PHD Nelson Leonardo Diaz Aldana E PDF

Aalborg Universitet
Coordinated Control and Management of Distributed battery-based energy Storage

Systems for Islanded Microgrids
Diaz Aldana, Nelson Leonardo
Publication date:
2016
Document Version
Publisher's PDF, also known as Version of record
Link to publication from Aalborg University
Citation for published version (APA):

Diaz Aldana, N. L. (2016). Coordinated Control and Management of Distributed battery-based energy Storage
Systems for Islanded Microgrids. Aalborg Universitetsforlag. Ph.d.-serien for Det Teknisk-Naturvidenskabelige
Fakultet, Aalborg Universitet
General rights
Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners
and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights.
? Users may download and print one copy of any publication from the public portal for the purpose of private study or research.
? You may not further distribute the material or use it for any profit-making activity or commercial gain
? You may freely distribute the URL identifying the publication in the public portal ?
Take down policy
If you believe that this document breaches copyright please contact us at [email protected] providing details, and we will remove access to
the work immediately and investigate your claim.
COORDINATED CONTROL AND
MANAGEMENT OF DISTRIBUTED
BATTERY-BASED ENERGY STORAGE
SYSTEMS FOR ISLANDED MICROGRIDS
BY
NELSON LEONARDO DÍAZ ALDANA
D ISS ERTAT ION S U B MITTE D 2016

COORDINATED CONTROL
AND MANAGEMENT OF
DISTRIBUTED
BATTERY-BASED ENERGY
STORAGE SYSTEMS FOR
ISLANDED MICROGRIDS
Ph.D. Dissertation
NELSON LEONARDO DIAZ ALDANA
Department of Energy Technology

Aalborg University, Denmark
December, 2016
Dissertation submitted: December 13, 2016
PhD supervisor: Prof. Josep M. Guerrero

Assistant PhD supervisor: Assoc. Prof. Juan C. Vasquez

PhD committee: Associate Professor Erik Schaltz (chairman)

Professor Antonio Moreno-Muñoz
Universidad de Córdoba, Spain
Professor Roberto Faranda
Politecnico di Milano, Italy
PhD Series: Faculty of Engineering and Science, Aalborg University
ISSN (online): 2246-1248

ISBN (online): 978-87-7112-846-8
Published by:
Aalborg University Press
Skjernvej 4A, 2nd floor
DK – 9220 Aalborg Ø
Phone: +45 99407140
[email protected]
forlag.aau.dk
© Copyright: Nelson Leonardo Díaz Aldana
Printed in Denmark by Rosendahls, 2016

Curriculum Vitae
Nelson L. Díaz
Received the B.S degree in Electronic Engineering from the Universidad Dis-
trital F.J.C in 2008, and the M.S. degree in Industrial Automation from the
Universidad Nacional de Colombia in 2011, Bogotá, Colombia. He is cur-
rently pursuing the Ph.D. degree from the Department of Energy Techno-
logy, Aalborg University, Aalborg, Denmark. He is member of the Research
Laboratory of Alternative Energy Sources, Universidad Distrital F.J.C. and
Microgrid Research Group, Aalborg University. His current research inter-
ests include microgrids and power converters control.
iii
Curriculum Vitae
iv
Abstract
The integration of variable renewable energy sources such as solar or wind

