Sustainable Energy, Grids and Networks 18 (2019) 100226

Contents lists available at ScienceDirect

Sustainable Energy, Grids and Networks

journal homepage: www.elsevier.com/locate/segan

Integrating renewable energy in smart grid system: Architecture,

virtualization and analysis
∗
Imane Worighi a,b,c , , Abdelilah Maach c , Abdelhakim Hafid d , Omar Hegazy a,b ,
Joeri Van Mierlo a,b
a
Vrije Universiteit Brussel (VUB), ETEC Department & MOBI Research Group, Pleinlaan 2, 1050 Brussel, Belgium
b
Flanders Make, 3001 Heverlee, Belgium
c
Mohammadia school of Engineers, Mohammed V University in Rabat, Morocco
d
University of Montreal, Computer Science Department, Montreal, Canada

article info a b s t r a c t

Article history: Renewable energy sources (RESs) and energy storage systems (ESSs) are the key technologies for smart
Received 28 August 2018 grid applications and provide great opportunities to de-carbonize urban areas, regulate frequency,
Received in revised form 22 April 2019 voltage deviations, and respond to severe time when the load exceeds the generation. Nevertheless,
Accepted 15 May 2019
uncertainty and inherent intermittence of renewable power generation units impose severe stresses
Available online 20 May 2019
on power systems. Energy storage systems such as battery energy storage system enables the power
Keywords: grid to improve acceptability of intermittent renewable energy generation. To do so, a successful
Smart grid coordination between renewable power generation units, ESSs and the grid is required. Nonetheless,
Micro-grid with the existing grid architecture, achieving the aforementioned targets is intangible. In this regard,
Virtualization coupling renewable energy systems with different generation characteristics and equipping the power
Storage energy systems with the battery storage systems require a smooth transition from the conventional power
Renewable sources system to the smart grid. Indeed, this coordination requires not only robust but also innovative controls
and models to promote the implementation of the next-generation grid architecture. In this context,
the present research proposes a smart grid architecture depicting a smart grid consisting of the main
grid and multiple embedded micro-grids. Moreover, a focus has been given to micro-grid systems
by proposing a ‘‘Micro-grid Key Elements Model’’ (MKEM). The proposed model and architecture
are tested and validated by virtualization. The implementation of the virtualized system integrates
solar power generation units, battery energy storage systems with the proposed grid architecture.
The virtualization of the proposed grid architecture addresses issues related to Photovoltaic (PV)
penetration, back-feeding, and irregularity of supply. The simulation results show the effect of
Renewable Energy (RE) integration into the grid and highlight the role of batteries that maintain the
stability of the system.
© 2019 Elsevier Ltd. All rights reserved.

1. Introduction paradigm shift to sustainable low carbon economy at European

Union, although, has come into conflict between the Northern,
Notwithstanding the Paris Agreement, a technological tran- Western Member states against Central and Eastern European,
sient from hydrocarbon-based power generating units to the causing an obstruction of swiftly advancing de-carbonization [1].
post-petroleum-based sources, there is intangible projective ev- To clarify the importance of integrated renewable energy sources,
idence of such transition in the world [1]. For instance, recent European Union set a goal of reaching 27% in gross final en-
studies into the projective period indicate that energy consump- ergy consumption from renewable energy sources by the end of
tion will increase from 663 to 736 quadrillion Btu between 2015 2030 [3]. Therefore, coupling of renewable energy sources (RESs)
and 2040 [2], with an expected increase in carbon dioxide annual and electric grid has gained momentum and is being widely ac-
emissions from 31.2 to 45.5 billion metric tons. Moreover, re- cepted as an alternative power supply. In Germany, expansion of
cent investigations demonstrate that the possibility of the energy renewable energies is a central pillar of the energy transition to-
wards a non-nuclear renewable system [4]. In addition, the global
∗ Corresponding author at: Vrije Universiteit Brussel (VUB), ETEC Department installed capacity of solar photovoltaic (PV) has dramatically in-
& MOBI Research Group, Pleinlaan 2, 1050 Brussel, Belgium. creased as part of a shift from fossil fuel-based power generations
E-mail address: [email protected] (I. Worighi). towards reliable, clean, efficient and sustainable fuels [5]. The

https://doi.org/10.1016/j.segan.2019.100226
2352-4677/© 2019 Elsevier Ltd. All rights reserved.
2 I. Worighi, A. Maach, A. Hafid et al. / Sustainable Energy, Grids and Networks 18 (2019) 100226

power plants [6]. Andújar et al. [7] explained two main rea-
Acronyms sons for justifying the transit towards coupling renewable energy
RESs Renewable energy sources sources with power plant-based fossil fuels. These reasons are [7]:
environmental concerns, and power losses. However, the stochas-
RE Renewable energy
tic and the intermittent behavior of the PV power generations
ESSs Energy Storage Systems
pose severe stresses to the grid lead to the instability in the
MKEM Micro-grid Key Elements Model electricity supply [8]. More precisely, the intermittent energy
PV Photovoltaic sources can fail to guarantee the continuity and reliability of the
RE Renewable Energy power supply [9]. Besides of the aforementioned challenges of
ICT Information and Communication Tech- PV integration, the back-feeding imposes tremendous operational
nology challenges in power systems [10]. When the local PV genera-
NIST National Institute of Standards and tion exceeds the local load demand, reverse power flows occurs.
Technology The reverse power induces a voltage rise within distribution
SGCN Smart Grid Communication Network networks [10]. Along with that, Calpa et al. [11] analyzed the
SGAM Smart Grid Architecture Model effects of high PV penetration as the main energy source for the
Spanish electrical grid. The authors highlighted the effect of high
SGCG Smart Grid Coordination Group
PV penetration which can lead to a reduction on the local con-
CSIRO Commonwealth Scientific and Indus-
sumption or even negative consumption. The results showed that
trial Organization
the electrical grid should be able to respond to this new shape
DG Distribution Generations by adapting the power generations, controlling consumption or
UCC Universal Charge Controller using storage systems [11]. Similarly, Cohen et al. [12] simulated
MAS Multiagent System a use case of Sacramento feeders and showed negative values for
EMS Energy Management System load demand and a largest reverse power flow. It was shown that
ESSs Energy Storage Systems feeder location (i.e. climate) has a stronger impact than feeder
SOS System of Systems type on the incidence of the reverse power flow and reductions
MG Micro-grid in peak loading [12]. Moreover, Nguyen and Kleissl [13] high-
lighted the impact of PV power generation units, can be reverse
LC Local Controller
power flow, high voltage level and increasing number of tap
HAN Home Area Network
operations. To deal with the reverse power flow, decentralized
BAN Building Area Network PV power generation units have been proposed instead of the
IAN Industrial Area Network centralized generation units [13]. Besides that, integration of re-
SDs Smart Devices newable power generation units as new distributed generations
SP Service Provider encompassing large scale at the transmission level, medium scale
HVAC Heating Ventilating and Air Condition- at the distribution level and small scale on commercial or res-
ing idential building can present challenges for the dispatch ability
PEVs Plug in Electric Vehicles and control ability of these resources and for operation of the
SM Smart Meter electricity system [14].
In this context, the traditional grid must be improved to cope
TOU Time of Use
with the increased penetration of PV and its inherent intermit-
RTP Real Time Pricing
tency. In this framework, energy storage systems can play a
SSM Supply Side Management significant role in meeting or mitigating the mentioned chal-
DC Direct Current lenges and dealing with the variations of PV. From technical
AC Alternating Current expertise point of view, the energy storage technology is consid-
IC Incremental Conductance ered as a one of the disruptive technologies that could change
SSM Supply Side Management the way the energy supply, for end-users [15]. For instance, by
SoC State of Charge installing a storage battery as an energy buffer, system stability
BESS Battery Energy Storage Systems can be improved [15]. In this sense, the generated PV power
can be stored, or fed into the micro-grid. Therefore, in order
TMY Typical Meteorological Year
to accommodate uncertainty of future realizations of demand
SG Smart Grid
and generation, sizing energy storage systems plays a prominent
GHG Greenhouse gas role in the micro-grid [16]. This can mitigate the intermittent
and fluctuations of PV power generation units. In addition, the
integration of energy storage systems during peak load periods
can be also useful to shift electrical demands from on-peak to
goal is to maintain grid stability with high penetration of RESs off-peak [17,18]. In this regard, it is becoming critical to integrate
while satisfying consumer demand. In this regard, it is expected analysis and design of power electronics and power systems in
that this integration could reduce carbon dioxide emissions and order to support this integration.
other air pollutants. As mentioned before, by successful coordi- The integration of new technologies into the conventional
nation between RESs and power systems, ESSs could improve the grid requires both innovative and robust modeling of various
reliability, security, and resiliency of micro-grid applications by components to incentivize and implement Smart Grid architec-
offering ancillary services such as peak shaving. tures with RESs. In order to upgrade the current power grid,
it is required to go through existing Smart Grid architectures.
1.1. Literature review Indeed, an architecture that could support RESs integration and
accommodate higher levels of variable ESSs. In this context, many
The need for integration of RESs into the power system is researchers have proposed broad definitions of smart grid. A
to provide a wide variety of socioeconomic and environmental smart grid can be defined as an upgraded electrical network, re-
benefits, and to minimize the GHG emissions from conventional lying on bi-directional communication infrastructure and power
 I. Worighi, A. Maach, A. Hafid et al. / Sustainable Energy, Grids and Networks 18 (2019) 100226 3

