Structured Planning of Hardware and Software Co-simulation Testing of

Smart Grids

Rami Elshinawy

, Rebeca P. Ram

ırez Acosta

, Jan S

oren Schwarz

and Sebastian Lehnhoff

OFFIS Institute, Escherweg 2, Oldenburg, Germany

Keywords:

Co-simulation, HIL, HTD, Information Model, SGAM, Simulation, Simulation Planning, Smart Grid.

Abstract:

The traditional electricity grid is ought to become a smart grid. One reason is the integration of volatile

renewable energy generation, which poses the demand for advanced communication and control technology.

This results in a change of the overall system dynamics, which in turn require testing to be done in early stages

and with a more holistic approach. However, there is a gap in the transition from system speciﬁcation and static

analysis to an experimental setup validating these speciﬁcation. This paper demonstrates ﬁrst steps towards a

structured methodology for deriving validation experiments for smart grids from initial system requirements.

Our approach aims to ﬁll the introduced gap by integrating Smart Grid Architecture Model (SGAM) for static

analysis, Holistic Testing Description (HTD) for experimental analysis, and additional tools in a new workﬂow.

Initial assessment of the workﬂow has been validated on its ability to select the most suitable components and

test beds to perform an experiment. Therefore, two case studies were selected, one in the electricity power

domain and another in the electricity market domain.

1 INTRODUCTION

The new power grid consists not only of physical

electrical equipment (e.g., transformer, cables, etc.),

but also Information and Communication Technol-

ogy (ICT) and the automation of these equipment

adds new domains and dimension to the complexity

of this system. The main reason behind the integra-

tion of these technologies is the introduction of decen-

tralised generation resources and renewable energy in

the efforts to dramatically reduce energy sector emis-

sion levels. These new concepts of energy generation

require constant monitoring and prediction, as they

are intermittent, weather dependent, and have limited

storage capacity. Due to their ﬂexibility and the dy-

namic of the grid, communication and control sys-

tems as well as new electricity market options need

to be tested at earlier stages.

This evolution transforms the traditional electri-

cal grid to a smart grid. This grid has physical de-

vices transferring power and software administrating

the interaction between them. The emerging Cyber

https://orcid.org/0000-0002-2803-2499

https://orcid.org/0000-0002-9876-8338

https://orcid.org/0000-0003-0261-4412

https://orcid.org/0000-0003-2340-6807

Physical Energy System (CPES) in hand comprises

of different disciplines and mixed technologies. A

complex conﬁguration, as described, demands testing

the integration of its constituents on a system level,

not only testing of certain aspects of it, addressing all

relevant domains. Moreover, the testing procedure is

required to follow a multi stage process alongside de-

velopment, and before roll out (Van Der Meer et al.,

2017). A well established approach for the analysis of

smart grids and the development of its components is

the use of simulations (Hartmann, 2009). These simu-

lations may be purely software based (Schloegl et al.,

2015) or can contain a real hardware setup (Nguyen

et al., 2017).

Despite the readily available tools for testing and

validation, developing a test for this complex system

is an issue of forming a clear test objective, besides a

speciﬁc and relevant multi domain test environment

(Blank et al., 2016). Test standards usually devel-

oped within speciﬁc context of a scientiﬁc or techni-

cal application. For example, organisations within au-

tomotive, thermal systems or electric power domains

each identify and maintain their speciﬁc standards,

test requirements, protocols, and test environments

(Heussen et al., 2019). However, multi domain testing

requires integration of a more fragmented knowledge,

which has been addressed in the context of the Eu-

Elshinawy, R., Acosta, R., Schwarz, J. and Lehnhoff, S.

Structured Planning of Hardware and Software Co-simulation Testing of Smart Grids.

DOI: 10.5220/0009820701970208

In Proceedings of the 10th International Conference on Simulation and Modeling Methodologies, Technologies and Applications (SIMULTECH 2020), pages 197-208

ISBN: 978-989-758-444-2

197

ropean ERIGrid project (M

aki et al., 2016; Heussen

et al., 2019; Van Der Meer et al., 2017). Thus, the dis-

ciplinary and methodological framing of experiments

is becoming a challenge itself, especially when form-

ing a ”holistic” test objective and designing experi-

ments for this objective.

The aim of this paper is to propose a workﬂow

for simulation experiments that reduces the gap in the

transition between speciﬁcation and static analysis of

a complex system on one hand, and operational test-

ing on the other hand (see section 3). This approach is

built upon integrating wide range of tools and knowl-

edge developed in the literature, taking into consider-

ation the increased number of stakeholders involved

and their business opportunities (see section 2). In

section 4, two case studies are presented for evalua-

tion of the workﬂow on experiments applied in dif-

ferent domains of the smart grid. Finally, the con-

tribution is concluded and ideas for future work are

elaborated in section 5.

2 FOUNDATIONS AND RELATED

WORK

The approach presented in this paper bases on differ-

ent tools and methods, which will be introduced in

the following. We will give a brief introduction to

the concept of co-simulation in section 2.1 and to the

Holistic Testing Description (HTD) procedure in sec-

tion 2.2. An overview of the Smart Grid Architec-

ture Model (SGAM) and its corresponding utilisation

from a use case description is given in section 2.3.

Approaches for simulation planning are summarized

in section 2.4. Finally, electricity market mechanisms

testing is brieﬂy discussed in section 2.5.

2.1 Co-simulation

As components in smart grid are highly intercon-

nected, test cases quickly become too complex for

pure analytical handling. Therefore, simulation based

experiments are important intermediate step for the

validation process (Steinbrink et al., 2018). As sim-

ulators have been developed to cover only one re-

search area or domain, simplifying the effect of other

connected domains, co-simulation has been devel-

oped, which consists of multiple simulators coupled

together by a software interface. Each simulator may

cover a different aspect of the smart grid. Together,

the simulators allow researchers to analyze complex

interactions and dynamics in more detail (Vogt et al.,

2018).

