 2 U
[ ( ) ( ) ]∑
Qload =Qn q 1
Un
+ q2
Un
+q 3 , qi=1
i =1

18
G-TVS

A large variety of load models exist in the literature but the models which are selected for
exchange must be compatible with the specifications of CGMES.

Maximum requirements:

See chapter 2.6 for requirements on lower voltage levels network modeling.
Load management protection e.g. load scheduling or tap changer blocking should not be
exchanged.

Recommendations:

Fundamentally you can distinguish between Static Load Models and Dynamic Load Models.

Static Load model is a model that expresses the active and reactive power at any instant of time
as functions of the bus voltage magnitude and frequency at the same instant. Static load models
are used both for essentially static load components (e.g. resistive and lighting load) and as an
approximation for dynamic load components (e.g. motor-driven loads). The exponential and
polynomial models are static model.

Dynamic load model is a model the expresses the active and reactive power at any instant of time
as functions of the voltage magnitude and frequency at past instants of time and, usually,
including the present instant. Difference or differential equations can be used to represent such
models.

Dynamic Studies have the scope of the time-dependent behavior from one static operation point
to another static operation point. Such studies require at least static load model or for more
accurate results a dynamic load model which is able to represent the time-dependent behavior
direct after a disturbance in the system.

There are a lot of different dynamic load models available. A wide collection of dynamic load
models is described in a Cigre publication named “Modelling and Aggregation of Loads in Flexible
Power Networks” [2.5.1]. Amongst others the following models are described: Exponential
Dynamic Model, Dynamic Induction Motor Model, Transfer Function Induction Motor Model and
the Composite Model.
The Exponential Dynamic Load Model provide reliable results for long term voltage stability
studies and for disturbances which cause a voltage drop at the load buses compare to a voltage
drop caused by a loaded tap changing transformer [2.5.1]. For this reason the Exponential
Dynamic Load Model is suitable to be used in remote areas or in areas with a small amount of
rotating loads.

The Dynamic Induction Motor Model is a fifth order three phase induction motor model and should
be only used for modelling large induction motors in very local studies.

The Transfer Function Load Model describes both the static load and the behavior of the
induction motor. It is possible to describe the load with a first, second and third order transfer
function.

First order transfer function:

19
G-TVS

k pf +T pf ∙ s k pu +T pu ∙ s
∆ P ( s )= ∆ f ( s) + ∆ U ( s)
1+T 1 ∙ s 1+T 1 ∙ s

k qf + T qf ∙ s k +T ∙ s
∆ Q ( s )= ∆ f ( s )+ qu qu ∆ U ( s )
1+T 1 ∙ s 1+T 1 ∙ s

The first order transfer function can be used for load models for dynamic studies in the influenced
area for more accurate results. However, each TSO can decide to use a more detailed load
model if it is required. For example in areas with a higher portion of induction motors the second
or third order transfer functions, which deliver more accurate behavior during the transients, can
be used [2.5.1]. Supposing the parameters for the second or third order transfer function are
known. Another option is to use a detailed Dynamic Induction Motor Model. This could be
necessary in areas with a very high portion of induction motors and for very local detailed studies.

2.6 Lowest voltage levels

 The level of power system representation should, in general, respect the following rules:
o All elements connected at 220 kV and above must be modelled explicitly.
o sufficient representation of the voltage levels between 30kV and 150 kV is necessary if this
part of the network plays a significant influence on voltage stability – significant amount of
distributed generation, DSR etc.
o Branches and substations of the network under the 30 kV voltage level should not be
represented in detail. Loads can be aggregated at the closest 30kV or above node.

The lowest voltage level required for modeling the network of influence in transient stability
studies, will allow to show the dynamic response of the power system after a large disturbance
with accuracy. This will provide reliable results in the oscillatory condition and distribution of
power flow through the lines interconnecting neighboring countries. Reduction of the system
“size” should not disregard the dominant dynamics provided by relevant generators and their
control systems for maintain voltage and frequency stable.

On the other hand, it is important to consider the displacement of conventional synchronous

generators by distributed generation (wind, solar, cogeneration, etc.) that is currently experienced
in the different electrical systems in ENTSOE and that will increase in the near future. In relation
to this, the network connection of this generation becomes very important, above all to take into
consideration the impact on the power system stability of a massive disconnection of such type of
distributed generation (e.g. 50.2 Hz issue).

It is important to be consistent with European Network Codes (NC). For instance, NC on

Requirements for generators indicates the need for simulation models for type C generators,
which are connected below 110 kV. The TSO is entitled to order models. The purpose of the
models is to be used in studies for continuous evaluation in system planning and operation.