exchange between suppliers and consumers, due to the perva- The objective was to improve Micro-grid operation stability under
sive incorporation of intelligent communication monitoring and irregularities of the input voltage. Similarly, in Ref. [27], distribu-
management systems [19]. It is the coexistence of power flow, tion generations (DG) and renewables with seasonal variation at
Information and Communication Technology (ICT) and financial different locations have been modeled by employing PSCAD soft-
transactions. In this way, the ICT in the smart grid can make the ware. The goal was to maintain the power quality of the system
existing infrastructure including generation, transmission, dis- when the load has fluctuation. The authors proposed a general-
tribution, and consumers more efficient and yet can guarantee ized approach to design, determine the capacity, required for the
the successful coordination between local distributed renewable micro-grids with metrics to meet the power quality indexes [27].
generation units and power system [20]. Thus, the smart grid
is an evolution of the power grid, based on the integration of 1.2. Motivation, objectives, and innovative contribution
new technologies, smart devices, advanced communications and
controls. In this regard, many researchers have proposed a de- Generally, a micro-grid is composed of renewable energy gen-
sign of Smart Grid architecture to support integration of RESs. erations, energy storage systems (ESSs), and loads, which can
In Ref. [21], a generic hierarchical architecture as a framework operate in grid-connected and stand-alone modes. The primary
of various energy management systems has been proposed to objective of the present article is to depict a smart grid ar-
deal with the challenges of the increase in penetration of the chitecture consisting of the main grid and multiple embedded
distributed renewable energy generation. Furthermore, the U.S. micro-grids. In this article, ‘‘Micro-grid Key Element model’’ is
Energy Independence and Security Act directed the National In- employed to model the above architecture that would be double-
stitute of Standards and Technology (NIST) to propose a NIST beneficial: (1) predicting the effect of energy sources on power
model defining the Smart Grid as a coexistence of seven do- system such as solar energy (2) minimizing the cost and risk of
mains in 2007 [18]. Whereas the NIST conceptual model pro- the proposed architecture before implementation. Moreover, a
vides a wide picture of how the fundamental components of virtualized power system would open new avenues for revenue
the smart grid connect and communicate and seems to be a generation as well as utilize computing and network resources
promising reference architecture for standardization of interfaces, more efficiently across the entire power grid. The virtualized grid
it still presents some drawbacks related to the lack of defi- system would provide increasing stability of the physical grid and
nition of Micro-grid systems with owners of energy sources. isolate problems more quickly.
Hence, the Smart Grid Coordination Group (SGCG), driven by According to the literature, the proposed architecture in this
the European Committee for Standardization (CEN), European article enjoys several advantages, compared to the recent stud-
Committee for Electrotechnical Standardization (CENELEC) and ies [28–31]. For instance, Sermakani et al. [28] proposed a
European Telecommunications Standardization Institute (ETSI) multiagent (mesh topology) system based on a decentralized
designed the Smart Grid Architecture Model (SGAM) and thus micro-grid control. However, the maintenance of the proposed
by international experts [22]. The SGAM as a three-dimensional topology is very difficult along with its tough administration.
framework consisting of domains, zones, and layers, provides a Moreover, the cost of implementation of such architecture is
structural approach for modeling smart grid use cases [23]. In higher than other network topologies, making it an expensive
option. In addition, Wang et al. [29] proposed a three-layer ar-
this sense, the NIST model has been extended by the European
chitecture of a smart grid, highlighting the energy trading among
Community which integrates a ‘‘Distributed Energy Resources’’
MGs by employing Peer-to-Peer (P2P) networks. Nevertheless,
(DERs) domain. The updated model reflects the growing im-
the unstructured networks such as P2P lead to difficulties with
portance of DER, which includes non-traditional sources such
communication resources, and limit the possibility for manag-
as customer-owned solar and wind power systems. These DERs
ing different plans [32]. To address the above drawbacks, the
with advanced functionalities enable power system designers to
present article proposes a Smart Grid as a system of multiple
incorporate ‘‘Micro-grids’’ into the existing grid architectures.
embedded micro-grids, including a double-layer controller. The
In this way, during the next decade, micro-grids will emerge
high-level of controller is the centralized controller at the main
as an alternative to the current centralized energy generation
grid, and the low-level of controller, known as a local controller
systems, because they can provide economic benefits through
is associated with the micro-grid level. In addition, Ref. [30]
avoiding long distance transmission [24]. Also, they can enhance
proposed a transformative architecture for a normal operation
the integration of small and medium size of DER units into the
and self-healing of networked micro-grids (MGs), composed of
electric grid [24]. Moreover, the micro-grids enable effectively
both cyber links for communication and a physical common
the main grid to disconnect different part of that when power
bus for power exchange. However, the financial flow has not
is disrupted. Hence the power system can be more adaptive and
been considered within this architecture. Furthermore, Ghiani
reliable against possible either fluctuation or fault. In this sense,
et al. [31] presented multidisciplinary key interactions in a Smart
a Micro-grid model is essential for after-the-fact event analysis.
Distribution Network (SDN) to the strong connection between
It can reduce system complexity and provide better insight into
the design and operation of SDN. Therefore, the connections
RESs integration studies. Furthermore, as one of the possible are established between three main fields: management and
solution to the inherent intermittent renewable power generation control system (MCS), monitoring system (MS), communication
units is coupled grid with ESSs, Wood et al. [25] has used the and processing system (CPS). Nevertheless, the market interaction
‘‘UltraBattery’’ technology, an entire new class of advanced lead– and financial transactions have not been taken into consideration
acid batteries invented by the ‘‘Commonwealth Scientific and in terms of necessary multidisciplinary approach to the design
Industrial organization’’ (CSIRO), to manage variability and shift of the smart grid architecture. In this regard, in addition to the
energy demand. This technology focused on Micro-grid systems above connections, the proposed smart grid architecture in the
to combine renewable energy sources with a storage system, thus present article has a connection with the market interaction and
providing multiple benefits, making renewable resources reliable financial transactions to address the above gaps.
and dispatchable [25]. Adonis et al. [26] presented an analysis The main contribution of the present article can be categorized
of control strategies developed for a Micro-grid control struc- as follows:
ture, when it is integrated with renewable energy sources. The
authors developed a control strategy for the load management 1. Presenting the Smart Grid as a System of multiple embed-
and performed an analysis through MATLAB/Simulink simulation. ded micro-grids;
4 I. Worighi, A. Maach, A. Hafid et al. / Sustainable Energy, Grids and Networks 18 (2019) 100226