Co-simulation consists of independently devel-

oped and implemented simulation models. Thus,

each simulator has its own solver and works simul-

taneously and independently on its own model. The

coupled simulators dynamically interact through their

model’s input and output variables, so that the out-

puts of one simulator become the inputs of the other

and vice versa (Palensky et al., 2017). The synchroni-

sation and execution process is controlled during run-

time by a master algorithm that orchestrate the simu-

lation.

The concept of co-simulation does not only focus

on pure software simulation. The setup might include

hardware and software interaction as in Hardware-in-

the-Loop (HIL) experiments. In this approach, a real

hardware setup for a domain (or part of a domain)

is coupled with a simulation tool to allow testing of

hardware components under realistic conditions. The

execution of the simulator in that case requires strictly

small time steps in accordance to the real-time con-

straints of the physical target (Nguyen et al., 2017).

2.2 Holistic Testing Description (HTD)

The project ERIGrid proposed a methodology, that

can be used to plan experiments in smart grid con-

text to account for multi domain systems and varied

experimental platforms, nonetheless improving repro-

ducibility of experiment results. This method is called

Holistic Testing Description (HTD) (Heussen et al.,

2019). The template-based approach consists of three

main documents that abstract the test objective from

the testbed. Each document gives a different view of

the system, ﬁlling these documents in order gradually

gives a more concrete view of this System Conﬁgura-

tion (SC). These documents are as follows (Heussen

et al., 2017; Van Der Meer et al., 2017):

1. Test Case (TC)

The starting point of HTD procedure is deﬁning

a test case. The inputs are the Generic System

Conﬁguration ((G)SC) and its corresponding Sys-

tem under Test (SuT), which describes the system

boundary in which the Object under Investigation

(OuI) lies, a list of use cases that could be realised

by this test object, and ﬁnally the test objective,

which declares the Purpose of Investigation (PoI).

From this, the Domain under Investigation (DuI)

could be deﬁned, additionally the use case de-

scription would help deﬁne the Functions under

Test (FuT) and choose the Function under Inves-

tigation (FuI).

2. Test Speciﬁcation (TS)

After identifying the SuT and the OuI, the test

speciﬁcation template is of help in identifying

SIMULTECH 2020 - 10th International Conference on Simulation and Modeling Methodologies, Technologies and Applications

198

the concrete Test Speciﬁcation System Conﬁgu-

ration (TS-SC) which is more granular in respect

of equipment number and speciﬁc connections.

The test design and the input/output parameters

are also identiﬁed in this stage.

3. Experiment Speciﬁcation (ES)

The ﬁnal stage is to map the testing requirement

and the SC to a testbed or collection of testbeds

in an integrated experiment, describing its conﬁg-

uration. It is important to be noted, that the test

system is separated from the experiment realiza-

tion, which could allow the test system to be real-

ized on separate testbeds.

The mapping procedure, that realizes the test de-

scription on a testbed, can be semi-automated as sug-

gested by (Heussen et al., 2019). This can be achieved

by developing a database that stores information ob-

jects about test laboratories and co-simulation soft-

ware as assets, and a method for selecting the appro-

priate tesbed and its integrated component for test re-

alization according to the test objective and test de-

veloper requirements.

The database will give information on the avail-

able testbed components and their connection possi-

bilities. The selection method describes a two-stage

process for deriving an experiment implementation

from a given test speciﬁcation. During the process the

test developers are asked to assess the degree of pre-

cision to which the experimental setup needs to repli-

cate various aspects of the test speciﬁcation (e.g., grid

topology, communication system, static and dynamic

parameters), by examining each aspect (component or

sub-system) of the test system and assigning one of

four different precision levels to it:

Precise. The respective component has to be

matched 1:1 (real hardware).

Equivalent. The respective component has to be

matched equivalently in a dedicated software sim-

ulation tool (e.g., grid topology modelled in Pow-

erFactory), or emulation based manner (e.g., com-

munication network emulator, real time electric

grid simulator, etc.)

Nominal. The respective component can be matched

in a software based manner with some deviations,

but they should only lead to marginal inﬂuences

on objective and results (e.g., CSV ﬁle time series

of a household load, storage module that can be

used as heat or electricity storage)

Irrelevant. The respective system aspect does not in-

ﬂuence the test objective and results.

The result of the assessment phase is pairing each sys-

tem aspect with a precision category. The assessment

can be used to communicate the ﬁxed implementa-

tion requirements of a test and to prioritize the rest of

the system aspects. These constraints, together with

the prioritization, enable an iterative search of the

database. Consequently, the above mentioned sug-

gestion is applied in our approach as the core of our

assisted testbed mapping procedure.

2.3 SGAM

In order to enable the seamless interaction between

automation components across all sectors of the

new smart grid the Gridwise Architecture Council

(GWAC) introduced the concept interoperability to

the electric power infrastructure (GridWise Architec-

tural Council, 2008) by deﬁning a framework consist-

ing of eight interoperability categories. This will re-

sult in a more cost effective integration of new com-

ponents, easier system update and replacement (CEN-

CENELEC-ETSI Smart Grid Coordination Group,

2014).

According to the GWAC, architecture is the next

category after a framework, ”architectures are the

blueprints for solutions addressing the issues identi-

ﬁed in the framework”. It is the right path towards

more speciﬁc solution but also neutral regarding tech-

nology (CEN-CENELEC-ETSI Smart Grid Coordi-

nation Group, 2012b). Accordingly, the Smart Grid

Architecture Model (SGAM) condensed the eight in-

teroperability categories into ﬁve layers. Each layer

is applied on the power grid, which has a hierarchy of

automation functionality (Neureiter et al., 2016).

