See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/312413618

Current procedures and practices on grid code compliance verification of

renewable power generation

Article in Renewable and Sustainable Energy Reviews · January 2017

DOI: 10.1016/j.rser.2016.12.051

CITATIONS READS

26 4,434

5 authors, including:

Agurtzane Etxegarai Pablo Eguia

University of the Basque Country University of the Basque Country
40 PUBLICATIONS 426 CITATIONS 137 PUBLICATIONS 1,376 CITATIONS

SEE PROFILE SEE PROFILE

E. Torres Araitz Iturregi

University of the Basque Country University of the Basque Country
67 PUBLICATIONS 691 CITATIONS 36 PUBLICATIONS 371 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Pablo Eguia on 29 December 2017.

The user has requested enhancement of the downloaded file.

Current procedures and practices on grid code compliance verification of renewable power
generation

Agurtzane Etxegaraia,∗, Pablo Eguia, Esther Torres, Garikoitz Buigues, Araitz Iturregi
a Department of Electrical Engineering, University of the Basque Country UPV/EHU, Bilbao

Abstract
Generation assets applying for grid connection must comply with certain grid code requirements. Grid code compliance verification
shall include revision of documentation covering technical data and models, checking of requested capabilities, and validation of
model performance. These procedures are singular regarding renewable power generation, due to their technical characteristics,
specific topologies and short experience. This paper aims to carry out an updated review of the international regulations and
current practices regarding the verification and certification of the electrical performance in renewable generation systems. Grid
code compliance can be verified by practical tests or by simulation of validated models. Therefore, this paper also encompasses
modelling and validation requirements, highlighting challenges caused by current procedures.
Keywords: Grid code, compliance, generator testing, staged fault testing, model validation, generic model, verification,
certification, wind power, renewable generation, commissioning.

Contents approval for its connection to a grid. Demonstrating grid code

compliance and achieving a grid connection agreement are, there-
1 Introduction 1 fore, essential milestones in the development of a power plant
project. The increasing penetration of renewable generation
2 Renewable power generation modelling and simula- plants formed by a large number of individual generating units
tion 2 poses a challenge to system operators, in terms of technical
2.1 Challenges regarding renewable power genera- singularities, connection process and plant modelling manage-
tion modelling . . . . . . . . . . . . . . . . . . 3 ment. In order to cope with these issues, specific compliance
2.2 Root Mean Square (RMS) against ElectroMag- procedures based on testing and simulation have already been
netic Transient (EMT) models . . . . . . . . . 5 established for Renewable Energy Sources (RES). The present
2.3 Generic models against proprietary models . . 6 paper introduces current procedures and practices on grid code
compliance verification for renewable power generation.
3 Renewable power generation model validation 7
Grid code compliance verification has a double objective.
3.1 Test data collection . . . . . . . . . . . . . . . 7
On the one hand, plant owners are responsible for demonstra-
3.2 Model simulation . . . . . . . . . . . . . . . . 8
ting compliance of the grid code to the relevant network opera-
3.3 Model validity acceptance rules . . . . . . . . 9
tor. And, on the other hand, network operators have to assess
4 Verification of the compliance of technical require- the compliance in order to ensure that the new plant will not
ments 9 adversely affect the secure operation of the power system. A
4.1 Review of compliance verification practices . . 10 grid code should be complemented by a good verification plan,
4.2 Certification procedures . . . . . . . . . . . . . 12 in order to avoid misinterpretations of the requirements. Ac-
4.2.1 Germany: FGW-TG8 procedure . . . . 12 cording to ENTSO-E [1], compliance testing is defined as the
4.2.2 Spain: PVVC procedure . . . . . . . . 13 process of verification that power generating facilities comply
with the specifications and requirements provided by this grid
5 Conclusions 15 code. It can be carried out before starting operation of new
installations. The verification should include the revision of
documentation (including technical data and models), the veri-
1. Introduction
fication of the requested capabilities of the facility by practical
Grid codes specify the electrical performance that genera- tests and simulation studies and, finally, the validation of the
tion assets must comply with in order to obtain the required model performance based on actual measurements. The grid
code compliance shall be maintained throughout the lifetime
∗ Corresponding
author. Tel.:+34 946017225 of the facility. Hence, power plants shall undergo periodical
Email address: [email protected] (Agurtzane compliance monitoring processes, in order to verify that their
Etxegarai)

Preprint submitted to Renewable and Sustainable Energy Reviews September 23, 2016
technical capabilities are maintained and simulation models are verification can be carried out by testing and simulation tech-
still valid. niques. Hence, this paper introduces requirements regarding
A grid code verification plan is as important as the regula- modelling and model validation in the first place . Modelling
tion itself and it should not need to leave room for interpreta- issues are highlighted in the paper, including: specific chal-
tion regarding how each requirement shall be assessed. Un- lenges involved when modelling renewable power plants, use
fortunately, not every grid code is complemented by a clear of RMS or EMT models, and the conflicting perspectives of
and detailed compliance verification plan. Grid code evolution operators and manufacturers regarding the disclosure of power
has been extensively studied in the literature, mainly focused plant models. Model validation consists of three main steps:
on technical requirements for large Wind Power Plants (WPP). data collection, simulation and the acceptance of the model va-
Several reviews including thorough analysis and comparison of lidity. The paper includes the most usual approaches for ac-
grid codes were conducted in last years, due to the continuous complishing each step. Finally, the paper reviews the current
revisions of regulatory frameworks, such as [2–13]. practices in selected countries to verify the compliance with
Tsili et al. conduct a comprehensive review including rele- technical requirements, such as voltage and frequency control,
vant European Transmission System Operators (TSOs), several active and reactive power control, frequency response or Low
operators in Canada, the Federal Energy Regulatory Commis- Voltage Ride-Through (LVRT).
sion (FERC) regulation in US and New Zealand in [2]. Re- The present study is a continuation of the work presented
ference [3] compares regulation for wind farms in European by the authors in [10]. Therefore, the countries under analysis
countries such as Denmark, Germany, Spain, UK and Ireland, are Australia, Denmark, Germany, Ireland, New Zealand, Spain
along with the technical rules in China. Mohseni and Islam and UK. The selection covers European countries with a signifi-
study countries with significant wind power penetration, such as cant penetration of RES and pioneering regulation frameworks.
Denmark, Germany, Spain, UK, Ireland, US, Canada or Nordic Among them, Ireland and UK are insular territories with weak
countries in [6]. Reference [8], focused on wind power inte- interconnections, where transient stability and frequency regu-
gration in Europe, not only does it compare national specifi- lation can be critical. New Zealand also belongs to this category
cations, but it also mentions grid code compliant wind gene- and, in Australia, wide parts of its power system are weakly
ration technologies. Diaz-Gonzalez et al. present a review of linked. So, the point of connection of large RES power stations
selected European grid codes and future trends regarding the will have low short-circuit values. This way, for a more insight-
participation of WPPs in frequency control in [9]. Singularities ful analysis, this paper includes a broad spectrum of countries
of the regulatory framework for renewable energy sources in with contrasting power system structures, RES penetration de-
weak and isolated power systems are analysed and compared in grees and network strength values.
[10, 12], including New Zealand, Ireland, UK, Denmark, insu- This paper is organised as follows. Section 2 describes and
lar territories in France and Spain, and Australia. Requirements compares renewable power generation asset modelling and si-
for ensuring transient stability and frequency regulation prove mulation requirements, reviewing most challenging aspects. Si-
to be critical in these power systems. [13] discusses the ad- mulation models must be accompanied by validation tests to
vanced grid code requirements for the integration of large scale show the validity of the models. Model validation practices and
photovoltaic power plants in the transmission system, given the practical set-ups are indicated in Section 3. Finally, procedures
expected growth in PV installed capacity in next years. The pa- to verify the conformity of RES installations with technical re-
per compares technical requirements to be met by PV installa- quirements (including voltage and frequency control, active and
tions in Germany, China, Romania, Puerto Rico, US and South reactive power control, frequency response and LVRT) are anal-
Africa. On the whole, literature has mainly analysed technical ysed in Section 4, with special emphasis on certification proce-
requirements to be met by large renewable power plants. dures in Spain and Germany. Section 5 presents the conclusions
However, grid code compliance verification remains a ma- of the study.
jor open issue, as pointed out by [2]. In the regulatory frame-
works, grid code verification as well as generation unit and sys- 2. Renewable power generation modelling and simulation
tem certification procedures are still at an early stage. Litera-
ture related to regulation verification and certification aspects is According to ENTSO-E [1], system operators may require
sparse, scattered or focused on a single country [4, 14]. Often, generating unit models for both steady state and dynamic simu-
it can also be found within grid code reviews [6, 15], even for lations (50 Hz component) as well as for electromagnetic tran-
distributed generation [16]. Several testing methods for com- sient simulations, where appropriate and justified. Model for-
pliance verification have been proposed in [17–20]. Regarding mat must also be provided, and model structure and block di-
conformity verification through simulation techniques, a review agrams shall be documented. Regarding dynamic simulations,
of modelling and simulation requirements for variable genera- models shall contain submodels of alternator and prime mover,
tion in the grid codes would also be useful [21]. Previous works speed and power control, voltage control, protection and con-
can be found in [22–24]. verters. Thus, system operators have introduced specific con-
Therefore, this paper aims to fill the gap and carry out an up- ditions for the inclusion of power plant models and simulation
dated review of the international regulations and current prac- procedures to be followed by manufacturers, in order to check
tices regarding the verification and certification of the electri- the viability of the installation and the conformity with grid
cal performance in renewable generation systems. Compliance code rules.
2
 Grid code requirements regarding data, modelling and si- • In some areas, such as Australia, it is becoming more
mulation have been previously reviewed in [22, 23]. [22] ga- common to connect renewable generation in weak points
thers practices by several system operators regarding modelling of the network. Concerns about simulation model ac-
requirements, ranging from Argentina, where non-confidential curacy under these circumstances is highlighted in some
and non-black box models are required for all WPPs above 10 references [14, 44].
MW, and Croatia, where no generator model is required before
connecting a generator. However, the enquiry was carried out Another singularity of RES modelling must be also noted.
in 2005, and since then, modelling requirements have evolved. Conventional power plants are constituted by either a single
Therefore, requirements regarding modelling and simulation in large unit or a few large units. In contrast, a variable generation
the countries under study have been described and compared in installation such as a WPP can be made up of multiple small
this section, including the application scope, model character- size generating units. Therefore, for large scale power system
istics, and simulation requirements. simulations it is often preferred to reduce the whole power plant
Table 1 indicates the documents containing modelling and to a single equivalent unit. Thus, aggregated simulation models
validation prescriptions required by system operators for re- are often accepted for large multi-generator power plants for
newable generation. Regulations are often complemented by practical reasons. The aggregation approach was evaluated in
guidelines with a more specific explanation. This is the case [45] and its validity was found adequate for wind farm transient
for most of the countries under study. Regarding generation as- studies, based on the National Renewable Energy Laboratory
sets, several approaches are used in the regulation under review: (NREL) equivalencing method [46], as long as wind speed is
technical regulation in Australia and New Zealand is techno- considered constant during the grid event. In some cases, when
logically neutral; Ireland has separate specifications for wind a WPP consists of different Wind Turbine Generator (WTG)
power, whereas in Denmark both wind and PV power must types or distinct clusters, using two or more equivalent genera-
meet specific requirements; and finally, in UK and Spain re- tors is also accepted [47]. However, the aggregation approach
quirements not only apply to wind power but also to other in- might not be adequate to represent a large wind farm by a sin-
termittent energy technologies. gle equivalent for all cases. This can be the case for large wind
Modelling issues for RES are analyzed below. Table 2 sum- farms with multiple feeders, where the response of individual
marises and compares modelling and simulation specifications or groups of turbines might be different, or power plants where
in the countries under study in the present paper. generators are operating at different speed. The equivalencing
of collector, transformer and compensation devices might show
2.1. Challenges regarding renewable power generation mode- the same issues.
lling The scope of application for modelling requirements nor-
In the traditional power systems, it was not necessary to mally includes both the generating unit and the complete gene-
include renewable power generation models in dynamic sim- ration system. In Australia, the complete power plant model in-
ulations, because penetration was still low. Thus, disconnec- cludes dynamic reactive power support devices and the power
tion of renewable generating units during disturbances was a plant controller. The German regulation indicates the mode-
usual practice. Nowadays, situation has changed and many grid lling requirements at generating unit level, whereas modelling
codes require manufacturers and generators to supply valid dy- fundamentals for power generation system are only described
namic models. However, several challenges have been reported as a framework for future application. Installation models shall
in the literature regarding renewable power generation mode- include generating units, transformers, cables, reactive power
lling [42, 43]: compensation systems and the external grid. In Ireland, the
wind farm model shall include the WTG models, converter con-
• Generators are usually based on power electronic devices. trols, reactive compensation and protection relays.
Thus, modelling can pose some issues, especially regar- Among the countries under study, in Australia all new ge-
ding control systems and algorithms that are often pro- nerating plants must provide validated models, regardless of the
prietary. technology. Requirements regarding modelling are technology-
neutral, analogously to technical requirements imposed on ge-
• Available models typically represent only large signal per- neration assets in the grid code [42]. In the Danish grid code,
formance, but the impact of wind farms on small signal synchronous generators, as well as wind farms with a power
performance needs also to be assessed. output greater than 1.5 MW, must supply a valid dynamic mo-
del. The minimum capacity for WPPs with the obligation to
• Performance under unbalanced network conditions, cau-
supply modelling information is set to 5 MW in the Republic
sed by unbalanced faults or asymmetric line impedances of Ireland, as well as in UK, where it applies to any power park
and loads, can significantly impact power electronic con- module type, i.e. generating units powered by an intermittent
trol systems, and making it difficult to model using only
power source. In Germany, all generators are subject to dy-
positive sequence models [42, 43]. Hence, some grid
namic model provision, unlike New Zealand where little infor-
codes require models that can represent both balanced
mation can be found with regard to asset modelling. In general,
and unbalanced situations.
generation assets with a power output smaller than 30 MW are
excluded from providing asset capability information.