2. Introducing a micro-grid consisting of multiple nano-grids; small residential power systems with renewable sources, storage
3. Coupling of the proposed grid structure with solar power systems, and domestic loads. The nano-grids and micro-grids are
generation units, and the effect of their penetration on the electric grids which correspond to the residential power sys-
power system is discussed; tems and neighborhoods, respectively. They are also connected
4. The proposed MKEM is optimally designed and accurately to the power distribution grid or to another micro-grid [31].
modeled in the power-systems simulation tool GRIDLAB-D; The micro-grid involves different technologies and contains loads,
5. Integrating ESSs into the proposed grid structure to main- distributed generators (PV or wind turbines), Distributed Sys-
tain system stability. tems, Electric Vehicles (EVs), ESSs and power electronic systems
[32,33]. In this regard, power electronic systems are required
To sum up, in the following paragraphs, a smart grid ar- to enable the micro-grid to operate in grid-connected or island
chitecture using embedded systems of micro-grids is presented modes. Moreover, these systems are needed to connect DER to
and proposed through MKEM modeling. The main objective of the nano-grids using inverters for PV or batteries to obtain the
introducing such architecture is to reduce system complexity and required frequency.
improve flexibility of the system. The introduced smart micro- To deal effectively with the challenges of integrating RESs and
grid is composed of renewable energy generations, energy storage nano/micro-grids with the electric grid, contemplating electric
systems (ESSs), and loads, which can operate in grid-connected source intermittency and load inconsistency, this study proposes
and stand-alone modes. Then, the proposed micro-grid model is a smart grid architecture, depicted in Fig. 1. The proposed ar-
implemented to test integration and penetration of RESs. chitecture consists of a double-layer controller, one of which
In addition, the rest of the article is organized as follows: is a centralized controller at the main grid level and the other
the proposed Smart Grid architecture is presented in Section 2. one is a local controller (LC) that controls the micro-grid level.
Section 3 depicts the elements of the proposed architecture by In centralized operation, each LC receives the set points of the
focusing on micro-grid and presenting a MKEM. Results and dis- corresponding Smart Grid (SG) Controller. However, this type of
cussion are provided in Section 4. Finally, the outcome of the
control has low reliability and redundancy [38]. In decentralized
article is summarized and concluded in Section 5.
operation, each local controller decides locally [39]. In the present
research, the proposed architecture uses a hybrid approach defin-
2. Proposed smart grid architecture
ing both the centralized and decentralized operations within the
main grid. Moreover, three flows can be distinguished: the power
The backbone of a smart grid, known as a complex System
flow which can be bidirectional in case of batteries or consumers
of Systems (SOS) [33], highlights ESS technology to reach better
with production capabilities, the information flow which con-
asset utilization, while to maintain reliable system operation,
sists of an ICT platform, and the financial flow related to the
demand response, and environmental protection through em-
commercial transactions including pricing and wholesaling [18].
ploying various generation types (e.g. solar and wind). In this
The proposed SoS architecture encompasses three levels which
regard, the main concept of the smart grid revolves around bi-
are: the macro-grid level referring to the main grid, the micro-
directional communication and power flow according to the def-
inition proposed by NIST Smart Grid Conceptual Reference Model, grid level and the nano-grid level. The macro-grid level contains
through seven domains which are: bulk generation, transmission, the transmission system with large scale generation systems, ESS
distribution, markets, operations, service provider, and customer. to store different kind of electricity such as an electrochemical
However, Smart Grids have to be analyzed from a SoS point or mechanical energy system, and large industrial consumers.
of view to realize their full potential and optimize the whole Moreover, a transmission system operator is used to transfer
system [33]. Arasteh et al. used the concept of SoS to model the energy in different forms such as natural gas or the electrical
expansion of DGs which may belong to the distribution compa- power, and a service provider is used to interact with the op-
nies or the private investors, considering the independent behav- erators and ensure proper functioning of the smart grid. At the
ior of each system with respect to other entities [34]. As private micro-grid level, two objectives can be distinguished considering
investors and distribution companies are autonomous, hetero- micro-grid operation modes. In grid-connected mode, financial
geneous and independently operable systems are connected to transactions are made using interactions with the wholesale mar-
provide a common goal, their behavior could be analyzed from kets. However, in islanded mode, the energy and economic flows
a SoS perspective to model their correlation. The SoS perspective of the micro-grid are managed using a local market named re-
aggregates different autonomous systems, interacting on various tail market, and energy efficiency services are provided. In both
levels, which are characterized by operational and managerial modes, the micro-grid guarantees in real time the stability and
independence [35]. A representative example of this perspective the security of the network. At the nano-grid level, peak-shaving,
is the energy management sector in which the Smart Grid is load leveling, and load-shifting are performed to reduce energy
known as a complex system composed of heterogeneous and losses and energy cost employing end user’s preferences, price
independent sub-systems (e.g., consumers, producers, prosumers, signals, management of local generation and local ESS. Moreover,
storages, etc.) that interact to compete or cooperate [36]. Dif- the demand response can be achieved through dynamic pricing
ferent strategies for coordinating constituent systems of a SoS to reduce the gap between supply and generation. The smart
exist such as centralized coordination, where a Central Controller meter is used for reading and transmitting measurements of the
(CC) coordinates constituent systems at the SoS level and all power consumption, generation, and voltage [40]. It also interacts
constituent systems have their own controllers for management with a two-way communication with retailers and controllers.
and operations, known as Local Controllers (LC) [37]. A decen- Moreover, it handles various bill models such as Time of Use
tralized strategy for SoS can also be used, where neither a CC (TOU) or Real Time Pricing (RTP).
nor a pre-specified rule or agreement for constituent systems ex-
ists [37]. Another strategy for SoS is a coordinating strategy which 3. The proposed micro-grid model
is considered as a hybrid strategy that mixes both centralized and
decentralized coordination. The micro-grid is a distribution system with local DERs which
In this regard, the Smart Grid should be considered as a is connected to the traditional centralized electrical grid but is
complex system where the main grid contains multiple micro- able to operate autonomously [41]. A micro-grid could be a kind
grids. These micro-grids consist of several nano-grids known as of smart grid equipped with advanced computer communication
 I. Worighi, A. Maach, A. Hafid et al. / Sustainable Energy, Grids and Networks 18 (2019) 100226 5

Fig. 1. The proposed smart grid architecture.

technologies and smart meters providing more flexibility and analyzed micro-grid operational scenarios using UML and defined
reliability for control and protection of the system [41]. Today, business use-cases, and technical use-cases referring to control
the micro-grids include different types of renewable sources to and management use-cases [46]. Business use-cases involve sell-
achieve environmental and economic benefits [42]. In this sense, ing and buying energy to/from external markets, selling balancing
the micro-grid has emerged as a flexible architecture for deploy- and ancillary services, providing islanding mode, and offering
ing distributed energy resources (DERs) that can meet wide-range communication platform for energy trading. However, the tech-
of needs for different communities [43]. Especially, PV system nical and management use cases include balancing supply and
is one of the most effective DERs in micro-grids [41]. More- demand on the varied time scales, auto configuration, monitor-
over, the micro-grids enhance market operations and consumer ing and state analyses, forecasting generation and consumption,
participation [44]. In this regard, the micro-grid energy markets and optimizing power flow to reduce losses. Furthermore, Sul-
allow small-scale participants such as consumers to actively trade tan et al. focused on two components of smart power system
energy within their community in real time [44]. Therefore, they that are transformers and smart meters and presented system
facilitate a sustainable, reliable, and local balance of generation requirements through UML use-case diagrams that are used to
and consumption [44]. In this context, Eq. (1) identifies the rela- describe actions and functionalities of the system [47]. More-
tionship between the generated power and the load power within over, in Ref. [48], the authors focused on modeling smart grid
a micro-grid at each time interval: functional requirements using use-case descriptors and UML class
∑ diagrams to easily track disturbances and outage events that can
Ploads = Ppv + PESS + Pmarket (1) happen in the electric grid environment. In this regard, the main
purpose of using model driven methodology and high level or
where Ploads , Ppv , PESS , and Pmarket are the load, the power gen- summary use-case diagrams is to clearly describe the functional
erated by the PV systems, the power provided by ESS, and the requirement of a system, help manage complexity and pave a
purchased and sold power from/to Micro-grid, respectively. To way for building smart grid applications [49,50]. In this context,
gather the requirements of a system including internal and ex- a ‘‘Micro-grid Key Elements Model’’ (MKEM) as high-level use-
ternal influences and highlight the role of RESs and ESSs, use case diagram combining business and control and management
cases and standards are considered as an efficient method to aspects is proposed.
determine requirements and to address interoperability issues. In The Fig. 2 shows the interactions of actors, the functionalities
this regard, Unified Modeling Language (UML) is recommended captured and the relationships among actors within business,
to allow the use within further model-based development efforts control and management levels. The proposed MKEM consists of
and to ease the customization by implementers in their own tool eight actors which are defined as follows:
chain [23]. For instance, Eger et al. identified essential use-cases
of a microgrid scenario by employing UML diagrams [45]. Further, 3.1. Consumer
the actors involved in the use-cases were divided into ‘‘people
& organization’’ such as prosumer, service provider, consumer, End-user or consumer can be classified into three groups: a
retailer, and ‘‘systems & devices’’. Similarly, Gopalakrishnan et al. residential customer, a commercial customer, and an industrial
6 I. Worighi, A. Maach, A. Hafid et al. / Sustainable Energy, Grids and Networks 18 (2019) 100226

Fig. 2. Proposed MKEM model.