2.3.1 Interoperability Layers

The ﬁve SGAM layers represent the ﬁrst dimension

of this three dimensional architecture model, and are

described as follows (CEN-CENELEC-ETSI Smart

Grid Coordination Group, 2012b):

Business. The business layer represents the business

view on the information exchange related to smart

grids. Regulatory and economic structures can be

mapped on this layer. It supports business execu-

tives in decision making related to business mod-

els and speciﬁc business cases.

Function. The function layer describes functions and

services including their relationships from an ar-

chitectural viewpoint. The functions are derived

by extracting the use case functionality, which is

independent from actors.

Information. The information layer represents the

information models, that are used to exchange in-

formation between functions. It contains infor-

mation objects and the underlying canonical data

Structured Planning of Hardware and Software Co-simulation Testing of Smart Grids

199

models. These represent the common semantics

for functions and services in order to allow an in-

teroperable information exchange via communi-

cation means.

Communication. The communication layer is to de-

scribe protocols and mechanisms for the interop-

erable exchange of information between compo-

nents in the context of the underlying use case.

Component. The component layer is the physical

distribution of all participating components in the

smart grid context. This includes system actors,

applications, power system equipment, protection

and telemetry devices, network infrastructure, and

any kind of computers.

The other two dimensions reside on the smart grid

plane as value creations chain (Domains) and automa-

tion pyramid (Zones) (Uslar et al., 2019) controlling

the energy supply chain. The SGAM primarily pro-

vides a general reference in how to architect smart

grids.

2.3.2 Mapping of Use Cases to SGAM

As software development is becoming an integral part

of a smart grid architecture, methods from this dis-

cipline has been adopted for modelling smart grids;

such as utilization of use cases for requirement engi-

neering. Today, the IEC 62559-2 Use Case Template

is a broadly accepted structure for describing smart

grid related use cases (Binder et al., 2019). A method

for use case mapping to an SGAM framework is in-

troduced in (CEN-CENELEC-ETSI Smart Grid Co-

ordination Group, 2012b; Neureiter et al., 2016) and

described in the following paragraphs.

The initial step is extracting information such

as name, scope and objective, use case diagram,

actor names, use case steps, information which is

exchanged among actors, and functional and non-

functional requirements. This information are en-

sured to be provided, if the use case template IEC

62559-2 is used. Actors can be of type devices, ap-

plications, persons, and organizations. These can be

associated to domains relevant for the underlying use

case and mapped to the component layer. The busi-

ness layer is intended to host the business processes,

services and organizations which are linked to the use

case to be mapped.

A use case consists of several sub use cases with

speciﬁc relationships, these sub use case can be trans-

formed to functions when formulating them in an

abstract and actor independent way. The informa-

tion layer consists of objects which are exchanged

between actors and derived from the use case de-

scription in form of use case steps and sequence di-

agrams. The communication layer contains protocols

and mechanisms for the interoperable exchange of in-

formation between the components.

2.4 Simulation Planning

While co-simulation and hardware experiments can

be used for testing the dynamic behavior of the smart

grid and the SGAM can be used for the static plan-

ning and evaluation of the smart grid, an integrated

process combining these approaches would be ben-

eﬁcial. Based on the more abstract planning in the

SGAM, concrete simulation scenarios should be de-

veloped, which allow more detailed testing.

Binder et al. have already worked on this chal-

lenge (Binder et al., 2019). The purpose of their

contribution is to allow a quick repetition between

the problem deﬁnition and the generation of code

using speciﬁc toolchain methodology. The method

adopted Model Driven Engineering (MDE) for auto-

mated and rapid code generation for simulation com-

ponents. However, their approach focuses on the gen-

eration of code based on activity diagrams and does

not support the integration of already existing simula-

tion components.

In their publication (Uslar et al., 2019), the authors

suggested that the SGAM view on the system could

be considered as input information for the speciﬁca-

tion of HTD test cases. Thus, a new workﬂow could

be realized that starts with an SGAM model and use

case based representation of a desired smart grid setup

and has test developers derive TCs from it, following

the HTD until the experiment implementation, result-

ing in the validation of all crucial parts of the system.

A process for the simulation planning based on

an information model and catalogs with available

co-simulation components is described in (Schwarz

et al., 2019). The information model describes the

data structure for the modeling of a co-simulation

scenario. The co-simulation components available

for coupling in a scenario are collected in a cata-

log. This approach is implemented based on Seman-

tic Web technologies. The information model is mod-

eled as an ontology and the catalogs are realized in a

Semantic Media Wiki (SMW). With the Page Forms

extension for the SMW a form for the deﬁnition of

components is build, which provides a questionnaire

for new components. With the catalogs of simulation

components, the development of co-simulation sce-

narios can be assisted, as described in (Schwarz and

Lehnhoff, 2019).

SIMULTECH 2020 - 10th International Conference on Simulation and Modeling Methodologies, Technologies and Applications

200

2.5 Electricity Market Testing

Electricity market plays a crucial role for the future

development of electricity grid. Thus, its testing is

of high importance for the development and imple-

mentation of future technologies. But, electricity mar-

kets are different from place to place and even in the

same region, it is possible to ﬁnd different market ap-

proaches. (Barroso et al., 2005) provided an overview

of electricity markets in 23 different countries, and

despite its differences, they classiﬁed it considering:

Market Clearing Process: in power pool, bilateral

contract or a mix of them.

Pricing Scheme: like single market price, nodal

pricing or zonal pricing.

But a wide variety of market models and arrange-

ments can be found. Therefore, to derive a market ex-

periment, a proper deﬁnition of the roles of the play-

ers involved as well as the speciﬁc market type has to

be deﬁned. This section will provide some hints about

how SGAM and HTD methodologies can be used to

perform different electricity market tests.

According to the Smart Grid Reference Archi-

tecture technical report the interoperability between

different actors in the electricity system, is essential

to facilitate a smart market (CEN-CENELEC-ETSI

Smart Grid Coordination Group, 2012b). They deﬁne

a smart market as the environment in which energy

products and services are freely traded by many mar-

ket actors.