Considering all these aspects, the following will be taken in account:

 The necessary lowest voltage level: 220 kV (Transient and Voltage stability)
 The recommended lowest voltage level: 110 kV (Transient stability)

20
G-TVS

 The recommended lowest voltage level: 30 kV (Voltage stability). Sufficient representation

is necessary if this part of the network plays a significant influence on voltage stability

Advantages and disadvantages concerning the choice of the lowest voltage levels above
indicated could be:

Advantages
 Reduce the amount of information to be exchanged by TSOs
 Reduce simulation times
 Avoid confidentiality issues when exchanging information about generators, above all the
dynamic data

Disadvantages

 Changes in the power system model dynamic behavior

 Get an unrealistic power flow through the interconnection lines
 Fake the equivalent inertia when distributed generation is being aggregated or simplified
as a negative load
 Increase / decrease the loss of generation in fault conditions.
 Miss oscillation modes between generators of different areas

Therefore, it is very important that the TSO has sufficient information of the system with the
purpose to make the most reliable studies. For this reason, the elements connected at 220 kV
and above must be modeled explicitly including at least:

 Substations’ regular topology and other relevant data by voltage level

 Transmission lines
 Transformer connecting the DSOs
 Phase-shifting transformer
 High voltage DC lines
 Reactors, capacitors and static VAR compensator
 Generators and loads

The network of influence with voltage below 220 kV may be represented with an equivalent
network according to the standards/methodologies defined by the relevant TSO. Each relevant
Network Operator and relevant TSO shall preserve the confidentiality of the information and
data submitted to them in compliance with the current national regulation.
An equivalent network is the result of a process that converts an initial network (with many
buses/elements) into a reduced order network that contains only the boundary buses and the
most relevant initial buses.
The equivalent network will include the demand and distributed generation aggregated
geographically in the relevant buses of connection (30 kV voltage level or above). The process
to aggregate both demand and distributed generation, it may be added technologically
(generators) and by sector if possible (demand: commercial, industrial and residential).
Network equivalents might be needed under two circumstances:
 To allow representing larger areas of major interconnected systems for stability studies
 To improve computational speed in simulations by reducing the size of scenarios.

An example of reduction of systems using an equivalent network could be the next:

21
G-TVS

Figure 1. Reduction of systems

2.7 Other components, controls and automata

 Parameters for grid elements must be provided, regardless of whether or not the elements
are under consideration, out of service or undergoing maintenance.
 Coordination of regulating devices: Multiple regulating devices (generators, switched
shunt devices, tap changers, etc.) controlling the node voltage at a single node, or multiple
nodes connected by zero impedance lines as described above, should have their
scheduled voltage and voltage control ranges coordinated. In addition, regulated node
voltage schedules should be coordinated with the schedules of adjacent nodes.
 Over and under voltage regulation: Regulation of voltage schedules (reference voltages)
exceeding the range of ±10 % of the normal operating voltage of the regulated node
should be avoided.
 Fixed shunts: All fixed shunt elements at nodes modelled in the load flow should be
modelled explicitly (not as loads or included with load).
 Switched shunts: Switched shunt elements at nodes modelled in the load flow should be
modelled explicitly.
 FACTS devices: They should be represented in load flows as continuously variable
switched shunts rather than as generators. If available FACTS should be modelled
explicitly, dynamic RMS model (manufacturer or model library type that represents the
static and dynamic behaviour of the manufacturer model).

22
G-TVS

3 Min and max Model requirements for remote dynamic model

3.1 RES

RES models shall be modelled for each RES Type separately. At least wind and PV have to be
distinguished. It is not necessary to model the voltage dependency for RES models in the remote
area due to the negligible voltage change caused by a fault in the influenced area.

RES PV Modelling

It is sufficient to aggregate the PV infeed at EHV – Level or if modelled at the HV-Level. To

achieve acceptable model behaviour it is sufficient to use a constant P and Q infeed. The voltage
dependency can be neglected due to the distance to the influenced area and due to the low
voltage level where the PV panels usually installed.

RES Wind Modelling

It is sufficient to aggregate the wind infeed at EHV – Level or if modelled at the HV-Level. To
achieve acceptable model behaviour it is sufficient to use a constant P and Q infeed. The
frequency dependency can be relevant if the portion of DFIG wind turbine is high. The voltage
dependency can be neglected.

3.2 HVDC
Proposal to use a steady state model, with the parameters for LCC and VSC technologies.
The HVDC lines can be modelled as a PQ or PV node.