3
 Table 1: Renewable generation modelling and validation requirements in the countries under study

Country Title Technology

National Electricity Rules [25]
Generating System Model Guidelines [26]
Australia
Data and Model Requirements for generating systems of less than 30
Any technology
MW [27]
Dynamic Model Acceptance Guideline [28]
Technical regulation 3.2.5. for wind power plants with a power output
Denmark Wind power
above 11 kW [29]
Technical regulation 3.2.2. for PV power plants with a power output
PV
above 11 kW [30]
FGW Technical Guidelines for Power Generating Units Part 3, Deter-
mination of electrical characteristics of power generating units and sys-
Germany
tems connected to MV, HV and EHV grids [31]
FGW Technical Guidelines for Power Generating Units Part 4, De-
mands on Modelling and Validating Simulation Models of the Electrical Any technology
Characteristics of Power Generating Units and Systems [32]
Republic of Ireland EirGrid grid code [33] Wind power
Connecting and Dispatching New Generation in New Zealand.
Any technology
New Zealand Overview [34]
Asset Capability Information Overview. Guideline [35]
P.O. 9.0 Información intercambiada con el operador del sistema [36] Any technology
Guı́a descriptiva del procedimiento de puesta en servicio [37] Any technology
Spain
Requisitos de los modelos de instalaciones eólicas, fotovoltáicas y todas
aquellas que no utilicen generadores sı́ncronos directamente conecta- Any non-synchronous generation
dos a la red [38]
Condiciones de validación y aceptación de los modelos [39] Any non-synchronous generation
The Grid Code [40] Any technology
UK
Guidance Notes-Power Park Modules [41] Any non-synchronous generation

4
 Table 2: Summary of dynamic model requirements in the countries under study

Country Size Aggregation Source code SW Time step

Australia All X > 30 MW PSS/E 1 ms
Denmark > 1.5 MW X > 25 MW χ Variable
Germany All X χ χ 10 ms
Ireland > 5 MW X χ PSS/E 5 ms
Spain > 10 MW X > 10 MW PSS/E 5 ms
UK > 5 MW X χ DigSILENT 10 ms