SGAM uses actors, which represents trading plat-

forms, for electricity and other electricity products

like grid capacity or ancillary services, to deﬁned

market testing. SGAM is also partitioned into hier-

archical zones, that model the information manage-

ment of the electrical process. The operation zone in

SGAM, coordinates the activities in the market zone

to ensure the safety and stability of the grid. Three ar-

eas for energy services are considered: Energy Mar-

ket (Commodity), Grid Capacity Market, and Flex-

ibility Market (Imbalance). (CEN-CENELEC-ETSI

Smart Grid Coordination Group, 2012b).

On the other hand, HTD clustered markets in

two areas: Energy Markets and Ancillary Services.

The components are actors with particular roles

like: BRPs, market operators, retailers, aggregators,

among others. The business models, market struc-

tures, and rules are considered as domain related con-

straints. In addition, a purely market perspective as

OuI is not the main focus of HTD.

Markets (including the stakeholders role) and en-

terprise zones in HTD, are more related to the ICT do-

main, in particular consider Information Technology

(IT), in which the components are functions aimed

at guaranteeing power system stability or energy bal-

ance according to the SuT.

Therefore, to test markets, the roles and system

actors need to be inline for a common understand-

ing of all parties. The Harmonized electricity market

Role Model (HRM) was chosen for most of the cases

(also referenced in SGAM) performed by ENTSO-E

(ENTSO-E, 2019a), in which an actor represents a

party that participates in a business transaction, and

a role is considered as the behavior of an actor or the

activities of the actors. This deﬁnition was also fol-

lowed for the market testing presented in this paper.

Finally, both methodologies can be used to de-

scribe an electricity market test, but so far they have

been focusing more on IT (communication devices

and control components) than electricity transactions

like bidding process or marginal cost calculations.

(Do Prado et al., 2019), highlighted the necessity of

new business models to ensure operation optimiza-

tion, ﬂexibility, customer integration, and sustainabil-

ity to promote markets liberalization. For this reason,

we propose a way of mapping the electricity mar-

ket test as part of the structured planning of a co-

simulation.

3 APPROACH

The identiﬁcation of suitable model and components

for a co-simulation setup is still a challenge for exper-

iments in co–simulation testbeds. The possibilities to

re-use components or models, or even to reproduce

the same experiments when sometimes models dif-

fer on their abstraction level increase this challenge

(Heussen et al., 2019). Our approach implements a

simulation planning mechanism using the concepts

summarised in section 2, that can help in the iden-

tiﬁcation of the most suitable components or models

to run the desired experiment. Additionally, it sug-

gests a workﬂow that can lead to seamless transition

between modelling of a system speciﬁcation and the

simulation setup validating this system. This ﬂow is

semi-automized in order to assist the test developer

in building the most suitable validation experiment,

enhancing the practicality and reducing the time re-

quired in ﬁlling the HTD template based documents.

The proposed process is described in detail in sec-

tion 3.1, while its technical integration is shown in

section 3.2.

3.1 Process

An overview of the proposed process is depicted in

ﬁgure 1. It shows the integration of HTD and SGAM

Structured Planning of Hardware and Software Co-simulation Testing of Smart Grids

201

LegendLegend

Use Case

Standard

Description

SGAM

Architectural Layer

Functional Layer

SGAM

Architectural Layer

Functional Layer

Holistic Testing Description (HTD)

Generic System Configuration (GSC)

FuT

PUC

Component Recommender

Test Case System Configuration (TS-SC)

OuI/FuI

Component

Catalogs

Information Model

ComponentsParametersConfiguration

Information Model

ComponentsParametersConfiguration

Experiment Questionnaire

Components + standards

Experiment system configuration (ESC)

Test

Case

Test

Specification

Experiment

Specification

Experiment

Catalog

SGAM

HTD Guideline

Inhouse Developed Tools

Future Work

Mapping/ Realization

Information Flow

Simulation

Use Case

Mosaik Simulation

Scenario

SESA-Lab

Simulation Setup

Simulation

Use Case

Mosaik Simulation

Scenario

SESA-Lab

Simulation Setup

Figure 1: Overview over the process and its alignment with the HTD.

(blue and orange boxes), which is described in section

3.1.1. How this integration can be assisted by ques-

tionnaires and catalogs (green boxes) is introduced in

section 3.1.2 to 3.1.3. Finally, the modeling of a con-

crete experiment (green boxes) is described in section

3.1.4 and the execution in 3.1.5 (yellow boxes).

3.1.1 Integration of HTD Test Case and SGAM

The key aspect of this workﬂow is the HTD method-

ology. The workﬂow is aligned with the methodol-

ogy and uses it as a guideline in its three stages de-

scribed in section 2.2. Following the procedure de-

scribed in section 2.3.2, use case description and its

corresponding mapping to the SGAM plan is the ini-

tial step towards modeling of a certain smart grid ar-

chitecture according to the requirements imposed by

the use case description. From an HTD point of view,

this system architecture can be considered as GSC.

Generally a use case description that describe the

system’s main functionality could be decomposed

into several sub use cases as mentioned in section

2.3.2. These granular Primary Use Cases (PUC)

(Neureiter et al., 2016) are transformed to functions

and a collection of them can be selected from HTD

prospective as FuT. Accordingly, the initial steps of

developing the ﬁrst stage of test case is achieved by

modelling of a system in SGAM and extracting PUC

as FuT.

3.1.2 Experiment Questionnaire

To document the system conﬁguration as well as the

system speciﬁcation, an experiment questionnaire has

been developed. This questionnaire is subdivided in

a part with general questions and different subsec-

tions for speciﬁc domains (e.g., elecricity grid, con-

trol, market, thermal, and ICT). Thus, it can be used

ﬂexibly depending on the domains to be considered.

The questionnaire helps also in selection of suitable

components to run a test.