customer. Therefore, three networks can be defined as follows: provide the required measurements for various control strate-
Home Area Network (HAN), Building Area Network (BAN) and gies [18]. By such devices, consumers can communicate with
Industrial Area Network (IAN). The HAN consists of the following power generation units and pave a way for controlling the load.
entities: smart devices (SDs), a home gateway (HG) and a service The Smart Meter can perform the following functionalities:
provider (SP) [51]. It is a network connecting devices able to send
and receive signals from other devices and applications. The HAN • Measure energy flow;
is thus an enabling factor for the evolution of the smart home and • Exchange information with the electricity supplier.
other applications, such as home energy management (HEM) sys-
3.3. DER
tems [52]. In addition, BAN and IAN are used for commercial and
industrial customers with focus on building automation, heating
DER generally include the distributed generation, and storage
ventilating and air conditioning (HVAC) and other industrial en-
systems. These components can provide benefits such as reducing
ergy management applications [53]. In this regard, the consumer
power losses in transmission system, maintaining grid stabil-
within a micro-grid can have a controllable load such as HVAC,
ity, and improving resiliency. DERs are essential components for
washing machine, water heater, electric vehicle, and an uncon-
the micro-grid whose role is to integrate RESs into the power
trollable load such as TV, lights, stove, and oven. These smart
system and to keep the balance between energy demand and
appliances provide the consumers with information on incentives
supply. Moreover, DERs can improve grid reliability, to better
and disincentives, offering choices and motivations, to change manage energy consumption and to achieve energy efficiency
their consumption profiles. The objective is to help balancing guidelines. However, the main disadvantage of using DERs is
supply and demand and to ensure reliability for dynamic pricing their intermittent nature. Therefore, modern electric grids have
and demand response services considering consumer preferences. to support the integration of the conventional generation tech-
This can be achieved by employing billing, bidirectional flows of nologies, renewable energy resources, energy storage devices and
energy, information and financial transactions. dynamic loads [54]. This interconnection highlights the concept
In this regard, the consumer can perform the following func- of micro-grid and mitigates the issues related to the intermit-
tionalities: tency behavior of RESs. To enhance the independency aspect,
ESSs are disconnected from DER and presented as independent
• Use electricity, generate electricity, store electricity; components in the proposed model.
• Engage in DSM to consume less energy during high- DER can perform the following functionalities:
consumption hours (peak hours) or shift its consumption (to
off-peak periods). • Distribute and produce energy;
• Perform Supply Side Management (SSM) with ESSs.
3.2. Smart meter
In this paper, the generated power Ppv by PV component can
be calculated via using solar irradiance and PV temperature as
A smart meter is an electronic device that measures energy
showed in Eq. (2):
consumption and exchanges the information with the electric-
ity supplier for monitoring and billing. This component can Ppv = Pn × (R ÷ Rref ) × [1 + Kt × (Tc − Tref )] (2)
 I. Worighi, A. Maach, A. Hafid et al. / Sustainable Energy, Grids and Networks 18 (2019) 100226 7

Fig. 3. PV system topology.

where Pn is nominal power of PV system at reference condi- • State analysis and subsequent actions such as: DSM and
tions, R is solar irradiance, Rref is solar irradiance at reference Supply Side Management (SSM) to balance supply and de-
conditions, Kt is temperature coefficient of PV systems, Tc is mand, forecast both energy generation, and consumption,
temperature of PV system ◦ C, Tref is temperature of PV system and Data management to perform monitoring.
at reference conditions, Rref is set to 2 MW/ft2 , Tref = 25 ◦ C.
3.6. Operator
Tc = Tamb + (0.0256 × R) and Tamb is the ambient temperature.
To maintain the stability of the system, the power output of The main strategic goal of operators is to achieve operational
PV system must follow some regulations to be within appropriate excellence by reducing costs involving operation, maintenance
limits as indicated in Eq. (3): and planning of the electric power transmission and distribution
networks [55]. Moreover, the operators are able to be operated
Ppv min (t ) ≤ Ppv (t ) ≤ Ppv max (t ) (3)
effectively when both consumers and generation units interact
Moreover, the PV system is connected to the grid via an with each other.
inverter employed to convert Direct Current (DC) to Alternating The operators can perform the following functionalities:
Current (AC). The topology used in this article is depicted in Fig. 3:
• Sell balancing and auxiliary services;
3.4. Retailer • Provide open trading through the forecasted price by the
service provider and bidding offered by Retailers;
The Retailer is used to sell energy to small lots. Indeed, energy • Determine generation and load schedule through data col-
can be bought at the Wholesale Market and sold to consumers lected by smart meters and forecast of weather and gener-
at the Retail Market using bi-directional transactions. The retail ation performed by the micro-grid service provider;
market provides various services to consumers such as creating • Aggregate data and status from micro-grid participants;
a small market for trading energy using bidding and auction for • Provide protection and restoration.
energy sale. Therefore, electricity consumers can easily switch
between different retail electricity providers. This encourages These functionalities are performed by means of smart meters
consumers to provide load reductions using convenient prices. which collect the required data such as the output of generation,
The retailer can perform the following functionalities: the load behavior, the weather condition, the voltage data and the
• Sell and buy energy; market prices. Moreover, demand response informs the consumer
• Provide bidding and offering. about its energy usage, allows financial transaction between the
consumer and the retailer and provides information about the
3.5. Local controller prices (low or high) which could be useful to determine the
behavior of the battery, whether to charge or discharge.
The LC interacts with SG Controller when the micro-grid is
connected to the main grid. Moreover, DER, storage, and con- 3.7. Service provider
trollable loads have a LC, and a two-way communication is
established with the SG Controller. Especially, at the consumer The Service Provider affords different kinds of services to the
premises, the LC has bidirectional communications with the operators to support them in the operation of the main grid. In
Smart Meter (SM) [40]. The controller is connected to the smart this regard, the Service Provider executes services to customers
meters installed in a home and connected to home appliances and utilities. The operators need to interact with various service
in order to reduce energy use and aggregate loads via the AMI suppliers for ensuring proper functionality of the smart grid. In
networks [53]. In this regard, the SM acting as a gateway is this context, the service supplier is a key player in the power
able to aggregate information related to power consumption and market in the framework of smart grids. Moreover, it takes part in
generation [40]. Therefore, the LC can also gather this information the energy market analysis and coordinates with the consumer.
and then send it to the SG Controller. In addition, the LC is The service provider can charge customers with time-varying
responsible for managing the decision of energy consumption prices such as TOU and RTP. Furthermore, it contributes to other
based on local information and the operator prospects. services such as weather forecasts.
One of the main advantages of employing LC is to provide flex- The Service Provider can perform the following functionalities:
ibility with the main grid in terms of successful interconnection
between RESs and the power system. Moreover, when an outage • Forecast Generation and weather;
occurs, the micro-grid should disconnect from the main utility • Forecast Price trend;
and maintain the normal operating condition. In this sense, the • Balance Supply and Demand;
Local Controller can perform the following functionalities:
3.8. Local ESS
• Switch to/from islanding mode;
• Optimize power flow through implementation of optimiza- Local ESS is a local provider of storage capacity for storing and
tion algorithms, and optimize Battery to Grid (B2G) and Grid delivering energy. It is needed to cover periods of time when the
to Battery (G2B) operations; sun is down, and the wind is not blowing [56]. Local Storage units
8 I. Worighi, A. Maach, A. Hafid et al. / Sustainable Energy, Grids and Networks 18 (2019) 100226

Fig. 4. Battery model.

ensure the balancing of volatile supply and demand. Pumped

Hydro Storage, Compressed Air Energy Storage, Batteries, Super-
capacitors and Electric Vehicles can be classified under ESS. In this Fig. 5. Block diagram of inverter model.
regard, Local ESS and energy management units are important to
avoid wasting the harvested RESs and to enhance its utilization.
The Local Storage can perform the following functionalities:

• Store Energy;
• Produce and consume energy;
• Manage Energy to balance supply and demand;
In fact, the ESS can provide power until it reaches a low state of
charge (SoC) then the main energy source turns on to charge the
ESS [57]. The means of storing that will be used in this work is Fig. 6. Battery based ESS model [60].
Battery Energy Storage Systems (BESS).
Fig. 4 shows the battery model. All elements in this model
depend on the battery state of charge (SoC) and the temperature to maximize the lifetime of BESS. Eqs. (7) and (8) illustrate these
(T). The battery system contains a package with NBs cells that are limits:
connected in series and NBP that are connected in parallel.
The parameters of the Li-ion battery are defined via look-up PESS min (t ) ≤ PESS (t ) ≤ PESS max (t ) (7)
tables based on experimental data. The terminal voltage of the SOCmin ≤ SOC ≤ SOCmax (8)
battery pack VB can be calculated as follows [58]:

VB = NBs (VOC − Ro IB − RP IP ) (4) 4. Simulation and performance evaluation

∫
1 IB
SoCB = SoCB (0) + dt (5) To evaluate the implementation of the proposed model and
3600 CB the MKEM, a residential distribution grid has been modeled
where: using power system modeling tools: MATLAB, and GRIDLAB-D.
dIP (IL − IP ) (IL − IP ) GRIDLAB-D is a widely used open-source power-systems sim-
= = (6) ulation tool. It was developed and maintained by the Pacific
dt RP ∗CP τP Northwest National Laboratory (PNNL) in the U.S [61,62]. It has
Ro is internal resistance, Cp is polarization capacitance, Rp is polar- various objects and modules to analyze the electric power flow,
ization resistance, VOC is open circuit voltage, and SoC is battery residential load and energy management. Furthermore, it can
state of charge. simulate both single-phase and three-phase balanced and un-
In this context, an inverter is used to convert DC from batteries balanced systems, with distributed customers [62]. GRIDLAB-
to AC. The input variables of the inverter is the substation voltage D includes models of various distribution system components
VSUB , the substation current ISUB , and the transformer Root Mean and can also be combined with other simulation tools such as:
Square (RMS) line to line voltage VLL [59]. In addition, the inverter MATLAB or NS3 [63], [64].
current will be a reference current set by the power flow con- To demonstrate the effectiveness of the proposed model and
troller depending on the substation voltage. The inverter model enhance the roles played by actors within a micro-grid, the ef-
is shown in Fig. 5: fect of RESs’ penetration has been analyzed. In that regard, the
Furthermore, BESS are connected to the network through a modified version of the IEEE 13 Node Test System has been used
power converter, to receive the energy of the system. Then this to show the effect of integration of PV systems into the grid.
energy will be stored in the battery. In this case, the batteries are IEEE 13 Node Test System has been modified to integrate lights,
not recharged externally. The subsystem is shown in Fig. 6 [60]. power loads and ZIP load models. Moreover, smart appliances
The outputs of the block are the battery voltage and, the state of such as multi-state physical load models of the HVAC and water
charge. heaters have been used. Furthermore, the Typical Meteorologi-
Besides that, BESS are characterized by a capacity C, and the cal Year data (TMY2) of WA-Seattle has been used for weather
state of charge (SoC) of BESS is defined using Eq. (5). Moreover, data [65]. Overall, the size of both renewable energy sources
the charging or discharging power of BESS and the amount of and energy storage systems depends on the location and the
stored energy must follow limitations to avoid any damage and weather used in the simulation. Subsequently, the outcome of the
 I. Worighi, A. Maach, A. Hafid et al. / Sustainable Energy, Grids and Networks 18 (2019) 100226 9

simulation setting provided a 13-node test system with total of

1247 single-family residences. Each residence has been equipped
with a water heater, a HVAC, ZIP models of different appliances
and triplex meters to measure the required data. Furthermore,
for each residential home, a controller has been used to con-
trol the smart appliances such as the water heater and HVAC.
These controllers are known as transactive controllers and can
interact with the market using price information [66]. In this
context, a local double-auction market has been used. In a double-
auction market, bidding is done from both the market players
i.e. customer and provider. Hence, a double-auction market uses a
two-way communication strategy where both buyers and sellers
have an opportunity to reach a transaction. In this regard, the
market operator collects, buys and sells bids to clear a double
auction, then finds a common cleared market price and quantity
and broadcasts this information back to the participants. As a
result, the existence of a double-auction market enhances the
involvement of buyers and sellers in defining the suitable price
and quantity of electricity.
In addition, given the proposed model, electric loads con- Fig. 7. Monthly bill with varying PV penetrations, on a sunny day.
nected at each node and voltage at the substation, a power flow
calculation has been performed to determine the steady state
node voltages and line currents at each point of the system. In
accept demand, supply bids and clear on five-minute intervals. In
this article, Newton–Raphson has been used. Among the various
numerical methods used for solving the power flow problem, addition, it can also be designed to manage capacity constraints
Newton–Raphson iterative method is the most widely used, since at substations.
it is more reliable, and the number of iterations required for The simulation was run for one month, and weather condi-
convergence is independent of the power system size [67,68]. tion of August from TMY data regarding WA-Seattle has been
Moreover, in the proposed model, one node is designed as used [65]. In addition, the preheating mode for HVAC was elimi-
a swing or slack bus. This node represents an infinite bus and nated.
provides the fixed voltage reference for the solver iterations. Fig. 7 shows the impact of PV on energy cost. It compares the
Furthermore, the scenario that used for the simulation is the system with and without solar power generation units on a sunny
integration of power generation units consisting of solar into day and shows the effect of introducing different capacities of PV
the grid using a virtualization of the proposed micro-grid model. systems (25%, 50%, 75% and 100%) on the monthly bill consisting
Energy demand and cost have been evaluated under different PV of the purchased energy price per month. As can be seen in Fig. 7,
penetrations. Moreover, analysis have been made to show the the accumulative monthly bill, collected by triplex meter (and
impact of RESs on energy cost, energy losses and energy demand. meters) is reduced when a high PV penetration is integrated. It
The power factor control of the converter was enabled to operate provides end-users with the financial incentives to increase PV
at a constant power factor. In addition, the inverter efficiency penetration, leading to a considerable reduction in monthly bill
was defined as 94% and the rated power for a one inverter was as shown in Fig. 7.
defined as 25 kVA. Moreover, the inverter is assumed to provide Moreover, the integration of RESs reduces transmission losses.
ideal maximum power-point tracking, capable of converting the RESs have the capacity to retain the energy loss at a minimum
maximum power from the solar installation. When the inverter level. The Fig. 8 shows power losses through the transformer
is controlled to produce less than this power, it is assumed the under different PV penetrations. Significant energy loss reduction
PV system is loaded to supply this power and the possible power can be achieved for this scenario with high PV penetrations. Note
output of the panel is not fully reached. In this regard, the inverter that the fluctuation in the power losses is associated with the
controls the power flow from the PV and the selected control agent-based simulation, the agents such as consumers, producers,
mode ‘CONSTANT_PF’ defaults to only generating real power, or grid operators, which are the decision-making entities and the
transiting all the available energy produced by the solar PV instal- main interconnections between the different flow networks, try
lation onto the grid. In this simulation, the PV penetration in the to reduce power losses and interact with the environment that
virtualized micro-grid has been measured by summing the power evolves over the time [69]. Therefore, the environment is affected
rating of all PV systems used in the residential power system. by the agents, leading to a small fluctuation [70]. Indeed, the
Indeed, the surface area of the PV system was defined as 5382
demand response strategy used by end-use customers modifies
ft2 and the daily insolation level that the cell was rated for was
the normal consumption patterns in response to incentives from
defined as 4 kWh/m2 .
grid operators along with the fluctuations of high penetration of
This article has focused on introducing BESS to improve the
PV systems [71].
key issues of several problems that have been introduced after
However, the integration of high PV penetrations into the
RESs integration into the grid and perform peak-shaving. Using
system induces some disturbances and issues on the distribution
GRIDLAB-D to simulate a summer residential load profile, com-
level. In the following, issues related to PV integration into the
bination of BESS and PV has been analyzed to provide the most
effective mitigation of restricting issues under high PV penetra- micro-grid are discussed based on the use cases presented in
tion. Therefore, to enhance their role in energy management; Table 1.
batteries have been used with inverters. Each inverter was linked
to a triplex meter. Moreover, a double-auction retail market has 4.1. Back-feeding
been implemented and augmented with a Real-Time Price (RTP)
for bill mode to enhance demand response and customer par- Rapid introduction of solar panels impacts the energy demand
ticipation in energy management. In this sense, the market can curve. This impact is most obvious during the middle of the
10 I. Worighi, A. Maach, A. Hafid et al. / Sustainable Energy, Grids and Networks 18 (2019) 100226

Fig. 8. Transformer power losses with varying PV penetrations, on a sunny day.

Fig. 9. Energy demand of the micro-grid in different PV Penetration scenarios,
Table 1 on a sunny day.
Studied use cases.
Use cases
Integrating different PV penetrations
Integrating 75% PV penetration and ESS
Integrating 50% PV penetration and ESS
Integrating PV with different PV efficiencies
Integrating PV with 0.15 PV efficiency and ESS