3.3 Conventional generators

Conventional generators have a distinct interaction to the entire network even in the remote area.
Due to this it is very important to use generator models with sufficient representation of the
dynamic behavior. In addition to the generator model, it is necessary to model all corresponding
controllers. This includes the voltage controller with the PSS function and both the over and under
excitation limiter. Furthermore, the speed controller is required. According to the scope of
investigation, the requirements to the generator and controller models are different. To represent
generators in the remote area it is sufficient to use standard models and parameters.

Generator and controller models:

 Recommendation to use Park's generator model with standard parameters for each fuel
type.
 Attention should be paid to the acceleration time constant T A. It is necessary to use T A of
the whole shaft system (not only TA of the generator).
 Structure and parameters of controllers mentioned above can be modelled with standard
controllers and parameters for each fuel type respectively for each excitation type.

23
G-TVS

3.4 Transformers (HOLD)

3.5 Loads

Transient and short term voltage stability:

For the remote area the load can be modelled by a constant power P and Q or constant
impedance.

Long term voltage stability:

For long term voltage stability the use of the static exponential load model can be meaningful (see
chapter 2.5 Loads).

Lowest voltage levels

The lowest voltage level required for modeling the remote area in transient stability and voltage
stability studies, will allow to show the dynamic response of the power system after a large
disturbance in the study area. This will provide reliable results in the oscillatory condition and the
frequency transient.

The remote area will be represented with an equivalent defined by the relevant TSO. The
elements that must be modeled explicitly are large loads and generators. This elements will be
connected in the most relevant boundary buses between both influence area and remote area.
The equivalents generators should take an inertia, reactance and generation according to the
study to provide results reliable in the inter-area oscillation behavior.

Considering all these aspects, the following will be taken in account to model the equivalents
elements:
 The necessary lowest voltage level: 400 kV and 220 kV (Transient and Voltage stability)

This equivalent is needed under the next circumstances:

 To allow representing larger areas of major interconnected systems for stability studies.
 To improve computational speed in simulations by reducing the size of scenarios.
 To avoid confidentiality issues when exchanging information.

24
G-TVS

4 Methods

4.1 Introduction

In order to make this method to be usable in the near future it is proposed that the determination
of the influence area is done by means of steady state computation. It is well understood that this
approach is not scientifically rigorous but static criteria based on conservative expert-based rules
could be used to define this influence area. The main objective of a static approach for the sizing
of the influence area is to allow CGMES models to be used for the steady state computation of
the criteria.

As short-term voltage stability and transient stability are greatly impacted by the support of
generating units in terms of short-circuit, it is proposed that every source of short-circuit current
having a significant contribution on the short-circuit current at the border of the study area is
modelled with sufficient detailed and not reduced. Therefore, it should be included in the influence
area.

In the context of short-term voltage stability, the influence area should be extended until, for any
considered incident, generator located at the border between the influence and remote area are
in voltage control mode with sufficient reactive power margin (PV nodes).

In the context of transient stability, the influence area has to cover the relevant transit corridors.
Bulk generation and load centers must be modelled with sufficient detail to investigate on
transient stability issues.

In the context of long-term voltage stability, the influence area should be extended until, for any
considered incident, transformer tap changer are not significantly observing voltage change at
their control.

It is important to recall that, depending on the power system context (e.g. generation and load
patterns, ...) and grid development stage (e.g. different grid models), the size of the influence area
for a type of study (i.e. Transient, short-term or long-term voltage stability) at a given location (i.e.
for a given study area) will be different. However, it is expected that the influence area belonging
to an external grid and to be considered by a given TSO for its internal studies will only slightly
evolve from year to year.

4.2 Steps to be conducted for the sizing of the influence area in the context of transient and short-term
voltage stability
As an illustration, the determination of the influence area can be reached by conducting the
following static computation.

Type of study Event Monitored criteria Threshold to be part

of the influence area

25
G-TVS

N-1 line, N-1 Delta reactive power Above x % of the

Generator in the study injection of generating machine nominal
area units reactive power
Type of node (PV, PQ) Each node switching
from PV to PQ
node(1)
Delta voltage angle at Above x deg
nodes
Transient and short- Load flow in N-0 Load on transit transit lines are
term voltage stability connected to the study loaded above x% of
area natural load / x kA

Import/Export of a
Power exchange of a grid area is above x
grid area % of its load
Short-circuit Contribution to short- Above x% of the
computation in the circuit power of short circuit power at
study area generating units and the relevant node.
active loads
N-1 line, N-1 Delta reactive power Above x % of the
Generator in the study injection of generating machine nominal
area units reactive power
Type of node (PV, PQ) Each node switching
from PV to PQ node
Long-term voltage
Delta voltage angle at Above x deg
stability
nodes
Voltage angle at node Above x deg
in a reference with a
node within the study
area
Delta Transformer tap Above x% of the
range

NOTE(1) : reactive power reserve should be considered in N-0 condition to exclude machines that
have no reactive power reserve in N-0.