Aggregated models are also accepted in the countries under network is large [14]. So, some simplifications need to be as-
review. Though, in Australia the model aggregation method- sumed. For instance, positive sequence EMT models are used
ology must be clearly specified and in Denmark, it must be for routine stability studies in Australia [14].
proved that aggregation does not significantly impact simula- When the power system topology and simulated disturbances
tion results. The aggregate models of wind farms must include are balanced, positive sequence models are adequate. However,
in some cases the central park level controller (Australia), as unbalanced conditions can affect power electronics and, there-
well as the collector network (Ireland). fore, are required to be studied. It can be performed by using
three-phase RMS or EMT simulations. In addition, positive se-
2.2. Root Mean Square (RMS) against ElectroMagnetic Tran- quence models might not be adequate for representing sufficient
sient (EMT) models details of the converter control system [14].
Static and dynamic simulations are needed for operational, Nevertheless, positive sequence RMS models must be handed
planning, interconnection and plant design purposes. System in generally by manufacturers . This is the case for large power
models are required at three general levels [32, 45]: plants in strong areas of Australia, Germany, and UK. In Den-
mark, it must be possible to use the simulation models for RMS
• Load flow and short-circuit models are used for basic de- balanced and unbalanced studies. Besides, the Danish grid code
sign studies. indicates that models must be valid for a frequency range of 47-
53 Hz and 0-1.4 p.u. of voltage.
• Positive sequence or RMS models have traditionally been Related to the model type, model minimum constants and
used for system integration studies and stability studies. simulation time steps are often specified. In Australia, tran-
Perfect balanced conditions are assumed and stability is- sient stability models must allow numerically stable and ac-
sues under study tend to be bounded within a small fre- curate performance for time step-sizes down to 1 ms. Time
quency band around the fundamental frequency of the constants of less than 5 ms should only be included if their in-
system. clusion is critical. Regarding model time constants, dynamics
• Detailed three-phase EMT level models are necessary to under 5 ms and 10 ms must be discarded in Spain and in UK,
study the effect of fast transients and electromagnetic in- respectively. In Ireland, simulation time steps must be higher
terference, which require higher frequency components. than 5 ms, whereas in Germany the limit is set on 10 ms, even
if it must be demonstrated that simulations with different incre-
The bandwidth of dynamic models is directly related to the ment sizes obtain equivalent results. Last, simulation models in
required simulation time-step and, thus, to the resulting simu- Denmark must be capable of using numerical equation solvers
lation speed. EMT models result into smaller time-steps and with variable time steps.
longer simulation times. However, they are necessary, because System operators often specify the compulsory software si-
phenomena such as Subsynchronous Resonance (SSR), Sub- mulation package. On the whole, the preferred option is PSS/E
synchronous Torsional Interaction (SSTI) or the study of be- from Siemens. In UK, the model structure and complexity must
haviour of variable generators in weak nodes of a power sys- be suitable to be integrated in Powerfactory from DigSILENT.
tem require detailed models. For instance, in Australia, detailed However, model could be implemented in the software package
EMT-type models must be provided when seeking assessment chosen by the manufacturer. In Denmark and New Zealand,
of the model for a Short-Circuit Ratio (SCR) lower than three. there is no indication about the software simulation package to
In addition, transient stability models should have a bandwidth be used. However, according to [43], the system operator in
of at least 0.05Hz to 10Hz. EirGrid regulation indicates that New Zealand performs steady-state, dynamic and transient net-
dynamic models must represent features likely to be relevant works analysis using DigSILENT Powerfactory.
to angular and voltage stability. However, using EMT-type mo-
dels for the whole system can be impracticable if the connecting
5
2.3. Generic models against proprietary models The development of generic models regarding variable gen-
Regarding dynamic modelling requirements, system opera- erators has been initially focused on wind power. During re-
tors and manufacturers have conflicting perspectives [23]. Sys- cent years, generic models representing different types of wind
tem operators prefer to use standard models representing the turbines and their controls have been published and provided
plant performance with an adequate accuracy and simple enough for power system studies [21, 45, 50, 51]. They allow simu-
to be included in large network simulations. On the contrary, lating the typical behaviour of wind generator types and con-
equipment manufacturers are concerned with achieving a high trol concepts, and are well suited for general power system
degree of accuracy and protecting their intellectual property. planning studies or feasibility studies. Wind speed variations
They are reluctant to disclose details of their products and, hence, have been commonly neglected in proposed generic WTG mo-
models are often not standardised. Propietary models include dels, as it can be considered constant during the simulation pe-
user-written positive sequence models and three-phase detailed riod (up to 20 seconds in typical transient stability simulations
equipment models. However, in recent times manufacturer- [49]). Generic models not only have been proposed for wind
specific models have become available in software tools such farms, but also for solar photovoltaic generation [52–55], reac-
as DigSILENT [43]. tive power compensation equipment (SVC, STATCOM) [56] or
In Australia, black box type representations are not accepted High Voltage Direct Current (HVDC) [57, 58]. In the United
by the system operator. Moreover, functional block diagrams States, many of the initiatives come from the Western Electric-
and model source code are compulsory for generating systems ity Coordinating Council (WECC), whereas the National Re-
over 30 MW, preferably in source code formats FORTRAN and newable Energy Laboratory (NREL) has been engaged in an
FLECS. In Denmark, black box modelling is allowed for in- extensive model validation project aimed at testing the models
dividual WTGs making up a wind farm with a power output against field measurements and refining the WECC generic mo-
greater than 25 MW. Simulation models must be supplied in dels if needed. Models proposed by the WECC are included in
the form of block diagrams using primarily transfer functions PSS/E, PSLF and PowerWorld software packages. Internation-
and accompanied by model descriptions. The source code must ally, IEC 61400-27-1 is an ongoing effort to standardise generic
be sent and encrypted parts are not acceptable. In Germany, a simulation models of individual wind turbines for transient sta-
grey box approach is adopted. Even if during generating unit bility simulation in large power systems. The main differences
certification black box models are accepted, they must be ac- with regard to WECC models is that a common modular struc-
companied by simplified open models. In Spain, user models ture applies to all wind turbine types and the turbine and plant
shall be provided as open code objects in FLECS or FORTRAN model are separated [59]. A future release of IEC 61400-27-2
programming languages. will include the plant controller.
During recent years, the need of harmonised generic mo- For generating systems of less than 30 MW, the Australian
del standards for the different parts and applications of power Energy Market Operator (AEMO) accepts the use of generic
systems has been enhanced [48]. Generic models are simpli- models for connection studies (e.g. standard IEEE models for
fied and publicly available. These models have a simple but synchronous generators or IEC/WECC generic wind farm mo-
comprehensive structure and constitute a generic structure ba- dels) provided that the model can reasonably represent the plant
sed on physical principles, which enable to emulate the design components of the generating system and the SCR is reasonably
of different manufacturers simply by changing appropriate pa- high [27]. In Ireland, user-written models must be supplied if
rameters [43]. Generic models must (1) allow a straightfor- there is no suitable library model. Similarly, in Spain a list with
ward exchange of model data between interested parties, (2) be preferred models regarding conventional generation assets, as
implementable and their performance comparable in different well as wind and photovoltaic power, reactive compensation
simulation programs, and (3) their parameters tunable to best systems such as Flexible AC Transmission Systems (FACTS)
represent the specific equipment, without having to reveal pro- and protection relays, has been released. For synchronous gen-
prietary information [49]. In addition, a generic model should erators, IEEE standard, CIM and PSS/E preferred models are
include external modules to be connected to the model, e.g. for indicated, whereas for renewable generation, generic models in
protection functions, and be valid for both strong and weak sys- PSS/E are stated. If none of these models allows to represent
tems. So, models should not behave erratically when the SCR adequately the dynamic performance, user-written models are
is low. A limit SCR value of 2 is cited in literature [49]. also accepted. In UK, the use of standard models is encoura-
According to [49], generic models should be described in- ged.
cluding connectivity diagrams (with links to external modules), However, generic models might be inadequate for studies
block diagrams used to represent the main components or other aiming to improve or assess equipment details [43]. Besides,
pertinent information, e.g., non-standard limit implementation. the validity of generic models for system frequency disturbances
The description of the models should be sufficient for program is still pending [45, 49]. Generic models have been tested against
developers to implement the models in positive sequence, large- vendor models, but should be compared to actual recorded dis-
scale, transient stability programs. Besides, the computation of turbance monitoring.
initial conditions must be described . Finally, model parame-
ters, test systems and operating conditions used for model eval-
uation need also to be provided.

6
3. Renewable power generation model validation necessary for generation assets connected to weak power grids
[14].
The purpose of model validation is to ensure the correct per- Currently, validation is mostly carried out by testing. Tests
formance of control systems and validate the computer models can be carried out on-site or off-site, via factory, laboratory or
used for stability analysis [60]. Consequently, model param- bed tests [68]. On-site or field tests are more realistic and, as a
eters must represent adequately the dynamic performance of consequence, they are generally preferred by system operators.
the device under modelling [45]. Along with the requirement In Australia model validation must be carried out with on-site
for simulation models, there is an increasing demand in the tests whenever possible. In Spain, the validation report shall
grid codes to assure the correspondance of models and physi- include results obtained under real tests, even if bench tests are
cal behaviour [61]. Table 1 indicates the documents containing accepted. In Ireland, both test types are accepted, because va-
validation conditions required by system operators, which are lidation is based on the comparison of simulation results with
strongly related to modelling requirements. Model validation is actual observed behaviour of a prototype or production WTG
compulsory in all the countries. under laboratory conditions and/or actual observed behaviour
There is considerable international experience regarding the of the real WTG as installed and connected to a transmission
validation of synchronous generator models [60, 62–65]. Va- or distribution network. In Germany, the use of bench tests is
lidation methods are mainly based on staged tests, involving permitted only if the behaviour of the generating unit is equal to
off-line tests such as enhanced short-circuit tests and partial the free-field measurement or when free-field testing equipment
load rejection tests [60], and frequency response testing cover- is used, as men tioned in standard IEC61400-21.
ing Standstill Frequency Response (SSFR), Open Circuit Fre- However, real tests might have some impact on the nearby
quency Response (OCFR) and On Line Frequency Response grid users, so significant cost and test-time reductions can be
(OLFR) described in [60, 66]. Both processes take significant obtained via factory testing [69]. Repeatability of the tests is
time and are expensive, and might be harmful for the generating also much better and, thus, it is much easier to identify any
units [65]. An on-line disturbance monitoring based method- equipment problem, if necessary.
ology is presented in [65], where validation is carried out by Model validation is required for different transient condi-
comparison with generator measurements recorded during sys- tions, depending on the country. Generally, dynamic models
tem events. Then, parameters can be optimised through an are validated by testing the generating unit performance under
automated iterative simulation approach. In addition, reactive faults. Thus, Low Voltage Ride-Through (LVRT) capability of
power capability, excitation system and governor testing is also the generation asset under test is also evaluated. In these cases,
extensively described in literature [60]. staged testing is a preferable option, with two factory imple-
There is less experience regarding validation with genera- mentation options [43, 70]:
tion based on renewable energy sources, and it is mainly fo-
cused on wind and photovoltaic systems. Validation must en- • Staged generator testing: it can be carried out using the
sure that models represent with sufficient accuracy the charac- turbine generator and controls alone, without the blades,
teristics of generating units, especially during severe transient during manufacturing process or at a dedicated test facil-
disturbances. ity.
Figure 1 illustrates the main steps that are necessary to eval-
• Staged full-scale testing: a full-scale turbine subjected to
uate the correspondence of a model with reality, based on a mo-
electrical disturbances at dedicated test facilities.
del definition [45, 61, 67]:

(1) Collect actual data from the devices under modelling, Factory test set-ups for short-circuit emulation are called
which can be recorded or measured during tests. voltage dip generators, which are able to emulate the actual
network impedance during a fault. Four types of emulators
(2) Simulate the same set of tests or events during the data
have been proposed in the literature [17–20, 69, 71]: genera-
collection process on the model.
tor based, shunt impedance based, transformer based, and full
(3) Compare measured or recorded data to simulation results converter based emulators. Full converters have the advantage
and decide if the validation is acceptable or the model to be able to emulate any voltage waveform [69], not only vol-
parameters have to be adjusted. tage dips but also active and reactive power steps, or voltage
Each of the steps will be further explained in next subsec- and frequency changes, as presented in [72]. Some commer-
tions. cial testing solutions have been patented such as the Megha for
LVRT tests or QuEST Lab, also able to carry out High Voltage
3.1. Test data collection Ride-Through (HVRT) or phase angle jump tests [73], which
are aimed for on-site tests.
In the early stages of the development of standard models,
The most common approach among the countries under re-
generic RMS models were validated against manufacturer de-
view is based on the shunt impedance based voltage dip gene-
tailed EMT models using a similar simulation tool [43]. It is an
rator as a voltage divider, proposed in standard IEC61400-21
easy method, but it has some shortcomings regarding the model
[74] (Figure 2). German and Danish validation tests are ba-
assumptions and the area of application. However, parallel vali-
sed on this standard. In the case of Germany, it can be used
dation of positive sequence RMS against EMT models could be
to validate both WTGs and photovoltaic generation units, for
7
 Model No