day when the irradiation is at its highest and leads to a sub-

stantial demand reduction as shown in Fig. 9. Capacities at 75%
PV penetration and greater than 75% PV penetration resulted in
PV back-feeding to the main grid during peak radiation times.
However, the night peak load remains unchanged even with high
PV penetration due to lack of irradiation. The Back-feeding occurs
when PV generation exceeds demand and losses on a feeder. In
this regard, integration of Renewable Energy (RE) systems causes
reverse power flows, i.e., feeding back into the grid as they are
generally connected near the load center. However, under high
penetration of distributed generation, this issue may cause local
overvoltage in the distribution grid [72]. Besides that, Fig. 9 shows
fluctuation in the energy demand, derived from using a double- Fig. 10. Energy demand of the micro-grid with, and without 75% PV penetration
auction market with demand response strategy, the RTP billing and BESS on a sunny day.
mode where prices are varied every 5 min and demand pat-
tern. The variation in the electricity price leads to the sustained
undamped fluctuation in the system [73]. Indeed, the demand improved reliability and power quality. Furthermore, integrating
response strategy could lead to such fluctuation as explained PV with storage systems provides potential benefits to the grid,
in [74,75]. Moreover, the output power fluctuation of RESs due such as peak shaving, load shifting, load leveling, voltage and
to the dependency of these sources on weather conditions could frequency regulation, outage protection, and other ancillary ser-
cause such oscillation in the energy flow [76]. In this regard, vices. In this regard, to achieve the above benefits, the existence
the fluctuation of electricity prices due to the RTP billing mode of energy storage in the presented scenario is mandatory to allow
along with the consumption’s dependency on the electricity price storing solar energy, mitigating back-feeding and using the stored
and RESs output power lead to the fluctuation in the energy energy at high energy prices.
demand. In the following, effective management of PV integration Therefore, using a virtual system of the proposed micro-grid,
to mitigate this issues and tackle Back-feeding problem has been batteries have been implemented on each transformer triplex
pointed out. meter. In addition, residential houses obtain the power from
In that regard, the increasing penetration of solar generation three sources; the power grid, the rooftop solar panel and the
imposes challenges on the existing distribution infrastructure battery which is charged from solar power generation units in
inducing shifts in peak demand. As the percentage share of gen- the middle of the day, when the sun is at its peak. The battery
erated PV energy is increasing in the total energy basket, it is can also be charged from the power grid during off-peak hours.
necessary to integrate energy storage with PV. Combining solar Fig. 10 shows the effect of integrating battery systems into the
PV with energy storage can provide an effective solution for reg- grid. The PV back-feeding has been removed by the introduction
ulating the load. This combination will have a positive impact on of BESS for 75% of PV penetration. Moreover, in Fig. 11, these
consumers, utilities and communities by means of effective cost, systems regulate the demand of energy and contribute to load
 I. Worighi, A. Maach, A. Hafid et al. / Sustainable Energy, Grids and Networks 18 (2019) 100226 11

Fig. 11. Energy demand of the micro-grid with 50% PV and BESS on a sunny
day.
Fig. 12. Energy demand of the micro-grid varying PV efficiency, on a sunny day.

shifting and peak shaving for 50% of PV penetration. Hence, when

using batteries, the demand of energy shifts into a more suitable
operating region.

4.2. Intermittent power due to the components’ efficiency

Non-renewable sources are known as reliable and controllable

sources, generating a known amount of energy. However, RESs
do not generate a stable energy quantity and are influenced by
various parameters. For instance, the performance of solar panels
depends on the quantity of sunlight, efficiency, air density, tem-
perature and other factors. Moreover, PV parameters are derived
from a physics-level perspective. The area, efficiency, panel type
and orientation, power production models, all these parameters
are used to determine how this installation can generate energy
as the solar conditions change throughout the day due to the
weather conditions. Subsequently, the main challenge in using
RESs is their intermittency and stochastic behavior [77].
Fig. 13. Energy demand of the micro-grid with 0.15 solar efficiency and BESS
Besides that, solar panels need to absorb sun rays to generate on a sunny.
electricity. Thus, the panels should have the best efficiency and
face for the best direction to maximize generation of energy.
Therefore, it is essential to determine the correct size of the PV to
5. Conclusion
satisfy the energy demand. This can be performed by considering
the efficiency of power conversion from the solar insolation to DC
power and the maximum power output of solar. In addition to In this article, a smart gird architecture was presented by
the previous requirements, supporting devices such as inverters means of System of Systems (SOS) technique. The proposed ar-
with effective controls are needed to allow better management of chitecture integrated several micro-grids and nano-grids, known
the produced energy. In this paper, the proposed model has been as small residential power systems with renewable sources, bat-
implemented using 100% PV penetration and different values of tery storage, and domestic load. Such integration can increase
solar efficiency. As shown in Fig. 12, high efficiencies have an the complexity of system modeling. By employing a micro-gird
apparent negative effect on the entire grid. key element approach, this article tried to address the system
Moreover, the efficiency at ‘‘0.15’’ and greater than ‘‘0.15’’ complexity. Moreover, the integration of renewable power gener-
induce significant demand reduction during peak hours. In Fig. 13, ation units was analyzed and investigated by virtualization of the
the intermittent power issue has been removed by integrating the introduced architecture through MKEM approach to test system
optimal size of BESS under an efficiency of ‘‘0.15’’. stability. The system was analyzed under various PV penetrations
As a result, the virtual system of the proposed micro-grid with different PV efficiency to demonstrate the effectiveness and
model provided a normal behavior once power generation units performance of the proposed framework, as well as the employed
consisting of solar energy, integrated into the grid. The local ESS approach. The result showed that the proposed model could
modeled in the proposed architecture, designed in the proposed reduce energy cost and power losses. Moreover, the proposed
MKEM and implemented in the virtual system performed its role. framework addressed the dependency of renewable power gen-
Consequently, the integration of RES and BESS in power systems eration units on weather condition by utilizing energy storage
has beneficial effect; this integration not only contribute to the systems to reduce and control the fluctuation of such distributed
reinforcement of the distribution networks and reduce power generations. Indeed, ESSs acted as not only load but also as
losses, but also perform peak shaving, load shifting, and reduce generation to shift and bring down the power flow. Furthermore,
energy cost. in this article, it was demonstrated that the MKEM can effectively
12 I. Worighi, A. Maach, A. Hafid et al. / Sustainable Energy, Grids and Networks 18 (2019) 100226

be utilized as a robust approach for reliability study, outage de- [12] M.A. Cohen, P.A. Kauzmann, D.S. Callaway, Effects of distributed PV
tection, advanced control, demand response, advanced metering generation on California’s distribution system, part 2: Economic analysis,
Sol. Energy 128 (2016) 139–152, http://dx.doi.org/10.1016/j.solener.2016.
infrastructure and storage management.
01.004.
Hence the present article proposed a smart grid architec- [13] D. Nguyen, J. Kleissl, Research on impacts of distributed versus centralized
ture to provide beneficial opportunities for grid-friendly deploy- solar resource on distribution network using power system simulation and
ment of ESSs in smart grid systems. Nevertheless, the following solar now-casting with sky imager, in: 2015 IEEE 42nd Photovolt Spec Conf
limitations need to be addressed in future studies: PVSC 2015, 2015, pp. 2–4, http://dx.doi.org/10.1109/PVSC.2015.7356208.
[14] Edvard, Smart grid deployment, what we’ve done so far 2012:https://
• Since battery energy storage systems have a significant im- electrical-engineering-portal.com/smart-gr. https://electrical-engineering-
portal.com/smart-grid-deployment-what-weve-done-so-far.
pact on the total operation cost and the life-time of the [15] A.A. Solomon, M. Child, U. Caldera, C. Breyer, How much energy storage
battery reduces during charging and discharging cycles, the is needed to incorporate very large intermittent renewables?, Energy
degradation cost of battery should be considered as a target Procedia 135 (2017) 283–293, http://dx.doi.org/10.1016/j.egypro.2017.09.
in cost-effectiveness analysis. 520.
• Vehicle electrification has recently drawn considerable at- [16] N.E. Mohammad Rozali, W.S. Ho, S.R.Wan. Alwi, Z.A. Manan, J.J. Klemeš,
M.N.S. Mohd Yunus, et al., Peak-off-peak load shifting for optimal storage
tention which raises concern about bi-directional power sizing in hybrid power systems using power pinch analysis considering
flow besides vehicle-to-grid and grid-to-vehicle functional- energy losses, Energy 156 (2018) 299–310, http://dx.doi.org/10.1016/J.
ities that need to be assessed. ENERGY.2018.05.020.
• As recent studies focus on technical and economic aspects [17] I. Worighi, A. Maach, A. Hafid, Modeling a smart grid using objects
interaction, in: Proc. 2015 IEEE Int. Renew. Sustain. Energy Conf. IRSEC
of PV-battery systems, further investigations are merited
2015, 2016, http://dx.doi.org/10.1109/IRSEC.2015.7454968.
in terms of voltage deviation and total cost of investment [18] I. Worighi, A. Maach, A. Hafid, Smart grid architecture and impact analysis
in such technologies by studying weighted multi-objective of a residential microgrid, Colloq. Inf. Sci. Technol. Cist (2017) 854–859,
functions. http://dx.doi.org/10.1109/CIST.2016.7805008.
[19] S. Aleksic, V. Mujan, Exergy cost of information and communication
equipment for smart metering and smart grids, Sustain Energy Grids
Declaration of competing interest
Networks 14 (2018) 1–11, http://dx.doi.org/10.1016/j.segan.2018.01.002.
[20] S. Selosse, S. Garabedian, O. Ricci, N. Maïzi, The renewable energy revo-
No author associated with this paper has disclosed any po- lution of reunion island, Renew. Sustain Energy Rev. 89 (2018) 99–105,
tential or pertinent conflicts which may be perceived to have http://dx.doi.org/10.1016/j.rser.2018.03.013.
impending conflict with this work. [21] I. Mauser, C. Hirsch, S. Kochanneck, H. Schmeck, Organic architecture for
energy management and smart grids, in: Proc - IEEE Int Conf Auton
Comput ICAC 2015, 2015, pp. 101–108, http://dx.doi.org/10.1109/ICAC.
Acknowledgment 2015.10.
[22] CEN/CENELEC/ETSI Joint Working Group on Standards for Smart
We acknowledge Flanders Make for the support to our team. Grids. CEN-CENELEC-ETSI Smart Grid Coordination Group: Smart Grid
Information Security 2012:1–107.
[23] F.P. Andrén, T.I. Strasser, W. Kastner, Engineering smart grids: Applying
References
model-driven development from use case design to deployment, Energies
10 (2017) 1–33, http://dx.doi.org/10.3390/en10030374.
[1] M. Ringel, M. Knodt, The governance of the european energy union: [24] M.J. Hossain, J. Lu, M.A. Mahmud, T. Aziz, Advanced decentralized der
Efficiency, effectiveness and acceptance of the winter package 2016, Energy control for islanded microgrids, in: 2014 Australas Univ Power Eng Conf
Policy 112 (2018) 209–220, http://dx.doi.org/10.1016/j.enpol.2017.09.047. AUPEC 2014 - Proc, 2014, pp. 11–15, http://dx.doi.org/10.1109/AUPEC.
[2] U.S. Energy Information Administration. International Energy Outlook 2014.6966534.
2017. 2017. doi:www.eia.gov/forecasts/ieo/pdf/0484(2016).pdf.
[25] J. Wood, Integrating renewables into the grid: Applying ultrabattery⃝ R