generators has meant a big breakthrough in the development of current so-
cieties, not only by reducing the environmental impact caused by conven-
tional sources based on fossil fuels. But also, by helping to support the gro-
wing energy demand and enabling the electrification of remote communities,
which restricted by technical and economic issues cannot be connected to the
utility grid. Due to the uncertainty and variability of the variable renewable
energy generators, new technical challenges, as well as prospective solutions,
have emerged.
Microgrids have appeared as an alternative for enabling flexible integra-
tion of variable renewable energy sources within a local power system in
which loads, generators, and energy storage systems operate coordinately,
for accomplish specific aims of common interest, such as: (i) supplying the
demand relying only on local resources, (ii) ensuring the regulation of the
local power system under different operating conditions, (iii) achieve an effi-
cient use of the renewable generation, and (iv) ensuring safe operating con-
ditions for the distributed units. However, due to the non-homogeneous
characteristics of the distributed generators and loads, it is not possible to
consider unified and homogeneous behaviours for all the distributed units.
Because of that, the coordination needs to consider the particularities of each
distributed units, such as the kind of its primary energy resource, its lim-
itations, and current operating conditions, as well as the requirements and
current operating conditions of the whole microgrid, in order to determine
appropriate individual actions of the distributed resources.
The coordination architecture contains the set of rules and control ac-
tions, which determine the responsibilities (roles) and the ways of interac-
tion (behaviours), that the distributed resources in a microgrid should fol-
low for ensuring the fulfilment of the common aims. This thesis explores
different coordination architectures designed for islanded microgrids based
on distributed renewable energy generators and battery-based storage sys-
tems. The proposed coordination architectures rely on expert systems, which
define the operation of the distributed resources and the whole microgrid
v
Abstract
within a finite number of operation modes, that ensure the appropriate charg-
ing/discharging profiles for energy storage systems based on batteries. Here,
centralized and decentralized architectures are explored and implemented by
means of deterministic automaton systems, which coordinate the changes of
the operation modes at the microgrid derived from the proper stages for
charging the storage devices.
The centralized architectures determine the coordinated actions relying
on a global perception of the whole microgrid. Because of that, grid-interactive
approaches can be easily deployed, in which the distributed units can be
operated as dispatchable sources based on simple equilibrium or optimiza-
tion functions. Alternatively, the coordination of roles among distributed
units can be determined by a centralized decision-making unit, emulating
the role of a coach in a soccer team. This approach has been widely used
for coordinating multi-robot systems such as robot soccer teams. From a
metaphoric point of view, this project has addressed the coordination pro-
blem of distributed resources in a microgrid like the coordination of a robot
soccer team. On the other hand, the proposed decentralized approaches are
purely distributed without the use of additional communication infrastruc-
ture. In this case, bus signalling strategies are used for coordinating the
operation of autonomous distributed generators. Additionally, the proposed
architectures allow cooperative operations between the distributed units, in
order to achieve appropriate active power sharing and ensure equalization of
the state of charge for distributed energy storage systems. This fact, allows
among other advantages, managing the distributed energy storage system
like a single aggregate system.
The performance of the proposed coordination architectures have been
verified experimentally and by means of hardware-in-the-loop simulation in
lab-scale prototypes of islanded microgrids, either in dc or ac configurations.
keywords:
Coordination architecture, Distributed Storage and Generation, Decision-Making,
Energy Storage Equalization, Finite State Machine, Islanded Microgrids.
vi
Resumé
Integrationen af variable vedvarende energikilder såsom sol eller vindmøller

har betydet et stort gennembrud i udviklingen af de nuværende samfund,
ikke kun ved at reducere miljøbelastningen forårsaget af konventionelle en-
ergikilder baseret på fossile brændsler. Men også, ved at hjælpe til at under-
støtte den voksende energibehov og muliggør elektrificering af fjerntliggende
samfund, som begrænset af tekniske og økonomiske spørgsmål kan ikke
forbindes til forsyningsnettet. På grund af den usikkerhed og variabilitet
af de variable energi generatorer vedvarende, har nye tekniske udfordringer
samt potentielle løsninger opstæet.
Microgrids har optrådt som et alternativ til at muliggøre fleksibel inte-
gration af variable vedvarende energikilder inden for et lokalt-power system,
hvor belastninger, generatorer og energilagring systemer fungerer sideord-
net, for udrette specifikke mål af fælles interesse, såsom: (i) leverer efter-
spørgslen stoler kun på lokale ressourcer, (ii) at sikre reguleringen af det
lokale elsystem under forskellige driftsbetingelser, (iii) opnå en effektiv ud-
nyttelse af vedvarende generation, og (iv) at sikre sikre driftsbetingelser for
de distribuerede enheder. Men på grund af de ikke-homogene karakter-
istika for de distribuerede generatorer og belastninger, er det ikke muligt
at overveje ensartede og homogene adfærd for alle de forskellige enheder.
På grund af dette, koordineringen skal overveje de særlige hver distribue-
rede enheder, såsom den form for sin primære energikilde, sine begræn-
sninger og aktuelle driftsforhold, samt de krav og aktuelle forhold i hele
elsystemet drifts, for at fastlægge passende individuelle handlinger distri-
buerede ressourcer.
Koordineringen arkitekturen indeholder det sæt af regler og kontrolforan-
staltninger, der bestemmer det ansvar (roller) og de måder for interaktion
(adfærd), at de distribuerede ressourcer i en elsystemet bør følge for at sikre
opfyldelsen af de fælles mål. Denne afhandling udforsker forskellige ko-
ordinering arkitekturer designet til islanded Microgrids baseret på distribue-
rede vedvarende energi generatorer og batteri-baserede storage-systemer. De
foreslåede koordinering arkitekturer stole på ekspertsystemer, som definerer
driften af distribuerede ressourcer og hele elsystemet inden et endeligt an-
vii
Resumé
tal driftstilstande, at sikre passende opladning/afladning profiler til energi-