The system has been described in its generic form,

the test developer has to deﬁne the SuT. This is real-

ized by the questionnaire the user has to go through in

order to give a closer look at the boundary of the SuT

with its constituent components and numbers, espe-

cially the OuI. The questionnaire helps to deﬁne test

case and test speciﬁcation terms such as DuI, OuI and

its corresponding FuI, a concrete PoI, and temporal

resolution. The output of the questionnaire represents

the domain independent TS-SC.

SIMULTECH 2020 - 10th International Conference on Simulation and Modeling Methodologies, Technologies and Applications

202

Domain Speciﬁc Questionnaire for the Eletricity

Grid. The supply of electricity is done through mul-

tiple stages and different voltage levels. It is important

when developing a test on an electrical equipment, to

know its position in the supply system and the work-

ing voltage level. The questionnaire is dividing the

grid to areas such as generation, transmission and dis-

tribution. Some tests also need grid data that simu-

lates realistic grids, but data from real grids is usually

not public. Therefore, benchmark grids exist in or-

der to test certain developed methodologies and the

questionnaire contains a database of these benchmark

grids to choose from.

Domain Speciﬁc Questionnaire for the Electricity

Market. Considering that markets can create op-

portunities for innovative business or solve grid prob-

lems, electricity markets can be designed as engineer-

ing tools, based on simulations (Ringler et al., 2016).

Therefore, a section in the experiment questionnaire

deals with the electricity market to determine the cur-

rent regulatory framework, based on the following at-

tributes:

Market Organization. It considers how the electric-

ity market is organized like: Power pools, bilateral

contracts, or a mix of them. In addition, it asks

for the type of scheduling (central or self) and the

contract’s capability to inﬂuence price formation

(physical market contract or only ﬁnancial market

contract).

Pricing Structure. Considers if the market model,

for the particular study calculates nodal prices or

zonal prices

Trading Time Frame. Scheduling, like day ahead,

intraday and real-time market. But depending on

the simulation, there are some models that focus

more on long-term (one to several years ahead),

medium-term (a week ahead up to two years) and

short-term (a week to a day ahead), for the market

planning or expansion analysis.

Capacity Allocation. Considers how much space

can market participants use on cross border lines

by keeping into consideration the grid topology.

Considers if the regulation asks or not for a con-

gestion management.

Actors Involved. Allowing the selection of the ac-

tors in the current regulation, based on the HRM

(ENTSO-E, 2019a).

The Focal Use Case Collection. (Rossi et al., 2016)

and the challenges for the retail market according

to (Do Prado et al., 2019).

This categorization was elaborated to ﬁt the differ-

ent markets types and use cases found in the liter-

ature (CEN-CENELEC-ETSI Smart Grid Coordina-

tion Group, 2012b; Uslar et al., 2019; M

aki et al.,

2016; CEN-CENELEC-ETSI Smart Grid Coordina-

tion Group, 2012a; Rossi et al., 2016). Considering

that a market test requires knowledge of the condition

of the electrical grid, when a market test is planned,

the questions regarding the required precision of the

grid components for the models as well as the grid

topology are asked.

3.1.3 Component Catalogs

Up to this stage, the developed system is independent

from any testbed environment. In theory, it could be

realized on any test environment. This separation of

test description and its realization on testbed aligns

with the HTD methodology. In order to realize the

test system on a test environment, component cata-

logs for hardware and software objects have been de-

veloped to give an overview of the availability of com-

ponents on certain testbed. In our case, these are the

in house environments such as co-simulation frame-

work mosaik (Steinbrink et al., 2019), or SESA-Lab

uscher et al., 2015). The catalog collects many dif-

ferent categorizations of the components, as described

in (Schwarz et al., 2019). The type of components

were deﬁned based on the documentation from ERI-

Grid (M

aki et al., 2016, p.32ff.), which contains a list

of different domains, areas, levels, components, and

attributes.

The recommendation process of the components

is conducted using the assessment criteria presented

in section 2.2 and a component recommender query-

ing the suitable components according to their pre-

cision level, accordingly the testbed environment is

chosen that contains components fulﬁlling the re-

quirements presented in the assessment phase. Ex-

amples of this are given in section 4.

3.1.4 Information Model

The ESC has to describe the detailed system conﬁg-

uration for a test. To do this in a structured and ma-

chine readable way, an information model (Schwarz

et al., 2019) is used, which allows to model data ﬂows

and parameters of simulation components. Due to

its ontological implementation the content of the in-

formation model is available for querying, which can

also be used for validation of a scenario (Schwarz and

Lehnhoff, 2019). For the modeling of hardware com-

ponents the information models was extended with

additional modeling options for the inputs and outputs

and the topology of the power system.

Structured Planning of Hardware and Software Co-simulation Testing of Smart Grids

203

Choose Coupling

Tool

Experiment

Setup

Coupling Tool Catalog

Experiment Catalog

Component Catalogs

Co-SiCoCa

SESALab-CoCa

Domains,

Components &

Evaluation Criteria

Information Model

Data flows

Transformation functions

Configuration

Experiment Questionnaire

Evaluation Criteria

Holistic Testing Description (HTD)

Requirements

SGAM

Figure 2: Technical Integration Diagram.

3.1.5 Simulation Execution

The development of an executable co-simulation sce-

nario is usually done manually. But based on the de-

scribed previous steps with a formal modeled scenario

based on the information model, the automatic gen-

eration of the executable software co-simulation sce-

nario would be possible and will be future work. For

the execution of a hardware co-simulation the infor-

mation model and the ﬁlled out experiment question-

naire could provide a setup for the manual implemen-

tation of the experiment.

3.2 Technical Integration

The technical implementation of the proposed pro-

cess is shown in ﬁgure 2 with the used technologies.