Step 1 – built a detailed Multi-TSO load flow model with detailed generation active and reactive
capability limits.
The model detail shall include; Sub-transmission and distribution level is represented as detail as
available for the TSO; Data for running 3-phase symmetrical short circuit calculations according to
IEC-60909.
Physical parameters shall be given for non–equivalent elements. Length of lines shall be
provided. Isolated elements, which are not topologically connected to the rest of the model, shall
not be present. This exclude areas operating in island mode, which would be connected to other
areas once the model is assembled.
All nodes shall have a non-zero nominal voltage.

Step 2 – Perform a load flow analysis, short circuit analysis, for the base case.

26
G-TVS

Step 3 – Perform a contingency analysis, N-1 for lines, N-1 generators in the study area.

Step 4 – Analyze the changes from Step 1 and Step 2 for the following parameters:

 ∆ Q from generators
 ∆ δ voltage angle at nodes
 Nodes changed from P,V to P,Q.
 Line loading, lines with loading equal or above SIL connected to the study area.

Step 5– Perform a short circuit calculation in the study area

 Short circuit contribution from generators and dynamic loads.

In order to provide an order of magnitude for the thresholds of the criteria here above and that
should be defined by ENTSO-e in the aim to harmonize the quality of the dynamic studies
performed across Europe, the influence area used by some TSOs is characterized by the
different numerical values.

In the context of Belgium, static short-circuit computation are computed for each Xnode (The
study area is considered to be the whole Belgium). The sum of short-circuit power of generating
units of remote area should not hold more than 5% of the total short-circuit power delivered by all
generating units. Furthermore, each generating unit individually contributing to more than 5% of
the total short-circuit power delivered by generating unit should be in the influence area. Finally,
N-1 security analysis is computed on the internal grid of Belgium. The sum of absolute deviation
of reactive power of generating unit of remote area should not hold more than 5% of the total sum
of absolute deviation of reactive power. Individual generating units from the remote area should
not deviate in reactive power by more than 5% (deltaQ/Snom) nor switch from PV to PQ.
Generating units individually contributing to more than 5% of the total additional reactive power
delivered by generating units should be in the influence area or in the study area. Finally,
generating units in reactive power limitation or switching from PV to PQ mode should be in the
influence area or in the study area.

Short circuit contribution from neighboring TSO areas has been assessed taking a model from the
national grid development plan 2023. The share von “external” short circuit contribution to
busbars in Germany is between ~15 to 60%. For this grid model no relevant generator unit
transient stability issues were identified. Thus, depending on regional particularities, a high share
of external short circuit contribution cannot be taken as a relevant criteria to define influence area
size.

During dynamic studies the longitudinal grid shape in the TenneT TSO grid including high and
long transits of several hundred kilometers it was seen that dynamic issues are not of relevance if
the current on the transit lines is below 2kA. In case of several lines in the transit corridor have a
current above 2 kA resulting from load flow calculations, stability assessment is required. It must
be clear that such a value highly depends on changes in topology and generation pattern.

4.3 Steps to be conducted for the sizing of the influence area in the context of long-term voltage stability

27
G-TVS

Step 1 – built a reference load flow model with detailed generation active and reactive capability
limits.
The model detail shall include; Sub-transmission and distribution level is represented as detail as
available for the TSO; It data for performing symmetrical short circuit calculations according to
IEC-60909, i.e. for generators it is necessary to have generator type, exciter type, positive
sequence and rated power, nominal voltage, power factor, X’’d, X’d, X’d/Ra, X0/ R0, Ra, R2, R0.
Electrical parameters and ratings for grid elements shall be provided, regardless of whether or not
the elements are under consideration, out of service or undergoing maintenance.
Physical parameters shall be given for non–equivalent elements. Length of lines shall be
provided. Isolated elements, which are not topologically connected to the rest of the model, shall
not be present. This exclude areas operating in island mode, which would be connected to other
areas once the model is assembled.
All nodes shall have a non-zero nominal voltage.

Step 2 – Perform a load flow analysis, short circuit analysis, for the base case.

Step 3 – Perform a contingency analysis, N-1 for lines, N-1 generators in the study area.

Step 4 – Analyze the changes from Step 1 and Step 2 for the following parameters:

 ∆ Q from generators
 ∆ δ voltage angle at nodes
 Nodes changed from P,V to P,Q.
 Line loading, lines with loading equal or above SIL.
 Changes in transformer tap settings.
 Analyze the change in active losses in individual lines.