lagringssystemer baseret på batterier. Her er centraliserede og decentralis-
erede arkitekturer udforskes og gennemføres ved hjælp af deterministiske
robot systemer, som koordinerer ændringerne i driftstilstande på elsystemet
stammer fra de rigtige trin til opladning af lagerenheder.
De centraliserede arkitekturer bestemme de koordinerede aktioner afhæn-
gige af en global opfattelse af hele elsystemet. På grund af dette, kan grid-
interaktiv tilgange let indsat, hvor de distribuerede enheder kan drives som
Afsendte kilder baseret på simple ligevægt eller optimering funktioner. Al-
ternativt koordinering af roller blandt distribuerede enheder, kan bestemmes
ved en centraliseret beslutningsproces enhed, efterligne rollen som coach i
et fodboldhold. Denne tilgang er ofte blevet brugt til at koordinere med
flere robotsystemer såsom robot fodbold hold. Fra et metaforisk set har dette
projekt rettet koordineringen problemet med distribuerede ressourcer i en el-
systemet som koordinering af en robot fodboldhold. På den anden side, er
de foreslåede decentrale tiltag rent distribueres uden brug af ekstra kommu-
nikationsinfrastruktur. I dette tilfælde er bus signalering strategier bruges til
at koordinere driften af autonome distribuerede generatorer. Derudover de
foreslåede arkitekturer tillader kooperative operationer mellem de forskellige
enheder, for at opnå passende aktive magtdeling og sikre udligning af staten
beregning for distribuerede energilagringssystemer. Dette faktum, tillader
blandt andre fordele, styre distribueret energilagringssystem som en enkelt
samlet system.
Udførelsen af de foreslåede koordinering arkitekturer er blevet verificeret
eksperimentelt og ved hjælp af hardware-in-the-loop simulering i lab-skala
prototyper af islanded Microgrids, enten i dc eller ac konfigurationer.
Nøgleord:
Koordinering arkitektur, Distributed Opbevaring og Generation, beslutning-
sprocesser, energilagring Udligning, Finite State Machine, Islanded Micro-
grids.
viii
Thesis Details and
Publications
Thesis Title: Coordinated Control and Management of Distributed

Battery-Based Energy Storage Systems for Islanded Mi-
crogrids.
Ph.D. Student Nelson Leonardo Díaz Aldana.
Supervisor: Prof. Josep M. Guerrero.
Co-supervisor: Prof. Juan C. Vasquez.
1st Authored Journal Papers:
• Díaz, N. L., Dragičević, T., Vasquez, J. C., & Guerrero, J. M. (2014).

Intelligent Distributed Generation and Storage Units for DC Microgrids
- A New Concept on Cooperative Control without Communications
Beyond Droop Control. IEEE Transactions on Smart Grid, 5(5), 2476-
2485 . DOI: 10.1109/TSG.2014.2341740.
• Díaz, N. L., Guarnizo, J., Mellado, M., Vasquez, J. C., & Guerrero, J. M.
(2016). A Robot-Soccer-Coordination Inspired Control Architecture Ap-
plied to Islanded Microgrids. IEEE Transactions on Power Electronics.
DOI: 10.1109/TPEL.2016.2572262.
• Díaz, N. L., Luna, A. C., Vasquez, J. C., & Guerrero, J. M. (2016). Cen-
tralized Control Architecture for Coordination of Distributed Renew-
able Generation and Energy Storage in Islanded AC Microgrids. IEEE
Transactions on Power Electronics. DOI: 10.1109/TPEL.2016.2606653.
• Díaz, N. L., Vasquez, J. C., & Guerrero, J. M. A Communication-less

Distributed Control Architecture for Islanded Microgrids with Renew-
able Generation and Storage. This paper has been submitted to IEEE
Transactions on Power Electronics.
ix
Thesis Details and Publications
Co-authored Journal Publications:
• Luna, A. C., Díaz, N. L., Graells, M., Vasquez, J. C., & Guerrero, J. M.
(2016). Cooperative Energy Management for a Cluster of Households
Prosumers. IEEE Transactions on Consumer Electronics.
• Luna, A. C., Díaz, N. L., Graells, M., Vasquez, J. C., & Guerrero, J. M.
(2016). Mixed-Integer-Linear-Programming Based Energy Management
System for Hybrid PV-wind-battery Microgrids: Modelling, Design and
Experimental Verification. I E E E Transactions on Power Electronics.
DOI: 10.1109/TPEL.2016.2581021.
Selected Publications in Proceedings with Peer Review:
• Díaz, N. L., Dragičević, T., Vasquez, J. C., & Guerrero, J. M. (2014).