Measurement Test data Sim. Yes Successful

Match?
Record results

Figure 1: General validation process

WTG
PCC nitude, rate of change of frequency and step response test on
Z1 grid voltage angle. In Denmark, aside from voltage drop tests,
model behaviour under voltage increase and frequency varia-
tions must be tested and results shall be obtained from test sta-
Z2
tions or measurement at commercial systems. In Germany, dy-
namic generating unit models are validated for LVRT require-
ment and at power station level, focuses on active and reactive
power. Finally, the conditions validated in Ireland must be simi-
Figure 2: Low voltage ride-through test equipment principle based on voltage
divider
lar to those of interest, such as short-circuit levels, severe faults,
voltage excursions, or large wind variations.
Test data can also be collected by opportunistic testing, also
which the power at the DC side could be supplied by a suitable called on-line monitoring [43]. Thus, measurement equipment
source. It has the advantage that it can be easily constructed and installed on-site records naturally power system disturbances
the setup is quite similar to a real fault situation, and thus, a re- used to validate the simulation models. However, transient mo-
alistic voltage dip is obtained [69]. The impedance Z2 emulates del validation requires three-phase fault events, which rarely
the fault impedance. The voltage dip starts when the circuit occur, and in any case, they seldom present the sufficient mag-
breaker is closed, and ends when the breaker opens and clears nitude. Therefore, validation against single-phase short-circuits
the fault current. The impedance Z1 is needed in order to limit has been carried out in some cases [43, 47], as measurements
the influence of the voltage dip in the supplying grid. In order were considered valid and it was reported to be a successful ap-
to be able to set the remaining voltage level during the dip, ac- proach also for synchronous generators [65]. Otherwise, long-
cording the grid code requirements, the impedance values have term monitoring approach can be used, and progressive model
to be adjusted accordingly, or a shunt impedance bank can also validation can be performed [14].
be used [18]. In the German regulation, the impedances em-
ployed in the testing equipment must have an X/R ratio of at 3.2. Model simulation
least 3. The SCR at the connection point must be at least 3. Model acceptance processes are often based on play-back
LVRT tests must be performed for three-phase and two-phase techniques [45], where t he simulation model is fed with recorded
faults at partial and full load. The test is aimed both for model measurements from the low voltage terminals of the actual de-
validation and compliance verification for LVRT requirement. vice. Accordingly, the same set of tests or events can be re-
Nonetheless, other test benches other than a short-circuit simu- produced. In that case, the grid impedance and the dynamic
lator based on the voltage divider principle, are permitted: grid behaviour of the grid is not included in the model. This choice
simulators or transformer-based testing equipment. is called open-loop model validation. Play-back techniques are
For weak connection points, on-site staged fault testing is more commonly used with positive sequence models [14]. Ho-
usually more appropriate [14]. Several methods exist for the wever, open-loop validation could lead to error, if the correla-
implementation of on-site staged fault testing, including the tion between wind speed and voltage is not taken into account
fuse-wire method, the dropped conductor method and the di- [75] and if the non-linear characteristics of the step-up trans-
rect earthing method. A practical set-up and results of the latter former are omitted [14].
method are presented in [14]. An analogous technique is reported in [47], in use by WECC
However, other non-fault disturbance tests are also required for some time. It is achieved with the aid of a modified cla-
in Australia, Denmark, Republic of Ireland or UK. In Aus- ssical generator model (GENCLS) capable of holding terminal
tralia, general model acceptance tests are fault (three-phase-to- voltage and frequency as specified in an input file with actual
ground) and non-fault disturbance tests, such as step response records. However, for some simulation software packages, it is
test on machine active and reactive power, or grid voltage mag-

8
not possible to use an external file to establish a fictitious vol- specifications are indicated for control system models, time do-
tage. So, two alternatives are possible: main responses including non-linear responses or performance
and responses to switching or controlled sequence events.
• Simulation of a similar event by adjusting disturbance Even if most notable approaches to model validation are
conditions. based on measurements, validity decision based on expert engi-
• Application of a specific voltage profile using a user-written neering judgment is considered in some countries, such as the
model. United States [45].
If simulation and test results show a different behaviour
The latter option is adopted for wind farm simulation in from the generation asset, model parameter values shall be ad-
Spain, in the case of grid code compliance verification. On the justed or derived upon the observance of test response or through
other hand, closed-loop model validation entails a farm level internal measurements [42]. However, unlike the case of syn-
model being connected to the rest of the network [14]. The va- chronous machines, for power electronic interfaced technolo-
lidation of an entire WPP is reported in [45], using recorded gies it is not generally possible to derive most model param-
voltage and current at Point of Common Coupling (PCC) level. eters directly from time-domain analysis of on-site test results
However, according to the experience of system operators in nor from frequency domain transfer function testing [14]. In-
Australia, a number of high-speed data recorders at several lo- deed, parameter tuning based on curve fitting against on-site
cations are necessary, including the low and medium voltage testing measurements has to take into account a wide range of
terminals of critical nodes [14]. operating conditions.
In Australia, the general model acceptance tests required are Type validation is admitted in most regulations. In Australia
fault disturbance tests (three-phase-to-ground fault) and non- validation on a single generating unit would be sufficient, as the
fault disturbance tests, such as step response test on machine same performance is observed in other units of the same type.
active and reactive power, or grid voltage magnitude, rate of However, factory and/or type tests alone are not sufficient for
change of grid frequency, and step response test on grid vol- model validation, so a long-monitoring program must also be
tage angle. The same case studies are required for wind farms established [42].
and synchronous generators, although additional case studies
are stated for each one: regarding LVRT for variable generation
4. Verification of the compliance of technical requirements
technologies, and regarding excitation systems, and governors
for synchronous generators. Model acceptance tests are also in- Manufacturers and generators have to show compliance with
dicated for dynamic reactive support plants and HVDC links. the technical connection requirements included in the grid codes
Model acceptance set-up for wind farms is shown in Figure 3. under force. There are two alternatives to carry out this compli-
A similar set-up is proposed for synchronous generation, where ance verification [1]: practical tests or simulation, provided that
the connection arrangement of the generating unit is slightly the generating unit model has been validated. Each approach
different. Models are expected to work for a range of the si- has its own advantages and disadvantages, and the risk and costs
mulation parameters rather than for specific settings. Fault dis- of the both methods have to be assessed. Testing represents the
turbance tests must consider factors such as fault duration, grid real behaviour of the power system, but it has a high cost and
SCR, grid X/R ratio, pre-fault load levels, or fault X/R ratio. can have side effects on the system if it is performed on-site. On
For each test, a combination of study cases are presented based the contrary, simulation does not involve any additional charge
on varying aforementioned factors. and it is harmless. However, validation of the model against
field measurements is necessary. In general, compliance verifi-
3.3. Model validity acceptance rules cation is performed through a combination of both techniques,
A perfect match between the measured and simulated res- including on-site testing, comparative simulation studies, long-
ponse is not expected, but an adequate match that captures the term monitoring and based on external test experiences [14].
relevant dynamics and properly represents the dynamic response However, on-site testing is the primary approach expected to be
of the plant [45]. Few grid codes indicate the minimum accu- applied [14].
racy level. For instance, in Denmark, the accuracy of simu- In addition, the relative size of the generating system has
lations models must be kept within ±10% for voltage, active to be taken into consideration, as well as the type of generation
power, active current, reactive power and reactive current. The technology and the location at which requirements are expected
actual accuracy shall be documented in the validation report. to be met. Regarding generation systems based on renewable
In Australia, accuracy requirements are detailed for both load energy sources, power plants are made up of several small size
flow and short-circuit models, and transient and oscillatory sta- generating units, but grid code requirements are expected to be
bility models. Regarding load flow models, the deviation of met at the PCC. Therefore, depending on the technical rule un-
the plant model from the actual plant response for active power der study, verification shall be carried out on two levels [76]: a
and reactive power must not exceed 10% of the total change in single generating unit, and the entire plant level.
that quantity, and the model must not show characteristics that This section reviews current grid code compliance rules in
are not present in the actual plant response. These accuracy re- the countries under study, as well as corresponding certification
quirements apply also for transient stability models and further procedures in Germany and Spain.

9
 Generating Substation Grid and fault Infinite bus
unit
Z s = R s + jX s
X s /R s = 3.10
Y/D Ubus D/Y Upcc
(1 − d) · Z s d · Zs

WTG
Z f = R f + jX f
X f /R f = 3

Fault

Figure 3: Test circuit for model acceptance testing

4.1. Review of compliance verification practices fault level, partial load rejection, or responsiveness of governor
This section introduces general requirements and practices system. The response to voltage disturbance, frequency control,
regarding generating system performance verification in order impact on network capability and voltage and reactive power
to comply with grid codes under force. Reference documents control can also be demonstrated through model validation.
for the countries under study are summarised in Table 3.
Denmark. In Denmark, verification requirements depend on
Australia. During commissioning applicants must demonstrate the generating unit technology and size, analogously to grid
that their generating system meets the performance standards. code requirements. Regarding WPPs and PV plants plant own-
Wherever practicable, the performance must be demonstrated ers are responsible for ensuring that generating assets comply
by test. However, these tests cannot demonstrate that the per- with technical regulation and to provide some documentation
formance standards are met under all system conditions. More- in accordance with the total rated power of the power plant at
over, some requirements cannot be demonstrated by test. the point of connection. Even small installations above 11 kW
The rules do not detail any specific commissioning test. In- require type-approval and power quality verification according
stead, as technologies, types and the specific installation may to standard IEC 61400-21 [74], apart from other documents.
vary from site to site, tests are expected to be tailored to the re- In addition, WPPs with a power output above 1.5 MW need
quirements of the equipment. Nonetheless, in order to assist the to verify the capability of the power plant to remain connected
applicant, typical tests for synchronous and non-synchronous during voltage drops, and dynamic simulation is an acceptable
machines are outlined based on former practices. For some of verification method.
the rules more than a testing and monitoring method is pro- Regarding electrical conditions, tolerance to frequency and
posed, along the required testing frequency and the basis for voltage deviations, and power quality are listed for verification,
compliance assessment. and many items can be verified by using simulation models.
Some of the performance standards can be fully demons- For instance, under normal operation, conditions wind farms
trated based on on-site tests: power quality, protection system, must withstand phase-angle jumps without disconnecting and
active power control, monitoring and control requirements and the compliance can be verified by using a simulation model.
power station auxiliary supplies. On the other hand, reactive LVRT requirements, regarding balanced and unbalanced short-
power cannot be fully demonstrated on-site for the full voltage circuits, can also be assessed with simulation models. In any
range, as well as the response of the generation system to fre- case, set-ups for testing or simulation are not indicated.
quency, and voltage disturbances or contingencies which are
unlikely to be demonstrated on field. Only the limits of the Germany. In Germany, active and reactive power generation
protection system related to these requirements may be proved. depending on the primary power supply, active power control
Regarding frequency control, the actual performance of the sys- for defined set points and frequency deviations, power quality,
tem under frequency variations is unlikely to be demonstrated performance during faults (LVRT), cut-in conditions and the
on-site. With respect to voltage and reactive power control, the performance of protective devices must be verified.
actual performance of the generating system during all oscilla- The aim of the LVRT test is to determine whether the ge-
tions considering all system conditions is also unlikely to be de- nerating unit is capable of detecting voltage dips and riding
monstrated on-site. The performance of the system may be par- through them, as well as providing current during the voltage
tially demonstrated through model validation. Some of the rules dip. The voltage dip detection methodology must be described.
can be tested by modelling and simulation of the plant, such as Testing, measurement and verification of the unit performance