The starting point is the experiment questionnaire in

a SMW, which represents the SGAM and HTD and

also asks for requirements and evaluation criteria of

the experiment. The ﬁlled out questionnaire is stored

in an experiment catalog in the SMW to make it avail-

able for reuse or comparison of experiments. The do-

mains, components, and evaluation criteria identiﬁed

in the experiment questionnaire can be exported as

RDF from the SMW, so that a direct integration in

the information model is possible.

Based on the requirements, which can be deduced

from the experiment deﬁnition, the user could also

be assisted in choosing a suitable coupling tool, e.g.,

mosaik or the SESA-Lab. For this purpose, a ques-

tionnaire for coupling tools has been developed in

the SMW, which makes the characteristics available

for querying with the Semantic Web query language

SPARQL. Especially, the decision between a pure

software co-simulation or an integration of hardware,

can be assisted based on the chosen precision levels

of the components.

Based on the information model, the chosen cou-

pling tool, and the component catalogs in the SMW

the user can get recommendations of suitable simula-

tion components for his experiment and an executable

software or hardware co-simulation scenario can be

build. In the future, the automation of this process

step will be further investigated.

4 EXAMPLES

To evaluate the applicability of our proposed ap-

proach, suitable case studies representing a typical

energy system example have been created. The ﬁrst

case is to test a new tool, that emulates the behavior

of a real component in the distribution grid. The sec-

ond case study presents a market model simulating a

trading platform for energy ﬂexibility.

4.1 Voltage Regulation on a

Distribution Feeder

In the document (CEN-CENELEC-ETSI Smart Grid

Coordination Group, 2012a) a list of systems that may

form a smart grid has been presented. The aim of the

document is to model smart grid systems or subsys-

tems and investigate the missing standardization in in-

formation and communication interoperability. This

SIMULTECH 2020 - 10th International Conference on Simulation and Modeling Methodologies, Technologies and Applications

204

Figure 3: Experiment questionnaire for case study 1.

list is well exhaustive and contains three types of sys-

tems that exists in modern day electricity grid:

Domain speciﬁc systems: generation, transmission,

distribution, distributed energy resources, cus-

tomer premises (CEN-CENELEC-ETSI Smart

Grid Coordination Group, 2012b)

Function speciﬁc systems: E.g., marketplace sys-

tems, demand ﬂexibility systems, smart metering

systems, weather observation and forecast sys-

tems

Systems that usually focus on administration features:

E.g., asset management, clock reference, commu-

nication management, device management

For the evaluation of the presented process, a sys-

tem conﬁguration for automation of the functions on a

feeder line in the electrical distribution grid has been

chosen. This system is generic enough to contain a

cluster of use cases that could be evaluated separately,

additionally an already in-house publication (Ansari

et al., 2019) that implemented a Test Case in this sys-

tem’s context has been developed. This makes the

feeder automation system an ideal case for the eval-

uation of the suggested process. The developed in

house test system investigates power quality regula-

tion by applying this system’s architecture.

Following the above suggested process, the sys-

tem has been already modeled in SGAM as in (CEN-

CENELEC-ETSI Smart Grid Coordination Group,

2012a) where the use case is chosen to be voltage reg-

ulation of the feeder. A list of PUCs is deﬁned that

collectively achieve this use case. From HTD per-

spective, SGAM modelling deﬁnes the generic sys-

tem conﬁguration GSC and FuT, a PoI is deﬁned to

verify the in-house developed virtual Operation Tech-

nologies (OT) such as virtual Remote Terminal Unit

(vRTU) ability to emulate real RTU device function-

ality. The developed questionnaire helps deﬁne the

SuT and give a TS-SC as shown in ﬁgure 3, addition-

ally the selected components precision level will de-

termine the experiment setup and the simulation envi-

ronment suitable to fulﬁll the test requirements.

In this case, SESA-Lab was the suitable testbed

for this experiment. The lab contains real time simu-

lator (opalRT) of the electricity grid which aligns with

the selected precision level requirements imposed by

the test developer, as shown in ﬁgure 4.

4.2 Market Model Testing

Today’s electricity market design requires evolution

to ﬁt new actors and new technologies (ENTSO-E,

2019b). Some solutions will need to evaluate the con-

gestion management, inclusion of locational signals,

visibility of the resources, trades close to real time,

ﬂexibility markets, among others.

For our market testing use case an active power

ﬂexibility trades scenario was created, in which the

European electricity market was characterized (power

pool price based). This use case idea was based on

(Meibner et al., 2019) and (Uslar et al., 2019).

A generic market trading platform was created in

our component catalog. This platform simulates local

energy exchange to enable energy ﬂexibility, that can

signiﬁcantly reduce network expansion costs. The

SuT was modeled in SGAM. Only the market, enter-

Structured Planning of Hardware and Software Co-simulation Testing of Smart Grids

205

Figure 4: Query to ﬁnd suitable components for case study 1.

Figure 5: Experiment questionnaire for market model testing (case study 2).

prise and operation zones of SGAM were deployed, in

the distribution domain. The idea of this use case was

to ﬁnd the suitable components (market trading plat-

form or markets models) to the performed the ﬂexi-

bility analysis modelled in SGAM.

The market trading platform will only match the

bids and reported as a business connection to the

trading partners. The trading partners were a Vir-

tual Power Plant (VPP), that controls and loads with

a ﬂexibility potential, and a Distribution System Op-

erator (DSO) as market agent, that tries to contract

ﬂexibility to avoid congestion.

The categorization helps to seek for a model that

allows price based power pools with Locational Mar-

ket Price (LMP), when the grid is modeled consider-

ing restrictions in lines. The precision level was se-

lected to nominal. Figure 5, shows the market cat-

egorization in the experiment questionnaire express-

ing how the market model or platform should be to

run the use case. These requirements will be used in

a query to propose the better components that ﬁt the

needs. As a result, mosaik was suggested as testbed,

and some other components already implemented in

the catalog were also called for the VPP as well as

loads, DER Units, and a market trading platform de-

scribed for this purpose.