The following sections will aim at quantifying the accuracy of such an approach with simulations
realized on a realistic CE test system. Faults will be applied on the study area of the reduced
dynamic model (based on the influence and remote minimum requirements) and same faults will
be applied on the original full dynamic model as reference (see chapter 4.5.2). Comparison of the
simulation results will show that the developed equivalent system has high accuracy in
representing the dynamic characteristics of the original system.

4.4 Example of application

The determination of the influence area is illustrated on a Dynamic Study Model (DSM) built by
the SPD working group to reproduce measured interarea oscillations in Continental Europe. This
model describes the whole synchronous interconnection and is available in different simulation
tool format. The generators dynamic models are adequate to represent inter-area oscillations.
According to technology type (hydro, gas turbine, ..) a fixed set of library models (generator +
excitation + PSS + governor) is assigned. Besides, each set has the same parameters in every
European country [ref. xxx].

28
G-TVS

The example of application illustrates the single generator stability (transient stability, short term
voltage stability) in the context of a moderate transit.

4.4.1 Determination of the size of the influence area based on static criteria.

The above mentioned static criteria are used to assess the extension of the influence and remote
area. These criteria allows to determine a first approach to the extent of the influence area
performing steady state analyses.
By looking a set of results from the load flow and short circuit analysis, a set of parameters can be
selected to test

4.4.2 Creation of the reference dynamic model

Step 1 – using a reference model (Initial dynamic model from SPD) and replacing the existing
controls with detailed controls throughout the interconnection by replacing all the SEXS voltage
controls by an IEEE ST7B control with an IEEE OEL control with distinct parameters for each
generator. The governor control and the PSSS are kept for they have been tuned in order to
reproduce measured inter-oscillation modes.
In order to perform short-circuit computations the basic voltage control (SEXS) is replaced by the
more complex IEEE ST7B standard voltage control [IEEE Std 421.5-2005] associated to an over
excitation limiter (IEEE OEL) with distinct parameters for each generator.

29
G-TVS

2 - on the reference model, perform a dynamic simulation of a short circuit at a generation

substation close to a border between two TSOs
3 - built a simplified model following the method here above
A – compute the steady –state short circuit currents and voltages for the short circuit
simulated at step 2
B – determine the substation of the influence area (see method above)
C – replace the voltage control in the remote area by the initial DSM SEXS control and the
uniform parameters
4 - on the simplified model, perform a dynamic simulation of the same short circuit

30
G-TVS

5 – compare the relevant curves of the two dynamic simulations (step 2 and step 4)

4.4.3 Validation of the size of the influence area and of the minimum modeling requirements of the
influence and remote area.

31
G-TVS

5 Implementation issues

5.1 Proposal for approach before CGMES

Static information is exchanged between TSOs with the CGMES model.
Dynamic information is exchanged between TSOs using Excel files with name of CGMES
standard models and customized parameters for minimum requirements. More detailed models
(static or dynamic) can be exchange if TSOs agree on a given software proprietary format.

5.2 Target approach with CGMES

Proposal should be in line with static and dynamic CGMES 2.4.13 for minimum requirements.

As the use of CGMES (Common Grid Model Exchange Standard) has been adopted by ENTSO-
E as a common format for exchange, and the development of the dynamic profile in CGMES
2.4.13, results along with the model requirements and the Guidelines for Network Modelling a
strong tool to facilitate model exchange for voltage and transient stability analysis within ENTSO-
E community.

As it has been stated before, the model content inside the influence area includes information
about the generator and the controllers defined in 2.3 have to be modelled by manufacturer
information or by a model library with the set of owner parameters that characterize the dynamic
behaviour of the manufacturer model.

For these purpose there are a number of controllers currently available and in testing process by
vendors. (see Appendix 6.1.4)

5.3 Model reduction guidelines

The network may be represented with an equivalent network according to the

standards/methodologies defined by the relevant TSO.

An equivalent network is the result of a process that converts an initial network (with many
buses/elements) into a reduced order network that contains only the boundary buses and the
most relevant initial buses.

As a reference the process for dynamic reduction of the influence area and remote area could be
the next:

Reduction of the influence area

The process for dynamic reduction of the influence area can be divided into three steps:

1. Identification of important generators and loads. Generators of the influence area are
important when tend to swing together following a disturbance in the study area. The
swing curves of the generators in the same area are most identical.
2. Aggregation of the rest of generators and loads. After important generators are identified,
the rest of generators can be aggregated to an equivalent generator. The equivalent
generator is the sum of the power of the generators in the same technology group,

32
G-TVS

separately –all the parameters based on the same MVA base. Therefore the inertia and
damping constant of the equivalent generator can be defined as the sum of the inertia and
damping constants of the generators. The equivalent load may be added by sector if
possible (demand: commercial, industrial and residential). It is also possible to use a more
detailed model if available.
3. Reduction of the network. Network reduction is the process of deleting some load buses
and branches in order to reduce the size and complexity of the network.