10
 Table 3: Grid code compliance verification in the countries under study

Country Title Technology

Commissioning requirements for generating systems [77]
Australia
Template for Generator Compliance Programs [78] Any technology
Technical regulation 3.2.5. for wind power plants with a power output
above 11 kW [29]
Denmark
Appendix 5.1. Wind power plants with a power output range of 1.5 MW
Wind power
to 25 MW [79]
Appendix 5.1. Wind power plants with a power output greater than 25
MW [80]
Technical regulation 3.2.2. for PV power plants with a power output
PV
above 11 kW [30]
FGW Technical Guidelines for Power Generating Units Part 3, Deter-
Germany mination of electrical characteristics of power generating units and sys- Any technology
tems connected to MV, HV and EHV grids [31]
Grid Code Compliance Test Procedure [81]
Republic of Ireland
EirGrid grid code [33] Wind power
New Zealand Companion Guide for Testing of Assets [82] Any technology
Procedure for verification, validation and certification of the require-
Spain ments of the PO 12.3 on the response of wind farms and photovoltaic Wind and PV power
plants in the event of voltage dips [83]
The Grid Code [40]
UK
Guidance Notes-Power Park Modules [41] Non-synchronous generation

can be carried out in accordance with standard IEC61400-21 and limits of the generating plant, and steady-state performance,
for any of the requirements. Test benches can be used for the including over-under frequency performance. Initial tests apply
following rules: fland reactive power provision based on pre- to all generators above 1 MW, but test types differ depending
defined set-points, transition behaviour of active and reactive on generation technologies. Detailed test programs for syn-
power provision based on predefined set-points, reduction of chronous generators and wind generators are indicated. Regar-
output power with overfrequency, PQ diagram and performance ding LVRT, the test entails applying a fault to the grid and mon-
during faults. itoring the wind farm response. The test must confirm that the
In some cases, if the safety of the unit is not guaranteed, co-ordinated control systems operate correctly and also allow
tests cannot be carried out with the unit running. In order to the validation of the model. In addition, this test must confirm
verify the voltage regulation requirement, the test can be carried that the wind farm stays connected during under frequency ex-
out on a test bench by means of a suitable grid or via adjustment cursions.
of the rated voltage in the control system. Flicker determination
can be carried out on-site, by using a test bench on the actual Republic of Ireland. The reference document encompasses ma-
grid or an AC network simulator. jor technical requirements for wind farms, including active power
Regarding the compliance at system level, only measure- management, transmission system voltage requirements, as well
ment of harmonic current is described in the reference docu- as signals, communication and control. In each section, a series
ment. of tests are defined, in order to be performed at wind farm level.
In Germany, the verification procedure is complemented by For each test the following items are described: purpose, instru-
a certification, as explained in Subsection 4.2. mentation, procedure and pass-criteria. All tests can be carried
out on-site without additional equipment, with exception of the
New Zealand. Generation assets must pass through testing at frequency response compliance test of wind farms. Since the
the commissioning stage. There is no compliance requirement grid frequency cannot be changed at will, this test requires to
in the connection rules, but an explanatory guide for asset test- be simulated by means of injection of a frequency signal into
ing has been released by the System Operator (Transpower) the wind farm controller to simulate appropriate changes of fre-
covering routine tests and commissioning tests. Routine tests quency.
are designed to ensure that the generators are able to meet the
Spain. Only verification regarding LVRT requirement included
technical requirements, as well as to verify operational ranges
in the Operation Procedure 12.3 is documented. The verifica-
11
tion procedure is explained in Subsection 4.2, in relation to the Certification is generally achieved in two steps. Firstly, a
certification procedure. type of generating unit will obtain a Type Certificate based on
one or more country specific grid codes according to the rele-
United Kingdom. Compliance processes for both synchronous vant certification procedure. A recognised certification system
generators and power park modules are included in the grid for WTGs is IEC 61400-22 [86], that includes the evaluation of
code document. Tests for the final operational notification must design, type-testing and manufacturing, as well as an optional
include reactive power capability tests, voltage control system type characteristic measurements module (power quality and
tests, governor or frequency control system tests and LVRT noise). Procedures for assessing compliance regarding power
tests for power plants above 100 MW. Reactive power capabi- quality requirements are gathered in IEC 61400-21 [74], inclu-
lity tests shall be performed by modifying the voltage set-point ding voltage quality (emissions of flicker and harmonics), vol-
of the voltage control. Voltage control system tests, that can tage drop response, power control (control of active and reactive
also be used to validate the excitation system or voltage sys- power), grid protection and reconnection time. The type certi-
tem model. The voltage control system shall be perturbed with fication process ends with the issuance of a certificate, main-
a series of step injections to the voltage reference and, where tained and verified over time.
possible, multiple up-stream transformer taps. And finally, fre- In a second step, a site specific Project Certificate will be
quency control system tests can also be used to validate the gov- issued for each power plant, based on site specific data and the
ernor or frequency control system model. Frequency modula- type certificate. In Europe, the most complete and documented
tion is possible by using a frequency injection signal. certification procedures regarding grid code verification are the
For each compliance test, the description, purpose, required Procedure for Verification, Validation and Certification of the
results and assessment criteria are given. However, measure- Requirements of the OP 12.3 on the Response of Wind Farms in
ment and acceptance requirements are not indicated. If a power the Event of Voltage Dips (PVVC) in Spain [83], and the Ger-
park contains two or more identical generating units, compli- man Technical Guidelines for Power Generating Units. Part 8.
ance testing may be reduced if the first unit completes the full Certification of the electrical characteristics of power genera-
testing. ting units and systems in the medium, high- and highest voltage
Regarding LVRT, manufacturers can demonstrate compli- grids [87], which describe the procedures to certify power ge-
ance using tests carried out with the facilities available. Ho- neration systems according their corresponding grid codes. A
wever, manufacturers are expected to replicate each fault type theoretical and practical comparison between both certification
(3-phase, phase-phase, two-phase to earth and single-phase to systems can be found in [4].
earth) with varying magnitudes. The tests should illustrate any
changes in characteristics or internal operating modes that de- 4.2.1. Germany: FGW-TG8 procedure
pend upon fault severity, such as active and reactive power fault The FGW-TG8 document [87] describes the procedure for
contribution and power recovery characteristic. The tests should the preparation and issue of a unit and system certification in
be performed on a single power park unit using a test circuit ba- accordance with the German grid connection regulations. The
sed on the voltage divider. scope of the guideline is limited to the electrical characteris-
Data and performance characteristics with respect to certain tics impacting load flow, grid stability and voltage quality in an
grid code requirements may be registered by manufacturer for electrical grid. The document is complemented by documents
specific non-synchronous generating units. It is called Man- [31] and [32].
ufacturer’s data and performance report, which covers LVRT As summarised by [4], applicants must provide the type
capability and the generating unit mathematical model. Simula- testing according to FGW-TG3 [31], proved by the test report
tion studies must be submitted to the system operator to demon- that includes measurement data, a comprehensive computer-
strate compliance with the connection conditions. The reactive based model of the power generating unit and an open model
power capability of the generator must be demonstrated by a of the power generating unit. The computer model may be a
load flow simulation study. On the other hand, voltage con- black box model and compatible with type tests in FGW-TG3
trol, reactive power stability and LVRT capability of power park in order to guide the verification of the model simulation ba-
modules shall be proved dynamic simulation series. sed on measuring results. Regarding the detail of the open mo-
del, it must be accorded beforehand between the certification
4.2. Certification procedures authority and the manufacturer, although in some cases, block
Showing compliance with grid codes is especially challen- diagrams may be adequate. However, in case of LVRT perfor-
ging regarding renewable power generation systems, and is best mance verification, fault detection must be described in detail.
done by compliance certification [84]. According to the inter- On the other hand, the computer-based model is validated by
national standard ISO/IEC 17000:2004, certification is a third- comparing simulation results to the test measurements. Simu-
party attestation related to products, processes, systems or per- lation runs must be carried out for a wide range of defined set-
sons, whereas attestation includes the issue of a statement, ba- points and grid conditions. Model validation is completed once
sed on a decision following the review, that fulfilment of spec- the open model has been checked.
ified requirements (e.g. guidelines, codes and standards) has The procedure also includes the procedure for generating
been demonstrated. The review itself covers the verification of system certification. To certify power generating systems the
the suitability, adequacy and effectiveness [85]. applicant must provide:
12
 • Details on all the units connected in the system, including dips: three-phase short-circuits at partial and full load, and two-
unit certificates, product certificates and/or test reports. phase isolated short-circuits at partial and full load. The defini-
tion and conditions in which the test will be carried out depend
• Details on the electrical components of the system, in- on the objective of the test: for model validation (general veri-
cluding the internal grid up to the connection point, and fication process) or observance of LVRT (particular verification
corresponding component certificates. process). In both cases, if the wind turbine is connected to a
• Details on the grid connection point, grid operator and weak point in the system, the voltage dip will be obtained with
connection regulation. The characteristics of the public no load, i.e. the generator disconnected from the dip generator.
grid, such as short-circuit power and impedance phase Model validation procedure. The generating unit model vali-
angle, will be provided by the system operator. dation consists of three steps:
4.2.2. Spain: PVVC procedure • Instantaneous voltage and current values are recorded for
The PVVC includes the verification, validation and certi- all the test categories. The duration of comparison win-
fication of wind farms, photovoltaic conversion systems and dow is a second, with 100 ms before the voltage dip.
FACTS. In the regulation under force, only wind farms are re-
• The manufacturers models must reproduce each of the
quested for LVRT and, accordingly, grid code verification is
tests carried out in the field. For that purpose, the test
focused on wind power. Two processes are possible to ver-
bench will be modelled as a voltage source set to the
ify the conformity with OP 12.3 for LVRT: the General Veri-
time series of the measured values, in the case of a WTG.
fication Process and the Particular Verification Process. The
For FACTS, the test bench can be modelled as a voltage
General Verification Process verifies that the generating unit re-
source and a current, so that they give exactly the same
mains connected during faults and that the requirements regar-
voltage and current as during the test. This is the afore-
ding active and reactive power are met, as stated on the pro-
mentioned playback technique. The integration step must
cedure. Three actions must be completed: testing, model vali-
be equal to or less than the time interval for the sampling
dation and, finally, wind farm simulation. Regarding test pro-
frequency of the measurements.
cedure, field tests are preferred, even though laboratory tests
could be accepted for FACTS. WTG simulation model validity • The results of the simulation model and the field tests
must be accredited by a model validation report confirmed by must match. The model will be considered validated when
measurements in the field tests. Next, the simulation models the difference between simulation and test for active and
of all dynamic elements of the wind farm have to be integrated reactive power does not exceed 10% in the 85% of the
inside a wind farm simulation model. Using this model, a wind cases.
farm simulation can be carried out evaluating its response. A
WTG with an accredited test report will constitute a unit type. Wind farm simulation procedure. In order to carry out a wind
farm simulation, the model of the wind farm will be based on
A generator of the same manufacturer and with the same char-
acteristics can avoid to have to repeat field tests. Wind farms validated WTG models. Existing reactive compensation de-
with a verified wind farms report will be considered a wind vices, cables, step-up transformer and internal lines will be also
farm type (i.e. project type). Figure 4 shows the three steps modelled. WTG aggregation is accepted. In addition, the inter-
of the general verification process, based on [4] . connection substations has also to be represented by a MV/HV
As an alternative to the general procedure, the particular transformer and the evacuation line until the PCC. The rest
verification process is based on a direct wind farm verification of the power system (the external grid) must be modelled so
by testing. As a consequence, model validation and wind farm that the fault clearance at the PCC reproduces the usual vol-
simulation can be avoided. tage profile in Spain: a sudden increase upon the clearing of
The particular verification process is faster and cheaper than the fault and a slower recovery afterwards. This profile will be
the general verification process [4]. Hence, manufacturers and considered fixed and independent of the location of the wind
wind farm operators may prefer this process. However, the gen- farm. The single line scheme is shown in Figure 5, based on
eral verification process is necessary whenever the wind tur- the reference document. The Union for the Coordination of the
bines are not capable of riding through the voltage dip (as de- Transmission of Electricity (UCTE) equivalent includes a syn-
fined in the particular process) and also if a compensating sys- chronous generator that reflects the UCTE system. To take into
tem is installed at the wind farm substation. account the dynamics of the closest grid, a synchronous gene-
rator is included, as well as a load, modelled as constant current
Test procedure. As testing equipment, the use of a voltage dip and constant impedance asset. Data of the synchronous gener-
generator using an inductive divider is recommended. Other ators and their excitation systems are indicated. The fault reac-
types of dip generators are accepted, but resulting residual volt- tance is adjusted so as to have a voltage magnitude of 0.2 p.u.
ages must be similar to those defined in the document. Four and 0.6 p.u. during respectively a three-phase and a two-phase
test categories are defined from the combination of partial and short-circuit. Parameter values can be found in the reference
full load operating point, and three phase and two phase voltage document. Voltage profile is shown in Figure 6. All sequence
impedances have been considered equal, as there is no specific
indication in the document.
13
 Field Test data Test report
test