5 CONCLUSION AND FUTURE

WORK

The concepts presented in this paper aim to enhance

existing approaches of smart grid testing and valida-

tion. The proposed workﬂow draws the initial steps

towards establishing a structured planning procedure

for validation in the smart grid domain, by combining

concepts of static smart grid planning using SGAM

and a holistic test case description method for in-

tegrating multi-domain objectives using HTD. The

SIMULTECH 2020 - 10th International Conference on Simulation and Modeling Methodologies, Technologies and Applications

206

workﬂow allows test developers to assess which tools

are most suitable to handle their interdisciplinary val-

idation challenges. Additionally, interdisciplinary

testing in a smart grid context becomes more trans-

parent and reproducible. The presented methodol-

ogy is applied internally to include our institute test-

ing infrastructures (mosaik and SESA-Lab), but it

could be extended to include other labs and soft-

ware co-simulation frameworks. Thus, test devel-

opers have the ﬂexibility of developing for example,

multi-platform tests. This can be facilitated by the use

of the Research Infrastructure Database developed in

ERIGrid (Kulmala et al., 2018).

As proof of concept we presented two relevant

case studies. The market model case study is used to

test the dynamics of a trading mechanism before ap-

plying it as regulation and could be sufﬁciently sim-

ulated in software based matter. The second case

study of voltage regulation in the distribution network

contains transient and dynamic behaviors that could

be simulated properly using hardware components.

The workﬂow was able to suggest the suitable test-

ing framework following the requirements imposed

by the assessment criteria.

The extension of the experiment questionnaire to

add more detailed questions regarding other domains

such as thermal, control, ICT, environmental and so-

cial domain as well as use cases to test its usability, is

a matter of future work. Additionally, further elabora-

tion of market models in the component catalog will

provide models for speciﬁc market studies.

ACKNOWLEDGMENT

This work is funded by the ’6. Energieforschungs-

programm der Bundesregierung’ under project ’MEO

- Modellexperimente in der operativen Energiesyste-

manalyse’ (Grant Agreement No. 03ET4078I).

REFERENCES

Ansari, S., Castro, F., Weller, D., Babazadeh, D., and Lehn-

hoff, S. (2019). Towards virtualization of operational

technology to enable large-scale system testing. In

Proceedings for 2019 Eurocon. IEEE.

Barroso, L. A., Cavalcanti, T. H., Giesbertz, P., and Pur-

chala, K. (2005). Classiﬁcation of electricity market

models worldwide. 2005 CIGRE/IEEE PES Interna-

tional Symposium, (i):9–16.

Binder, C., Fischinger, M., Altenhuber, L., Draxler, D., Las-

tro, G., and Neureiter, C. (2019). Enabling architec-

ture based Co-Simulation of complex Smart Grid ap-

plications. Energy Informatics, 2(1).

Blank, M., Lehnhoff, S., Heussen, K., Bondy, D. E., Moyo,

C., and Strasser, T. (2016). Towards a foundation for

holistic power system validation and testing. IEEE

International Conference on Emerging Technologies

and Factory Automation, ETFA, 2016-November:1–4.

uscher, M., Claassen, A., Kube, M., Lehnhoff, S., Piech,

K., Rohjans, S., Scherfke, S., Steinbrink, C., Ve-

lasquez, J., Tempez, F., and Bouzid, Y. (2015). Inte-

grated Smart Grid simulations for generic automation

architectures with RT-LAB and mosaik. 2014 IEEE

International Conference on Smart Grid Communica-

tions, SmartGridComm 2014, pages 194–199.

CEN-CENELEC-ETSI Smart Grid Coordination Group

(2012a). First Set of Standards. Technical report.

CEN-CENELEC-ETSI Smart Grid Coordination Group

(2012b). Smart Grid Reference Architecture. Tech-

nical report.

CEN-CENELEC-ETSI Smart Grid Coordination Group

(2014). Methodologies to facilitate Smart Grid sys-

tem interoperability through standardization , system

design and testing. (October):1–120.

Do Prado, J. C., Qiao, W., Qu, L., and Ag

uero, J. R. (2019).

The next-generation retail electricity market in the

context of distributed energy resources: Vision and in-

tegrating framework †. Energies, 12(3).

ENTSO-E (2019a). The Harmonised Electricity Market

Role Model, Version: 2019-01. pages 1–21.

ENTSO-E (2019b). Vision on Market Design and System

Operation towards 2030.

GridWise Architectural Council (2008). GridWise Interop-

erability Context-Setting Framework. Technical re-

port.

Hartmann, A. K. (2009). Practical Guide to Computer Sim-

ulations. World Scientiﬁc.

Heussen, K., Bondy, D. E. M., Nguyen, V. H., Blank,

M., Klingenberg, T., Kulmala, A., Abdulhadi, I. F.,

Pala, D., Rossi, M., Carlini, C., van der Meer,

A., Kotsampopoulous, P., Rigas, A., Khavari, A.,

Tran, Q. T., Moyo, C., and Strasser, T. (2017). D-

NA5.1 Smart grid conﬁguration validation scenario

description method. Technical report, H2020 ERIGrid

project.

Heussen, K., Steinbrink, C., Abdulhadi, I. F., Nguyen,

V. H., Degefa, M. Z., Merino, J., Jensen, T. V., Guo,

H., Gehrke, O., Bondy, D. E. M., Babazadeh, D.,

ostl Andr

en, F., and Strasser, T. I. (2019). ERI-

Grid Holistic Test Description for Validating Cyber-

Physical Energy Systems. Energies, 12(14):2722.