Reduction of the remote area

The process for dynamic reduction of the remote area can be divided into two steps:

1. Aggregation of the generators and loads to an equivalent generator and load. Therefore
the inertia and damping constant of the equivalent generator can be defined as the sum of
the inertia and damping constants of the generators.
2. Reduction of the network. Network reduction is the process of deleting buses and
branches in order to reduce the size and complexity of the network.

5.4 Model maintenance / data exchange triggering

Once the influence area has been defined and agreed, between two or more TSOs, the model
exchange process can be started.

As a reference the model definition and content the models proposed in G-NM can be used.
Inside the influence area: the Faraday model is the more suitable for voltage and transient
stability analysis,

Model description for influence area:

- It contains data for performing dynamics analyses, e.g. generator model, governor model,
related control systems – AVR, exciter, PSS, etc.
- The load flow part of the model is the same as in Tesla model type. However, it could be
derived from Norton or Ampere (in case Faraday should be node-breaker kind) as well.
- A Faraday model contains, manufacturer data for the generators and tuned parameters,
i.e. the most detailed data that the model authority (TSO) has.
- The information of the parts of loads that are constant impedance, constant current and
constant power shall be provided.
- All aggregated of non-conventional generation (solar, wind, biomass, etc.) shall be
represented separately by technology type if these generation types are aggregated in the
Tesla model type. In the Tesla model type the Sub transmission and distribution level
represented as detailed as available for the TSO.

33
G-TVS

In the remote area: the Coulomb model is the more suitable for voltage and transient stability
analysis,

Model description for remote area:

- The load flow part of the model is the same as in Volt model type (dynamic data is added).
- It contains data for performing dynamics analyses, e.g. generator model, governor model,
related control systems – AVR, exciter, PSS, etc.
- It can contain manufacturer data for the generators or standard representation of the
control equipment.

For the Volt model the following rules apply:

[MP.V-1] For load flow the information of Equipment, Topology, voltage profiles and
boundary points shall be used. If available, the geographical information and single line
diagrams can be provided.

Note : for CGMES this information corresponds to: EQ, TP, SV, SSH, DL, GL, extensions.
Boundary points, as defined in BP database, BP_EQ, BP_TP, extensions.

[MP.V-2] Step-up transformers for generators or condensers are not explicitly

represented as transformers in the models whether the high voltage side is connected at
220 kV or above. The transformer impedance can be neglected or, added to the
generator’s impedance.
[MP.V-3] Detailed representation of the non-radial network of Nominal voltage less
than or equal 150 kV and greater than or equal to 110 kV150 kV, 132 kV and 110 kV
levels is not necessary.
[MP.V-4] Branches and substations of the network under the 110 kV voltage level will
not be represented in detail. Loads would be aggregated at the closest extra-high voltage
(EHV) node to represent the actual loading on the connection point. (Losses should be
considered.) Embedded generation should be represented as a load+generation bus.
[MP.V-5] Radial connections below 220 kV – except generators and condensers
connected by step up transformers – shall be aggregately represented at its EHV or HV
network connection point.
[MP.V-6] Generators can be modelled at their terminal voltage level or connected
directly to a EHV-HV bus. In this case an equivalent impedance shall be used for
generator and transformer.
[MP.V-7] In case of more than one generator, load, etc. connected to a particular
node, no aggregation of generation and loads shall be made.
[MP.V-8] Generating units connected to 110 kV network or above with maximum
capacity greater than 75 MW for Continental Europe, 30 MW for Nordic and Great Britain,
15 MW for Baltic and 10 MW for Ireland shall be modelled explicitly.
[MP.V-9] For embedded generation not fall under MP.V-8 must be modelled in line
with MP.V-4 for each technology type and node.
[MP.V-10] Rules defined in Operation Handbook, Policy 3 (A.3) for ratings are applied
in case of no precise description defined by the WG responsible for the process. Ratings

34
G-TVS

should be omitted for model elements, which are part of an electrical equivalent
(EquivalentEquipment). The ratings must be omitted if no rating is available or it is infinite.

[MP.V-11] Topology processing to obtain the bus branch model shall be consistent
with operational data.

[MP.V-12] Retained breakers can be present in the model.

[MP.V-13] Ratings shall be included for; breakers, bus bars, generators, transformers
and lines.

[MP.V-14] Individual models submitted shall be solved without overloads before a

contingency study.