Simulation Results Yes Validation report

Match?

Model Sim. Wind farm report

Figure 4: General Verification Process

UCTE bus Grid bus PCC bus

Fault

Figure 5: PVVC: model of the equivalent electrical grid

14
 dip must meet specified requirements and the residual stress
1 level and time during the load test must be as indicated. The
voltage dip profile depends on whether the system is connected
Voltage(p.u.)

to a strong or weak PCC.

0.5
5. Conclusions
3PhG
The increase in renewable generation plants formed by a
2Ph
large number of individual generating units poses a challenge to
0
0 2 4 6 8 10 system operators, in terms of connection process and plant mo-
delling management. In order to cope with these issues, com-
Time(s)
pliance procedures based on testing and comparative simulation
Figure 6: Voltage at the PCC bus during faults studies have already been established. Nonetheless, grid code
conformity verification for RES remains an open issue: in the
regulatory frameworks, procedures are still at an early stage and
For each of the four test categories, it must be proved that: literature about the subject is still sparse, scattered or focused
on a single country. Therefore, this paper has reviewed and
• The wind farm remains connected during the voltage dip.
analysed current procedures and usual practices, including sys-
Therefore, the simulation model must include the protec-
tem operators in Australia, Denmark, Germany, New Zealand,
tion relays. If the wind farm model is not based on unit
Republic of Ireland, Spain and United Kingdom.
aggregation, the loss of generated active power must not
In the regulatory framework under study, positive-sequence
exceed 5% of the pre-fault value.
RMS models of renewable power plants must be provided, even
• Voltage and current levels at WTG terminals must be for low capacity installations. EMT-type models are also nec-
compared to test values and differences must be within essary for the study of small signal phenomena or under weak
admitted error tolerances. grid conditions. Model aggregation is nowadays widely ac-
cepted, whereas black-box models are not admitted in some
• Exchanges of active and reactive power must be performed countries. The preferred option as compulsory simulation soft-
as described in the technical requirement. Measurement ware is PSS/E from Siemens.
techniques and power calculating methodologies for test- Along with the specifications for simulation models, man-
ing and simulation are indicated in the reference docu- ufacturers must assure the correlation of models with the phys-
ment. ical behaviour of real power plants or generating units. There-
fore, validated models are necessary. Generally, dynamic mo-
For existing wind farms, simplified WTG models can be
dels are validated by the generating unit performance under grid
used, as usually no data to model the installations exists. If the
faults. Thus, LVRT capability can be also assessed. The most
wind turbines have an accredited test report, general library mo-
common approach is the use of a shunt-impedance dip gene-
dels can be included in the wind farm simulation. The models
rator, although other test benches are also permitted. Neverthe-
will consist of a current/voltage source and protections, either
less, for weak connection points, on-site staged fault testing can
so as to meet the limits of the report. If the requirements to use
be more adequate.
library models are not fulfilled, validated models of WTGs or
Comparison of compliance verification practices is a de-
FACTS must be provided by the manufacturers and the valida-
manding task, since procedures are related to heterogeneous
tion must be carried out according to the PVVC.
grid codes, where the detail degree, structure and even termi-
Verification of requirements for photovoltaic plants. Annex II nology are discordant. According to the study presented in this
of the reference document indicates the verification of the re- paper, conformity specifications are largely vague and general
quirements in P.O. 12.3 for photovoltaic plants. The testing in regulation. Procedures, instrumentation and pass-criteria of-
process is based on feeding the AC side of the photovoltaic ten are not explicit nor clear. Generally, conformity to techni-
conversion system by a system that simulates voltage dips. The cal rules must be demonstrated by compliance tests, although in
conversion system will be tested by a continuous source at its Denmark and Australia some provisions can be proved through
DC part, either consisting of a set of photovoltaic modules or a simulation. In Spain, certification procedures for wind farms
DC power supply. Regarding the voltage dip generator, it can be accept simulation as a means to verify compliance of LVRT
an inductive generator similar to the set-up proposed for WTGs clause. The performance of renewable power plants under fault
and FACTS, or a power electronics device or other device able must be verified in all the countries under study, whether for
to simulate a variable AC voltage with the profile defined in the model validation or compliance verification.
technical requirement. It can be concluded from this study that grid code verifi-
Regarding test validation criteria, active power must be wi- cation, model validation and certification procedures are still
thin indicated ranges, the system disconnection must be less under development in many countries, contrastingly to rather
than one in three consecutive tests, injected current during the stable, harmonised and complete grid codes. Each regulation