Kulmala, A., M

aki, K., Rinne, E., Gehrke, O., Heussen, K.,

Bondy, E., Verga, M., Sandroni, C., Pala, D., Nguyen,

V. H., Besanger, Y., Blank, M., Buescher, M., Findrik,

M., Smith, P., Rigas, A., Khavari, A., Cali, M., Sos-

nina, M., Rikos, E., Bhandia, R., Abdulhadi, I., and

Tran, Q. T. (2018). D-NA5.2 Partner proﬁles. Techni-

cal report, H2020 ERIGrid project.

aki, K., Blank, M., Heussen, K., Bondy, E., Rikos, E.,

Rodriguez, E., Merino, J., Blair, S., and Strasser,

T. (2016). D-JRA1.1 ERIGrid scenario descriptions.

Technical report, H2020 ERIGrid project.

Structured Planning of Hardware and Software Co-simulation Testing of Smart Grids

207

Meibner, A. C., Dreher, A., Knorr, K., Vogt, M., Zarif,

H., Jurgens, L., and Grasenack, M. (2019). A

co-simulation of ﬂexibility market based congestion

management in Northern Germany. International

Conference on the European Energy Market, EEM,

2019-September.

Neureiter, C., Uslar, M., Engel, D., and Lastro, G. (2016).

A standards-based approach for domain speciﬁc mod-

elling of smart grid system architectures. 2016 11th

Systems of Systems Engineering Conference, SoSE

2016, pages 1–6.

Nguyen, V. H., Besanger, Y., Tran, Q. T., Nguyen, T. L.,

Boudinet, C., Brandl, R., Marten, F., Markou, A.,

Kotsampopoulos, P., van der Meer, A. A., Guillo-

Sansano, E., Lauss, G., Strasser, T. I., and Heussen, K.

(2017). Real-Time Simulation and Hardware-in-the-

Loop Approaches for Integrating Renewable Energy

Sources into Smart Grids: Challenges & Actions.

Palensky, P., Van Der Meer, A. A., L

opez, C. D., Joseph, A.,

and Pan, K. (2017). Cosimulation of Intelligent Power

Systems: Fundamentals, Software Architecture, Nu-

merics, and Coupling. IEEE Industrial Electronics

Magazine, 11(1):34–50.

Ringler, P., Keles, D., and Fichtner, W. (2016). Agent-

based modelling and simulation of smart electricity

grids and markets - A literature review. Renewable

and Sustainable Energy Reviews, 57(May 2016):205–

215.

Rossi, M., Carlini, C., Pala, D., Sandroni, C., Strasser, T.,

Rikos, E., Kari, M., Kulmala, A., Rigas, A., Kotsam-

popoulos, P., Van der Meer, A., Bhandia, R., Nguyen,

V. H., Heussen, K., Bondy, D. E. M., Gehrke, O.,

Kosek, A. M., Degefa, M. Z., V

oller, S., Høverstad,

B. A., and Rodr

ıguez, E. (2016). D-JRA1.2 Focal

use case collection. Technical report, H2020 ERIGrid

project.

Schloegl, F., Rohjans, S., Lehnhoff, S., Velasquez, J., Stein-

brink, C., and Palensky, P. (2015). Towards a classiﬁ-

cation scheme for co-simulation approaches in energy

systems. Proceedings - 2015 International Sympo-

sium on Smart Electric Distribution Systems and Tech-

nologies, EDST 2015, pages 516–521.

Schwarz, J. S. and Lehnhoff, S. (2019). Ontological in-

tegration of semantics and domain knowledge in en-

ergy scenario co-simulation. IC3K 2019 - Proceedings

of the 11th International Joint Conference on Knowl-

edge Discovery, Knowledge Engineering and Knowl-

edge Management, 2(Ic3k):127–136.

Schwarz, J. S., Steinbrink, C., and Lehnhoff, S. (2019).

Towards an Assisted Simulation Planning for Co-

Simulation of Cyber-Physical Energy Systems. In

7th Workshop on Modeling and Simulation of Cyber-

Physical Energy Systems (MSCPES), pages 1–6,

Montreal.

Steinbrink, C., Blank-Babazadeh, M., El-Ama, A., Holly,

S., L

uers, B., Nebel-Wenner, M., Ram

ırez Acosta,

R. P., Raub, T., Schwarz, J. S., Stark, S., Nieße, A.,

and Lehnhoff, S. (2019). CPES testing with mosaik:

Co-simulation planning, execution and analysis. Ap-

plied Sciences, 9(5).

Steinbrink, C., Schlogl, F., Babazadeh, D., Lehnhoff, S.,

Rohjans, S., and Narayan, A. (2018). Future per-

spectives of co-simulation in the smart grid domain.

2018 IEEE International Energy Conference, ENER-

GYCON 2018, pages 1–6.

Uslar, M., Rohjans, S., Neureiter, C., Pr

ostl Andr

en, F.,

Velasquez, J., Steinbrink, C., Efthymiou, V., Migli-

avacca, G., Horsmanheimo, S., Brunner, H., and

Strasser, T. I. (2019). Applying the smart grid archi-

tecture model for designing and validating system-of-

systems in the power and energy domain: A european

perspective. Energies, 12(2).

Van Der Meer, A. A., Palensky, P., Heussen, K., Bondy,

D. E., Gehrke, O., Steinbrinki, C., Blanki, M., Lehn-

hoff, S., Widl, E., Moyo, C., Strasser, T. I., Nguyen,

V. H., Akroud, N., Syed, M. H., Emhemed, A., Ro-

hjans, S., Brandl, R., and Khavari, A. M. (2017).

Cyber-physical energy systems modeling, test speci-

ﬁcation, and co-simulation based testing. 2017 Work-

shop on Modeling and Simulation of Cyber-Physical

Energy Systems, MSCPES 2017 - Held as part of CPS

Week, Proceedings.

Vogt, M., Marten, F., and Braun, M. (2018). A survey and

statistical analysis of smart grid co-simulations. Ap-

plied Energy, 222(September 2017):67–78.

SIMULTECH 2020 - 10th International Conference on Simulation and Modeling Methodologies, Technologies and Applications

208