[MP.V-15] Individual models shall be solved with tap changers, reactive power limits,
Switched shunts and phase shifter transformers enabled.

5.5 Model exchange/ legal requirements

The cooperation between Network Operators shall take in to account the next confidentiality
obligations:

1. Each Relevant Network Operator and Relevant TSO shall preserve the confidentiality of
the information and data exchanged and shall use them exclusively for the purpose they
have been submitted.

2. Notwithstanding the above, disclosure of such information and data may occur in case a
Relevant Network Operator or a Relevant TSO is compelled under EU or national law to
disclose it, under the conditions set forth in the relevant legislation. Such disclosure shall
be reported to the owner of such information and data.

3. In case of disclosure for other purposes than those described in paragraph 1 and 2, a
Relevant Network Operator or a Relevant TSO shall seek the consent of the owner of
such information and data. The Relevant Network Operator or Relevant TSO shall provide
the motivation for this disclosure. This consent cannot be unreasonably withheld.

35
G-TVS

6 Appendix

6.1.1 Abbreviations and definitions

6.1.2 Difference between transient, short term and long term voltage stability

The IEEE/CIGRE Join Task Force on Stability Terms and Definitions has classified power system
stability in categories according to the next figure:

Figure. Classification of power system stability according to IEEE/CIGRE Join Task Force on
Stabilty Terms and Definitions (IEEE 2004)

Large-disturbance rotor angle stability (transient stability)

Large-disturbance rotor angle stability or transient stability, as it is commonly referred to, is

concerned with the ability of the power system to maintain synchronism when subjected to a
severe disturbance, such as a short circuit on a transmission line. The resulting system response
involves large excursions of generator rotor angles and is influenced by the nonlinear power-
angle relationship. Transient stability depends on both the initial operating state of the system and
the severity of the disturbance. Instability is usually in the form of aperiodic angular separation
due to insufficient synchronizing torque, manifesting as first swing instability. However, in large
power systems, transient instability may not always occur as first swing instability associated with
a single mode; it could be a result of superposition of a slow inter-area swing mode and a local-
plant swing mode causing a large excursion of rotor angle beyond the first swing. It could also be
a result of nonlinear effects affecting a single mode causing instability beyond the first swing.
The period of interest in transient stability studies is usually 3 to 5 seconds following the
disturbance. It may extend to 10-20 seconds for very large systems with dominant inter-area
swings. Transient stability is categorized as short term phenomena.

Voltage stability

Voltage stability refers to the ability of a power system to maintain steady voltages at all buses in
the system after being subjected to a disturbance from a given initial operating condition. It
depends on the ability to maintain/restore equilibrium between load demand and load supply from

36
G-TVS

the power system. Instability that may result occurs in the form of a progressive fall or rise of
voltages of some buses. A possible outcome of voltage instability is loss of load in an area, or
tripping of transmission lines and other elements by their protective systems leading to cascading
outages. Loss of synchronism of some generators may result from these outages or from
operating conditions that violate field current limit.

Small disturbance voltage stability

Small-disturbance voltage stability refers to the system’s ability to maintain steady voltages when
subjected to small perturbations such as incremental changes in system load. This form of
stability is influenced by the characteristics of loads, continuous controls, and discrete controls at
a given instant of time. This concept is useful in determining, at any instant, how the system
voltages will respond to small system changes. With appropriate assumptions, system equations
can be linearized for analysis thereby allowing computation of valuable sensitivity information
useful in identifying factors influencing stability. This linearization, however, cannot account for
nonlinear effects such as tap changer controls (dead bands, discrete tap steps, and time delays).
Therefore, a combination of linear and nonlinear analyses is used in a complementary manner.

Short-term voltage stability involves dynamics of fast acting load components such as induction
motors, electronically controlled loads, and HVDC converters. The study period of interest is in
the order of several seconds, and analysis requires solution of appropriate system differential
equations; this is similar to analysis of rotor angle stability. Dynamic modelling of loads is often
essential. In contrast to angle stability, short circuits near loads are important.

Large disturbance voltage stability

Large-disturbance voltage stability refers to the system’s ability to maintain steady voltages
following large disturbances such as system faults, loss of generation, or circuit contingencies.
This ability is determined by the system and load characteristics, and the interactions of both
continuous and discrete controls and protections. Determination of large-disturbance voltage
stability requires the examination of the nonlinear response of the power system over a period of
time sufficient to capture the performance and interactions of such devices as motors, under load
transformer tap changers, and generator field-current limiters. The study period of interest may
extend from a few seconds to tens of minutes. Long-term voltage stability involves slower acting
equipment such as tap-changing transformers, thermostatically controlled loads, and generator
current limiters. Stability is usually determined by the resulting outage of equipment, rather than
the severity of the initial disturbance. Instability is due to the loss of long-term equilibrium (e.g.,
when loads try to restore their power beyond the capability of the transmission network and
connected generation), post-disturbance steady-state operating point being small-disturbance
unstable, or a lack of attraction toward the stable post-disturbance equilibrium (e.g., when a
remedial action is applied too late). Where timing of control actions is important, this should be
complemented by quasi-steady-state time-domain simulations.