15
provision should be complemented by compliance tests inclu- [7] T. Ackermann, A. Ellis, J. Fortmann, J. Matevosyan, E. Muljadi, R. Pi-
ding description, purpose, testing set-up, required results, and wko, P. Pourbeik, E. Quitmann, P. Sorensen, H. Urdal, B. Zavadil, Code
Shift: Grid Specifications and Dynamic Wind Turbine Models, IEEE
assessment criteria. Thus, compliance verification should not Power and Energy Magazine 11 (2013) 72–82.
leave room for interpretation. However, even if compliance pro- [8] C. Sourkounis, P. Tourou, Grid Code Requirements for Wind Power Inte-
cesses are well documented in some countries like UK, mostly gration in Europe, in: Conference Papers in Energy, 2013. doi:10.1155/
there is usually little information and documents are not clear 2013/437674.
[9] F. Dı́az-González, M. Hau, A. Sumper, O. Gomis-Bellmunt, Participation
nor definite. of wind power plants in system frequency control: Review of grid code
Recently, RES certification have been launched in some requirements and control methods, Renewable and Sustainable Energy
countries, such as Germany and Spain. The German regula- Reviews 34 (2014) 551–564.
tion encompasses generating unit and system certification. The [10] A. Etxegarai, P. Eguia, E. Torres, A. Iturregi, V. Valverde, Review of
grid connection requirements for generation assets in weak power grids,
process includes type testing verification, as well as the provi- Renewable and Sustainable Energy Reviews 41 (2015) 1501–1514.
sion of simulation models of the assets under certification. On [11] P. E. Sutherland, Ensuring Stable Operation with Grid Codes: A Look
the other hand, Spanish procedure is based on requirement OP at Canadian Wind Farm Interconnections, IEEE Industry Applications
12.3. Thus, wind farm certification is based on the verification Magazine 22 (2016) 60–67.
[12] E. M. G. Rodrigues, G. J. Osório, R. Godina, A. W. Bizuayehu, J. M.
of LVRT requirement based on simulation models of the whole Lujano-Rojas, J. P. S. Catalão, Grid code reinforcements for deeper rene-
wind farm, which are made up by previously validated wind wable generation in insular energy systems, Renewable and Sustainable
turbine models. Energy Reviews 53 (2016) 163–177.
[13] A. Cabrera-Tobar, E. Bullich-Massagué, M. Aragüés-Peñalba, O. Gomis-
However, simulation procedures are scarcely documented Bellmunt, Review of advanced grid requirements for the integration of
and basically focused on LVRT requirement, since fault dis- large scale photovoltaic power plants in the transmission system, Rene-
turbances are also used for model validation. On the other wable and Sustainable Energy Reviews 62 (2016) 971–987.
hand, modelling requirements are more widely standardised, [14] B. Badrzadeh, A. Halley, Challenges associated with assessment and
testing of fault ride-through compliance of variable power generation in
although likely not adequate for weaker connection points and Australian national electricity market, IEEE Transactions on Sustainable
EMT studies. The development and use of adequate generic Energy PP (2014) 1–9.
non-synchronous generator models would certainly smooth the [15] D. Popovic, I. Wallace, International Review of Fault Ride Through for
way to straight-forward certification procedures. Conventional Generators, Technical Report 16010829, Kema, 2010.
[16] F. Kalverkamp, B. S.-v. d. Brelie, T. D. Nguyen, T. Mertens, M. Meuser,
Comparative analysis of European Grid Codes and compliance standards
for distributed power generation plants with respect to future require-
Acknowledgement ments of ENTSO-E and CENELEC, in: International ETG Congress
2015; Die Energiewende - Blueprints for the new energy age; Proceed-
The authors acknowledge the financial support from the Uni- ings of, 2015, pp. 1–6.
versity of the Basque Country UPV/EHU (projects EHU 13/66, [17] C. Veganzones, J. A. Sanchez, S. Martinez, C. A. Platero, F. Blazquez,
EHUA15/25 and UFI 11/28), as well as the Basque Government D. Ramirez, J. R. Arribas, J. Merino, N. Herrero, F. Gordillo, Voltage
dip generator for testing wind turbines connected to electrical networks,
(ELKARTEK15/42). Renewable Energy 36 (2011) 1588–1594.
[18] Y. Yang, F. Blaabjerg, Z. Zou, Benchmarking of voltage sag generators,
in: IECON 2012 - 38th Annual Conference on IEEE Industrial Electron-
References ics Society, 2012, pp. 943–948. doi:10.1109/IECON.2012.6389164.
[19] N. Espinoza, M. Bongiorno, O. Carlson, Novel LVRT Testing Method for
[1] ENTSO-E, ENTSO-E Network Code for Requirements for Grid Connec- Wind Turbines Using Flexible VSC Technology, IEEE Transactions on
tion Applicable to all Generators, 2016. Sustainable Energy 6 (2015) 1140–1149.
[2] M. Tsili, S. Papathanassiou, A review of grid code technical requirements [20] D. Xiang, J. Turu, S. Muratel, T. Wang, On-site LVRT Testing Method for
for wind farms, IET Renewable Power Generation 3 (2009) 308–332. Full Power Converter Wind Turbines, IEEE Transactions on Sustainable
[3] M. Altin, O. Goksu, R. Teodorescu, P. Rodriguez, B.-B. Jensen, L. Helle, Energy PP (2016) 1–1.
Overview of recent grid codes for wind power integration, in: Optimiza- [21] Lars Lindgren, Jörgen Svensson, Lars Gertmar, Generic models for Wind
tion of Electrical and Electronic Equipment (OPTIM), 2010 12th Interna- Power Plants. Needs and previous work, Technical Report Elforsk rapport
tional Conference on, 2010, pp. 1152–1160. doi:10.1109/OPTIM.2010. 12:47, Elforsk, Stockholm, 2012.
5510521. [22] Phil Southwell, Zoran Bozic, Results of survey of requirements on ge-
[4] M. P. Comech, M. Garcia-Gracia, S. Martin Arroyo, M. A. Mar- nerator data and the need for confidentiality, Cigre Electra magazine 224
tinez Guilen, Wind farms and grid codes, in: G. Krause (Ed.), From Tur- (2006) 44–45.
bine to Wind Farms - Technical Requirements and Spin-Off Products, In- [23] E. Energy, AEMO, Wind Integration: International Experience. WP2:
Tech, 2011, pp. 17–41. URL: http://www.intechopen.com/books/ Review of Grid Codes, Technical Report, Ecar Energy and AEMO, 2011.
from-turbine-to-wind-farms-technical-requirements-and- [24] F. Kalvercamp, M. Meuser, B. Schowe-von der Brelie, J. Stueken, Wind
spin-off-products/wind-farms-and-grid-codes. farm simulation and certification of its grid code compliance - a re-
[5] M. Gustavo, J. Gimenez, Technical and regulatory exigencies for grid view of lessons learned from the German experiences, in: LSI Work-
connection of wind generation, in: G. O. Suvire (Ed.), Wind Farm shop 2011, 2011. URL: http://www.fgh-gmbh.com/cms/upload/
- Technical Regulations, Potential Estimation and Siting Assessment, qrcode/Paper_LSIWorkshop2011_FGH_WIW11-180_Final.pdf.
InTech, 2011. URL: http://www.intechopen.com/books/wind- [25] Australian Energy Market Commission AEMC, The National Elec-
farm-technical-regulations-potential-estimation-and- tricity Rules, 2016. URL: http://www.aemc.gov.au/Electricity/
siting-assessment/technical-and-regulatory-exigencies- National-Electricity-Rules/Current-Rules.html.
for-grid-connection-of-wind-generation. [26] Generating System Model Guidelines, Technical Report, Australian En-
[6] M. Mohseni, S. M. Islam, Review of international grid codes for wind ergy Market Operator AEMO, 2008.
power integration: Diversity, technology and a case for global standard,
[27] Data and Model Requirements for generating systems of less than 30 MW,
Renewable and Sustainable Energy Reviews 16 (2012) 3876 – 3890. Technical Report, Australian Energy Market Operator AEMO, 2013.
[28] Dynamic Model Acceptance Guideline, Technical Report, Australian En-