[3] Brussels, Second Report on the State of the Energy Union COMMISSION
technology in MW scale energy storage solutions for continuous variability
STAFF WORKING DOCUMENT Monitoring progress towards the Energy
management, in: 2012 IEEE Int Conf Power Syst Technol POWERCON 2012,
Union objectives – key indicators. Swd 2017;1.
2012, pp. 1–4, http://dx.doi.org/10.1109/PowerCon.2012.6401258.
[4] S. Wurster, C. Hagemann, Two ways to success expansion of renewable
[26] A. Kowalczyk, A. Wlodarczyk, J. Tarnawski, Microgrid energy management
energies in comparison between Germany’s federal states, Energy Policy
system, in: 2016 21st Int Conf Methods Model Autom Robot MMAR 2016,
119 (2018) 610–619, http://dx.doi.org/10.1016/j.enpol.2018.04.059.
2016, pp. 157–162, http://dx.doi.org/10.1109/MMAR.2016.7575125.
[5] H. Akbari, M.C. Browne, A. Ortega, M.J. Huang, N.J. Hewitt, B. Norton, et al.,
[27] Q. Fu, L.F. Montoya, A. Solanki, A. Nasiri, V. Bhavaraju, T. Abdallah, et al.,
Efficient energy storage technologies for photovoltaic systems, Sol. Energy
Microgrid generation capacity design with renewables and energy storage
(2018) 1–25, http://dx.doi.org/10.1016/j.solener.2018.03.052.
addressing power quality and surety, IEEE Trans. Smart Grid 3 (2012)
[6] J. Benedek, T.T. Sebestyén, B. Bartók, Evaluation of renewable energy
2019–2027, http://dx.doi.org/10.1109/TSG.2012.2223245.
sources in peripheral areas and renewable energy-based rural develop-
[28] S. Sermakani, M. Thangaraja, Power Demand Optimization in Smart Grid
ment, Renew. Sustain Energy Rev. 90 (2018) 516–535, http://dx.doi.org/
via Wireless Networks. J. Electr. Electron. Eng. n.d. 18–23.
10.1016/j.rser.2018.03.020.
[7] J.M. Andujar, F. Segura, T. Dominguez, Study of a renewable energy [29] N. Wang, W. Xu, Z. Xu, W. Shao, Peer-to-Peer Energy Trading among
sources-based smart grid. requirements, targets and solutions, in: 3rd Microgrids with Multidimensional Willingness 2018:1–22. http://dx.doi.
IEEE Conf Power Eng Renew Energy, ICPERE 2016, 2017, pp. 45–50, http: org/10.3390/en11123312.
//dx.doi.org/10.1109/ICPERE.2016.7904849. [30] Z. Wang, B. Chen, J. Wang, C. Chen, Networked microgrids for self-healing
[8] G. Notton, M.L. Nivet, C. Voyant, C. Paoli, C. Darras, F. Motte, et al., Intermit- power systems, IEEE Trans. Smart Grid 7 (2016) 310–319, http://dx.doi.
tent and stochastic character of renewable energy sources: Consequences, org/10.1109/TSG.2015.2427513.
cost of intermittence and benefit of forecasting, Renew. Sustain Energy [31] E. Ghiani, A. Serpi, V. Pilloni, G. Sias, M. Simone, G. Marcialis, et al.,
Rev. 87 (2018) 96–105, http://dx.doi.org/10.1016/j.rser.2018.02.007. A multidisciplinary approach for the development of smart distribution
[9] V. Vega-Garita, A.P. Harsarapama, L. Ramirez-Elizondo, P. Bauer, Physical networks, Energies (2018) 11, http://dx.doi.org/10.3390/en11102530.
integration of PV-battery system: Advantages, challenges, and thermal [32] T. Wauters, Turck.F. De, C. Develder, Overlay networks for smart grids, IEEE
model, in: 2016 IEEE Int Energy Conf ENERGYCON 2016, 2016, http: Smart Grid Res.: Commun. (2013).
//dx.doi.org/10.1109/ENERGYCON.2016.7514038. [33] A.J. Lopes, R. Lezama, R. Pineda, Model based systems engineering for
[10] A.S. Bouhouras, K.I. Sgouras, P.A. Gkaidatzis, D.P. Labridis, Optimal active smart grids as systems of systems, Procedia Comput. Sci. 6 (2011) 441–450,
and reactive nodal power requirements towards loss minimization under http://dx.doi.org/10.1016/j.procs.2011.08.083.
reverse power flow constraint defining DG type, Int. J. Electr. Power Energy [34] H. Arasteh, M.S. Sepasian, V. Vahidinasab, P. Siano, Sos-based multiobjec-
Syst. 78 (2016) 445–454, http://dx.doi.org/10.1016/j.ijepes.2015.12.014. tive distribution system expansion planning, Electr. Power Syst. Res. 141
[11] M. Calpa, M. Castillo-Cagigal, E. Matallanas, E. Caamaño Martín, Á (2016) 392–406, http://dx.doi.org/10.1016/j.epsr.2016.08.016.
Gutiérrez, Effects of large-scale PV self-consumption on the aggregated [35] G.M.P. Wanderley, M.H. Abel, E.C. Paraiso, J.P.A. Barthès, MBA: A system
consumption, Procedia Comput. Sci. 83 (2016) 816–823, http://dx.doi.org/ of systems architecture model for supporting collaborative work, Comput.
10.1016/j.procs.2016.04.171. Ind. 100 (2018) 31–42, http://dx.doi.org/10.1016/j.compind.2018.04.011.
 I. Worighi, A. Maach, A. Hafid et al. / Sustainable Energy, Grids and Networks 18 (2019) 100226 13