Main difference between small disturbance and large disturbance voltage stability in the terms of
considered period of time and other element influence is given in the following table.

37
G-TVS

Voltage
Small disturbance Large disturbance
Stability

PV and QV curves (load

flow results) of the
PV and QV curves (load flow results) faulted state
dV/dQ sensitivities
Long Term Long term dynamic
long term dynamic models including tap-
models including tap-
changers, var-control, excitation limiters, etc.
changers, var-control,
excitation limiters, etc.
Dynamic models (short
term) special importance
Short Term
on dynamic load
modelling
Reactive power
contribution of Large (thermal overload Limited by
synchronous capabilities) overexcitation limiters
generation
Switchable No contribution
High contribution
shunts (switching times too high)

SVC/TSC High contribution High contribution

Often voltage instability and rotor-angle instability occur together, i.e. one may lead to the other
and the distinction may not be clear, although there are some differences based on which this
phenomena can be distinguished. It is of common experience that rotor angle stability as well as
voltage stability is affected by reactive power control. The small disturbance voltage stability
usually leads to increase of rotor angle. However, there is a very important difference between
the rotor angle stability and the voltage stability – the former is basically the generator stability
while the latter is usually load voltage stability. In a large interconnected power system, voltage
collapse of load is possible without loss of synchronism of any generator. While comparing the
transient voltage stability with transient rotor angle stability, it may be mentioned that transient
voltage stability is also related to rotor angle stability though voltage collapses focus on load
mainly and deal with the magnitude of the load voltage; the transient voltage stability focuses on
generators and power angles mainly. The cause of angle instability is usually located near to the
generator and may also involve voltage collapse. On the other hand, if the voltage collapse is in
the load area, the collapse not involve angle stability. In short, it can be said that the main
difference between voltage stability and angle stability is that voltage stability depends on the
balance of reactive power demand and generation in the system where as the angle stability
mainly depends on the balance between real power generation and demand.

38
G-TVS

6.1.3 References

[1] Network Code on Requirements for Grid Connection Applicable to all Generators (RfG)
[2] Network Code on Operational Security (OS)
[3] Guidelines Network Modelling- ENTSO-E
[4] IEEE/CIGRE Joint Task Force on Stability Terms and Definitions, “Definition and Classification
of Power System Stability. IEEE Transactions on power systems, vol. 19, no.2, May 2004.
[5] [2.5.1] Modelling and Aggregation of Loads in Flexible Power Networks WG C4.605
[6]
[7]
[8]
[9]
[10]

6.1.4 Existing CGMES controllers in development and testing

Governor:

GovSteam0
GovHydroIEEE0

GovHydroIEEE2

GovSteamIEEE1

GovCT1
GovCT2

GovGAST

GovGAST1

GovGAST2

GovGAST3

GovGAST4

GovGASTWD

GovHydro1

GovHydro2

GovHydro3
GovHydro4
GovHydroDD

39
G-TVS

GovHydroFrancis

GovHydroPelton

GovHydroPID

GovHydroPID2

GovHydroR

GovHydroWEH

GovHydroWPID

GovSteam1

GovSteam2

GovSteamCC
GovSteamEU

GovSteamFV2

GovSteamFV3

GovSteamFV4

GovSteamSGO

PSS:

PssIEEE2B
PssIEEE1A
PssIEEE3B

Pss2ST

Pss1

Pss1A

Pss2B

40
G-TVS

PssIEEE4B

Pss5

PssSH

PssSK

PssPTIST3

PssSB4

PssELIN2

PssPTIST1
PssWECC

OEL/ UEL:

OverexcLimIEEE
OverexcLim2
OverexcLimX1
OverexcLimX2

UnderexcLimIEEE1
UnderexcLimIEEE2
UnderexcLim2Simplified
UnderexcLimX1
UnderexcLimX2

Also the following models are available and in testing:

Load:

LoadComposite
LoadGenericNonLinear
LoadAggregate
LoadStatic
LoadMotor

Wind Turbine-IEC 61400-27:

41
G-TVS

Type 1A
Type 1B
Type 2
Type 3A
Type 3B
Type 4A
Type 4B

42