16
 ergy Market Operator AEMO, 2013. tus report, in: Power Systems Conference and Exposition (PSCE), 2011
[29] Energinet.dk, Technical regulation 3.2.5. for wind power plants IEEE/PES, 2011, pp. 1–8. doi:10.1109/PSCE.2011.5772473.
with a power output greater than 11 kW, 2010. URL: http: [50] P. Pourbeik, A. Ellis, J. Sanchez-Gasca, Y. Kazachkov, E. Muljadi,
//www.energinet.dk/SiteCollectionDocuments/Engelske% J. Senthil, D. Davies, Generic stability models for type 3 and 4 wind tur-
20dokumenter/El/55986-10_v1_Grid%20Code%203%202%205_v% bine generators for WECC, in: 2013 IEEE Power Energy Society General
204%201-30%20%20September%202010.pdf. Meeting, 2013, pp. 1–5. doi:10.1109/PESMG.2013.6672398.
[30] Energinet.dk, Technical regulation 3.2.2 for PV power plants with a [51] J. J. Sanchez-Gasca, Generic wind turbine generator models for WECC
power output above 11 kw, 2015. URL: https://www.energinet.dk/ - a second status report, in: 2015 IEEE Power Energy Society General
SiteCollectionDocuments/Engelske%20dokumenter/El/14- Meeting, 2015, pp. 1–5. doi:10.1109/PESGM.2015.7285645.
17997-39%20Technical%20regulation%203.2.2%20for%20PV% [52] WECC Renewable Energy Modeling Task Force, Generic So-
20power%20plants%20with%20a%20power%20output%20above% lar Photovoltaic System Dynamic Simulation Model Specification,
2011%20kW,%20Rev.%201.pdf. Technical Report Sandia Contract #1047506, Western Electric-
[31] FGW Fördergesellschaft Windenergie und andere Erneuerbare Energien, ity Coordinating Council Modeling and Validation Work Group,
FGW technical guidelines for power generating units part 3, determina- 2012. URL: http://www.powerworld.com/files/WECC-Solar-
tion of electrical characteristics of power generating units and systems PV-Dynamic-Model-Specification-September-2012.pdf.
connected to mv, hv and ehv grids, 2013. URL: http://www.wind- [53] R. T. Elliott, A. Ellis, P. Pourbeik, J. J. Sanchez-Gasca, J. Senthil, J. We-
fgw.de/. ber, Generic photovoltaic system models for WECC - A status report,
[32] FGW Fördergesellschaft Windenergie und andere Erneuerbare Energien, in: 2015 IEEE Power Energy Society General Meeting, 2015, pp. 1–5.
FGW technical guidelines for power generating units part 4, demands doi:10.1109/PESGM.2015.7285992.
on modelling and validating simulation models of the electrical char- [54] S. I. Nanou, A. G. Papakonstantinou, S. A. Papathanassiou, A generic
acteristics of power generating units and systems, 2013. URL: http: model of two-stage grid-connected PV systems with primary frequency
//www.wind-fgw.de/. response and inertia emulation, Electric Power Systems Research 127
[33] EirGrid, EirGrid Grid Code, Technical Report v. 6.0., EirGrid, 2015. (2015) 186–196.
[34] Connecting and Dispatching New Generation in New Zealand. Overview, [55] M. Patsalides, V. Efthymiou, A. Stavrou, G. E. Georghiou, A generic
Technical Report, Transpower New Zealand Ltd, 2007. transient PV system model for power quality studies, Renewable Energy
[35] T. N. Z. Ltd., Asset Capability Information Overview. Guideline, Techni- 89 (2016) 526–542.
cal Report Version 4.1., Transpower New Zealand Ltd., 2010. [56] P. Pourbeik, D. Sullivan, A. Bostrom, J. Sanchez-Gasca, Y. Kazachkov,
[36] Boletı́n Oficial del Estado, Resolución de 18 de diciembre, 2015. URL: J. Kowalski, A. Salazar, A. Meyer, R. Lau, D. Davies, E. Allen, Generic
http://www.ree.es/sites/default/files/01_ACTIVIDADES/ model structures for simulating static var systems in power system studies
Documentos/ProcedimientosOperacion/RES_VAR_20151218_ - a WECC task force effort, IEEE Transactions on Power Systems 27
Participacion_en_servicios_de_ajuste_y_aprobacion_POs. (2012) 1618–1627.
pdf. [57] S. Cole, R. Belmans, A proposal for standard VSC HVDC dynamic mo-
[37] REE, Guı́a descriptiva del procedimiento de puesta en servicio, Technical dels in power system stability studies, Electric Power Systems Research
Report Version 1.0, REE, 2011. URL: http://www.ree.es. 81 (2011) 967–973.
[38] REE, Requisitos de los modelos de instalaciones eólicas, fotovoltaicas y [58] C. Hahn, A. Semerow, M. Luther, O. Ruhle, Generic modeling of
todas aquellas que no utilicen generadores sı́ncronos directamente conec- a line commutated HVDC system for power system stability studies,
tados a la red, Technical Report, 2010. in: T D Conference and Exposition, 2014 IEEE PES, 2014, pp. 1–6.
[39] REE, Condiciones de validación y aceptación de los modelos, Technical doi:10.1109/TDC.2014.6863308.
Report, REE, 2010. URL: http://www.ree.es. [59] P. Sorensen, B. Andresen, J. Fortmann, P. Pourbeik, Modular structure
[40] N. G. E. Transmission, The grid code, 2016. of wind turbine models in IEC 61400-27-1, in: 2013 IEEE Power and
[41] Guidance notes. Power Park Modules, Technical Report, National Grid, Energy Society General Meeting (PES), 2013, pp. 1–5. doi:10.1109/
2013. PESMG.2013.6672279.
[42] P. Ravalli, J. Leung, Dynamic model requirements and model valida- [60] L. Hajagos, J. Barton, R. Berube, M. Coultes, J. Feltes, G. Lanier, S. Pat-
tion in the Australian National Electricity Market, in: 2011 IEEE Power terson, L. Pereira, P. Pourbeik, A. Schneider, R. Jones, Guidelines for
and Energy Society General Meeting, 2011, pp. 1–5. doi:10.1109/PES. generator stability model validation testing, in: IEEE Power Engineer-
2011.6039169. ing Society General Meeting, 2007, 2007, pp. 1–16. doi:10.1109/PES.
[43] S. Zhao, N.-K. Nair, Assessment of wind farm models from a transmis- 2007.386134.
sion system operator perspective using field measurements, IET Renewa- [61] J. Fortmann, S. Engelhardt, J. Kretschmann, C. Feltes, I. Erlich, Valida-
ble Power Generation 5 (2011) 455–464. tion of an RMS DFIG simulation model according to new german model
[44] J. Feltes, B. Fernandes, Wind turbine generator dynamic performance validation standard FGW TR4 at balanced and unbalanced grid faults,
with weak transmission grids, in: 2012 IEEE Power and Energy Society Proceedings / 8th International Workshop on the Large-Scale Integration
General Meeting, 2012, pp. 1–7. doi:10.1109/PESGM.2012.6345675. of Wind Power into Power Systems as well as on Transmission Networks
[45] M. Asmine, J. Brochu, J. Fortmann, R. Gagnon, Y. Kazachkov, C.-E. for Offshore Wind Wind Power Farms : 14-15 October 2009, Bremen,
Langlois, C. Larose, E. Muljadi, J. MacDowell, P. Pourbeik, S. Seman, Germany (2009).
K. Wiens, Model validation for wind turbine generator models, in: 2011 [62] L. Pereira, Introduction and background to synchronous unit testing and
IEEE Power and Energy Society General Meeting, 2011, pp. 1–1. doi:10. model validation in the WSCC, in: IEEE Power Engineering Society
1109/PES.2011.6038947. 1999 Winter Meeting, volume 1, 1999, pp. 151–156 vol.1. doi:10.1109/
[46] E. Muljadi, S. Pasupulati, A. Ellis, D. Kosterov, Method of equivalencing PESW.1999.747441.
for a large wind power plant with multiple turbine representation, in: [63] R. Rifaat, Independent power producers (IPP) perspectives and expe-
2008 IEEE Power and Energy Society General Meeting - Conversion and riences with WSCC requirements for generator model validation tests,
Delivery of Electrical Energy in the 21st Century, 2008, pp. 1–9. doi:10. in: Conference Record of the 2000 IEEE Industry Applications Con-
1109/PES.2008.4596055. ference, 2000, volume 2, 2000, pp. 924–931 vol.2. doi:10.1109/IAS.
[47] E. Muljadi, A. Ellis, Validation of wind power plant models, in: 2000.881942.
2008 IEEE Power and Energy Society General Meeting - Conversion [64] L. Hajagos, Guidelines for generator stability model validation testing,
and Delivery of Electrical Energy in the 21st Century, 2008, pp. 1–7. in: IEEE Power Engineering Society General Meeting, 2003, volume 3,
doi:10.1109/PES.2008.4596053. 2003, pp. –1294 Vol. 3. doi:10.1109/PES.2003.1267335.
[48] T. Gehlhaar, T. Wehrend, H. Lankowski, D. Schulz, Need of harmonized [65] P. Pourbeik, Automated parameter derivation for power plant models
generic model standards for flexible grids in a smart future, in: Proceed- from system disturbance data, in: IEEE Power Energy Society Gen-
ings, Aarhus, Denmark, 2011. eral Meeting, 2009. PES ’09, 2009, pp. 1–10. doi:10.1109/PES.2009.
[49] A. Ellis, Y. Kazachkov, E. Muljadi, P. Pourbeik, J. Sanchez-Gasca, De- 5275649.
scription and technical specifications for generic WTG models - a sta- [66] P. Dandeno, H. Karmaker, C. Azuaje, M. Glinkowski, I. Kamwa, S. Salon,

17
 R. Saunders, S. Umans, Experience with standstill frequency response [87] FGW Fördergesellschaft Windenergie und andere Erneuerbare Energien,
(SSFR) testing and analysis of salient pole synchronous machines, IEEE FGW technical guidelines for power generating units part 8, certification
Transactions on Energy Conversion 14 (1999) 1209–1217. of the electrical characteristics of power generating units and systems in
[67] J. Fortmann, L. Cai, S. Engelhardt, J. Kretschmann, Wind turbine mod- the medium-, high- and highest-voltage grids, 2013. URL: http://www.
eling, LVRT field test and certification, in: 2011 IEEE Power and En- wind-fgw.de/.
ergy Society General Meeting, 2011, pp. 1–7. doi:10.1109/PES.2011.
6039567.
[68] J. Niiranen, S. Seman, J. Matsinen, R. Virtanen, A. Vilhunen, Technical
Paper: Low voltage ride-through testing of wind turbine converters at
ABB helps wind turbines meet the requirements of IEC61400-21 more
quickly, Technical Report, ABB, 2013.
[69] J. Niiranen, Experiences on voltage dip ride through factory testing of
synchronous and doubly fed generator drives, in: 2005 European Con-
ference on Power Electronics and Applications, 2005, pp. 11 pp.–P.11.
doi:10.1109/EPE.2005.219725.
[70] Y. Coughlan, P. Smith, A. Mullane, M. O’Malley, Wind turbine modelling
for power system stability analysis-a system operator perspective, IEEE
Transactions on Power Systems 22 (2007) 929–936.
[71] C. Wessels, R. Lohde, F. Fuchs, Transformer based voltage sag generator
to perform lvrt and hvrt tests in the laboratory, in: Power Electronics and
Motion Control Conference (EPE/PEMC), 2010 14th International, 2010,
pp. T11–8–T11–13. doi:10.1109/EPEPEMC.2010.5606830.
[72] Abdullah Al Mahfazur Rahman, Muhammad Usman Sabbir, Grid Code
Testing by Voltage Source Converter, Master’s thesis in the master degree
programme, electric power engineering, Chalmers University of Technol-
ogy, Göteborg, Sweden, 2012.
[73] For Optimal Renewable Energy Systems S.L. 4FORES, 2014. URL:
http://www.4fores.es.
[74] IEC-61400-21, Wind turbines, part 21: Measurement and assessment of
power quality characteristics of grid connected wind turbines, 2008.
[75] Jan Pierik, Johan Morren, Tim van Engelen, Sjoerd de Haan, Jan Bozelie,
Development and validation of wind farm models for power system stud-
ies. alsvik wind farm results, in: Conference Proceedings, Athens,
Greece, 2006. URL: http://www.ewea.org/ewec2006/allfiles2/
956_Ewec2006fullpaper.pdf.
[76] T. Gehlhaar, Grid code compliance beyond simple lvrt, in: Germanischer
Lloyd, 2012.
[77] Commissioning requirements for generating systems, Technical Report,
Australian Energy Market Operator AEMO, 2012.
[78] Template for Generator Compliance Programs, Technical Report, Relia-
bility Panel AEMC, 2009.
[79] Energinet, Appendix 5.1. Wind power plants with a power out-
put range of 1.5 MW to 25 MW. Technical regulation for grid
connection TF 3.2.5, Technical Report Version 4.1, 2010. URL:
http://www.energinet.dk/EN/El/Forskrifter/Technical-
regulations/Sider/Regulations-for-grid-connection.aspx.
[80] Energinet, Appendix 5.2. Wind power plants greater than 25 MW.
Technical regulation for grid connection TF 3.2.5, Technical Report
Version 4.1, 2010. URL: http://www.energinet.dk/EN/El/
Forskrifter/Technical-regulations/Sider/Regulations-
for-grid-connection.aspx.
[81] EirGrid, Grid Code Compliance Test Procedure, Technical Report, Eir-
Grid, 2010.
[82] Companion Guide for Testing of Assets, Technical Report, Transpower,
2012.
[83] REE, Procedure for verification, validation and certification of the re-
quirements of P.O.12.3 on the response of wind farms and photo-
voltaic plants in the event of voltage dips, Technical Report Version
9, 2011. URL: http://www.aeeolica.org/uploads/documents/
1306-pvvc-n9-english.pdf.
[84] Aniruddha Das, Matthias-Klaus Schwarz, Grid connectivity issues and
the importance of GCC, 2013.
[85] Mike Woebbeking, IEC TS 61400-22 (First Revision of IEC WT 01) The
new standard for Wind Turbines and Wind Farms – Onshore and Off-
shore, Technical Report, Germanischer Lloyd Industrial Services GmbH,
Business Segment Wind Energy (GL), 2008. URL: http://www.gl-
group.com/pdf/IEC_TS_61400-22_Woeb.pdf.
[86] IEC International Electrotechnical Commission, IEC 61400-22 ed1.0.
wind turbines - part 22: Conformity testing and certification,
2010. URL: http://webstore.iec.ch/webstore/webstore.nsf/
artnum/044160!opendocument.

18

View publication stats