[36] M. Antal, C. Pop, T. Cioara, I. Anghel, I. Salomie, F. Pop, A system of systems [56] O. Hegazy, Advanced Power Electronics Interface and Optimization for Fuel
approach for data centers optimization and integration into smart energy Cell Hybrid Electric Vehicles Applications Omar Hegazy 2012.
grids, Future Gen. Comput. Syst. (2016) http://dx.doi.org/10.1016/j.future. [57] E. Technology, Energy Recovery Technologies in Public Transport Ricardo
2017.05.021. Barrero 2012.
[37] W.L. Barnes, M. Calvin, K. Linville, Y.N. Song, R.H. Xu, B. Zhao, et al., [58] O. Hegazy, M.A. Monem, P. Lataire, J. Van Mierlo, Modeling and analysis of
Coordination of constituent systems for functionalizing systems of sys- a hybrid PV/second-life battery topology based fast DC-charging systems
tems: An exploration, Procedia Comput. Sci. 114 (2017) 375–383, http: for electric vehicles, in: 2015 17th Eur Conf Power Electron Appl EPE-ECCE
//dx.doi.org/10.1016/j.procs.2017.09.051. Eur 2015, 2015, pp. 1–11, http://dx.doi.org/10.1109/EPE.2015.7311727.
[38] Y. Yoldaş, A. Önen, S.M. Muyeen, A.V. Vasilakos, İ. Alan, Enhancing smart [59] O. Hegazy, R. Barrero, P. Van den Bossche, M.El. Baghdadi, J. Smekens, J.
grid with microgrids: Challenges and opportunities, Renew. Sustain Energy Van Mierlo, et al., Modeling, analysis and feasibility study of new drivetrain
Rev. 72 (2017) 205–214, http://dx.doi.org/10.1016/j.rser.2017.01.064. architectures for off-highway vehicles, Energy 109 (2016) 1056–1074, http:
[39] M. Noussan, Performance based approach for electricity generation in //dx.doi.org/10.1016/j.energy.2016.05.001.
smart grids, Appl. Energy 220 (2018) 231–241, http://dx.doi.org/10.1016/j. [60] J.S. Janosy, Keynote speaker-4: The intelligent electricity network of the
apenergy.2018.03.092. future: Smartgrid, in: 2015 17th UKSim-AMSS Int Conf Model Simul, 2015,
[40] C. Gouveia, D. Rua, F. Ribeiro, L. Miranda, J.M. Rodrigues, C.L. Moreira, et p. 6, http://dx.doi.org/10.1109/UKSim.2015.110.
al., Experimental validation of smart distribution grids: Development of a [61] US Department of Energy. GRIDLAB-D 2019. http://www.gridlabd.org.
microgrid and electric mobility laboratory, Int. J. Electr. Power Energy Syst. [62] K. Mahmud, U. Amin, M.J. Hossain, J. Ravishankar, Computational tools
78 (2016) 765–775, http://dx.doi.org/10.1016/j.ijepes.2015.12.005. for design, analysis, and management of residential energy systems, Appl.
[41] I. Goroohi Sardou, M. Zare, E. Azad-Farsani, Robust energy management of Energy 221 (2018) 535–556, http://dx.doi.org/10.1016/j.apenergy.2018.03.
111.
a microgrid with photovoltaic inverters in VAR compensation mode, Int. J.
[63] M. Vogt, F. Marten, M. Braun, A survey and statistical analysis of smart
Electr. Power Energy Syst. 98 (2018) 118–132, http://dx.doi.org/10.1016/j.
grid co-simulations, Appl. Energy 222 (2018) 67–78, http://dx.doi.org/10.
ijepes.2017.11.037.
1016/j.apenergy.2018.03.123.
[42] M.C. Magro, M. Giannettoni, P. Pinceti, M. Vanti, Real time simulator for
[64] A.S.N. Huda, R. Živanović, Large-scale integration of distributed genera-
microgrids, Electr. Power Syst. Res. 160 (2018) 381–396, http://dx.doi.org/
tion into distribution networks: Study objectives, review of models and
10.1016/j.epsr.2018.03.018.
computational tools, Renew. Sustain Energy Rev. 76 (2017) 974–988, http:
[43] A. Hirsch, Y. Parag, J. Guerrero, Microgrids: A review of technologies, key //dx.doi.org/10.1016/j.rser.2017.03.069.
drivers, and outstanding issues, Renew. Sustain Energy Rev. 90 (2018) [65] W. Marion, K. Urban, Userś Manual TMY2 (Typical Meteorological Year)
402–411, http://dx.doi.org/10.1016/j.rser.2018.03.040. 1961-1990. n.d.
[44] E. Mengelkamp, J. Gärttner, K. Rock, S. Kessler, L. Orsini, C. Weinhardt, De- [66] S. Behboodi, D.P. Chassin, N. Djilali, C. Crawford, Transactive control of fast-
signing microgrid energy markets: A case study: The brooklyn microgrid, acting demand response based on thermostatic loads in real-time retail
Appl. Energy 210 (2018) 870–880, http://dx.doi.org/10.1016/j.apenergy. electricity markets, Appl. Energy 210 (2018) 1310–1320, http://dx.doi.org/
2017.06.054. 10.1016/j.apenergy.2017.07.058.
[45] K. Eger, J. Götz, R. Sauerwein, A Von Jagwitz, D. Boëda, O. Arce, et al., [67] C.C. Oliveira, A. Bonini Neto, C.R. Minussi, D.A. Alves, C.A. Castro, New
FINSENY Microgrid Scenario Building Blocks. 1. 2011. representation of PV buses in the current injection Newton power flow,
[46] A. Gopalakrishnan, A.C. Biswal, Animated operational scenarios for micro- Int. J. Electr. Power Energy Syst. 90 (2017) 237–244, http://dx.doi.org/10.
grid systems using scenario visualization and simulation tools, in: 2014 1016/j.ijepes.2017.01.027.
IEEE Glob. Humanit. Technol. Conf. - South Asia Satell. GHTC-SAS 2014, [68] N.C. Yang, H.C. Chen, Decomposed Newton algorithm-based three-phase
2014, pp. 123–128, http://dx.doi.org/10.1109/GHTC-SAS.2014.6967570. power-flow for unbalanced radial distribution networks with distributed
[47] M. Sultan, N.A. Zafar, UML based Formal Model of Smart Transformer energy resources and electric vehicle demands, Int. J. Electr. Power Energy
Power System 8 (2017) 304–10. Syst. 96 (2018) 473–483, http://dx.doi.org/10.1016/j.ijepes.2017.09.042.
[48] K.M. P.Ranganathan K.E. Nygard, UML Design patterns in a smart grid [69] M. Jdeed, Master Thesis 2016.
prakash, in: Proc. ISCA 26th Int. Conf. Comput. their Appl. CATA 2011, [70] D.P. Chassin, J.C. Fuller, N. Djilali, GridLAB-D : An Agent-Based Simulation
2011. Framework for Smart Grids 2014;2014.
[49] E. Ebeid, M. Valov, R.H. Jacobsen, Model-driven design approach for [71] Modeling of GE Appliances in GridLAB-D : Peak Demand Reduction 2012.
building smart grid applications, in: Proc - 19th Euromicro Conf Digit Syst [72] X. Jin, K. Baker, D. Christensen, S. Isley, Foresee: A user-centric home
Des DSD 2016, 2016, pp. 260–267, http://dx.doi.org/10.1109/DSD.2016.94. energy management system for energy efficiency and demand response,
[50] I. Kaitovic, S. Lukovic, Adoption of model-driven methodology to aggre- Appl. Energy 205 (2017) 1583–1595, http://dx.doi.org/10.1016/j.apenergy.
2017.08.166.
gations design in smart grid, in: IEEE Int Conf Ind Informatics, 2011,
[73] R.U.I. Tan, V.B. Krishna, D.K.Y. Yau, Z. Kalbarczyk, Integrity Attacks on
pp. 533–538, http://dx.doi.org/10.1109/INDIN.2011.6034936.
Real-Time Pricing in Electric Power Grids 2015;V:1–33.
[51] J. Shen, C. Wang, T. Li, X. Chen, X. Huang, Z.H. Zhan, Secure data uploading
[74] X. Zhou, J. Shi, Y. Tang, Y. Li, S. Li, K. Gong, Aggregate Control Strategy for
scheme for a smart home system, Inf. Sci. (N.y) 453 (2018) 186–197,
Thermostatically 2019. http://dx.doi.org/10.3390/en12040683.
http://dx.doi.org/10.1016/j.ins.2018.04.048.
[75] T. Coletta, R. Delabays, L. Pagnier, P. Jacquod, Large Electric Load Fluc-
[52] G.M. Toschi, L.B. Campos, C.E. Cugnasca, Home automation networks: A
tuations in Energy-E ffi cient Buildings and how to Suppress them with
survey, Comput. Stand. Interfaces 50 (2017) 42–54, http://dx.doi.org/10.
Demand Side Management 2016:1–6.
1016/j.csi.2016.08.008. [76] T. Uehara, M. Datta, Frequency Control Method using Automated Demand
[53] P De Martini, L. Kristov, Distribution systems in a high distributed resoures Response for Isolated Power System with Renewable Energy Sources
future: planning, market design, operation and oversight, Future Electr. 2017:1–14. http://dx.doi.org/10.1515/ijeeps-2016-0248.
Util. Regul. Gr (2015) 1–66. [77] F. Akhtar, M.H. Rehmani, Energy replenishment using renewable and
[54] O.D. Montoya, A. Garcés, F.M. Serra, DERS integration in microgrids using traditional energy resources for sustainable wireless sensor networks: A
VSCs via proportional feedback linearization control: Supercapacitors and review, Renew. Sustain Energy Rev. 45 (2015) 769–784, http://dx.doi.org/
distributed generators, J. Energy Storage 16 (2018) 250–258, http://dx.doi. 10.1016/j.rser.2015.02.021.
org/10.1016/j.est.2018.01.014.
[55] M. de Reuver, T. van der Lei, Z. Lukszo, How should grid operators govern
smart grid innovation projects? an embedded case study approach, Energy
Policy 97 (2016) 628–635, http://dx.doi.org/10.1016/j.enpol.2016.07.011.