Fuzzy-Logic-Based Gain-Scheduling Control for State-of-Charge Balance
of Distributed Energy Storage Systems for DC Microgrids. In 2014 IEEE
Applied Power Electronics Conference and Exposition (APEC 2014) .
Proceedings of a meeting held 16-20 March 2014, Fort Worth, Texas,
USA. . (pp. 2171-2176 ). IEEE Press. (I E E E Applied Power Electronics
Conference and Exposition. Conference Proceedings).
• Díaz, N. L., Wu, D., Dragičević, T., Vasquez, J. C., & Guerrero, J. M.
(2015). Fuzzy droop control loops adjustment for stored energy balance
in distributed energy storage system. In Proceedings of the 2015 9th
International Conference on Power Electronics and ECCE Asia (ICPE-
ECCE Asia). (pp. 728 - 735 ). IEEE Press.
• Díaz, N. L., Wu, D., Dragičević, T., Vasquez, J. C., & Guerrero, J. M.
(2015). Stored Energy Balance for Distributed PV-Based Active Gener-
ators in an AC Microgrid. In Proceedings of the 2015 IEEE Power &
Energy Society General Meeting. (pp. 1-5). IEEE Press.
• Díaz, N. L., Coelho, E. A. A., Vasquez, J. C., & Guerrero, J. M. (2015).

Stability Analysis for Isolated AC Microgrids Based on PV-Active Gen-
erators. In Proceedings of the 2015 IEEE Energy Conversion Congress
and Exposition (ECCE). (pp. 4214 - 4221 ). [7310255] IEEE Press.
• Díaz, N. L., Luna, A. C., Vasquez, J. C., & Guerrero, J. M. (2015). Energy
Management System with Equalization Algorithm for Distributed Energy
Storage Systems in PV-Active Generator Based Low Voltage DC Micro-
grids. In Proceedings of the 2015 IEEE First International Conference
on DC Microgrids (ICDCM). (pp. 293-298 ). IEEE Press.
• Díaz, N. L., Luna, A. C., Vasquez, J. C., & Guerrero, J. M. (2015). Equa-
lization Algorithm for Distributed Energy Storage Systems in Islanded
x
AC Microgrids. In Proceedings of the 41th Annual Conference of IEEE

Industrial Electronics Society, IECON 2015. (pp. 004661 - 004666). IEEE
Press.
This present report combined with some of the above listed scientific pa-
pers has been submitted for assessment in partial fulfilment of the PhD de-
gree. The thesis is based on the submitted or published scientific papers
which are listed above. Parts of the papers are used directly or indirectly in
the extended summary of the thesis.
The scientific papers are not included in this version due to copyright
issues. Detailed publication information is provided above and the interested
reader is referred to the original published papers. As part of the assessment,
co-author statements have been made available to the assessment committee
and are also available at the Faculty of Engineering and Science, Aalborg
University.
xi
xii
Acknowledgement
This document contains a summary of the Ph.D. project, entitled “Coordi-

nated Control and Management of Distributed Battery-Based Energy Storage
Systems for Islanded Microgrids”, which is carried out at the Department of
Energy Technology, Aalborg University, under the supervision of Prof. Josep
M. Guerrero and Associate Prof. Juan Carlos Vasquez. I would like to express
my sincere thanks to the Universidad Distrital F.J.C from Bogotá Colombia,
to the Colombian Administrative Department of Science, Technology and In-
novation (Colciencias), and the Department of Energy Technology, Aalborg
University for the financial support during the period of study and for giving
me this amazing opportunity for developing my professional and personal
life.
There are not words for expensing my gratefulness to my supervisor,
Prof. Josep M. Guerrero, for provided me the opportunity to work under
his supervision and guidance and also for offering me his trust and support
throughout my stay in Aalborg, not only under academic point of view but
also under different aspects of my stay in Aalborg. I also express my deepest
gratefulness to Associate Prof. Juan Carlos Vasquez, for his kind support and
advice throughout my Ph.D. He always offered a friendly and constructive
environment for improving my work.
Especial thanks to Dr. César Leonardo Trujillo; for all his support, en-
couragement and unconditional friendship, to Dr. Mehdi Savaghebi; for his
confidence and opportunity to share unforgettable moments with him and
his family, to Dr. Yajuan Guan, Dr. Lexuan Meng, Dr. Chendan Li and to
other members of the Microgrid Research Group for their encouragement,
support, friendship and for the nice time that they shared with me. I also
extend my thanks to Prof. Ernane Coelho, Dr. Guillermo Guarnizo, Dr.
Tomislav Dragičević, Prof. Francesc Guinjoan Gispert and to other members
of the research group EPIC (Energy Processing and Integrated Circuits) of
the Universitat Politècnica de Catalunya for allowing me to collaborate, ex-
change experiences, and expand my knowledge. I also give my recognition to
all the staff members of the Energy Technology Department and on behalf of
them especial thanks to Tina Larsen and Corina Busk Gregersen for making
xiii
Acknowledgement
possible all my process during my Ph.D. study.

I would be required a separate chapter to express my more sincere thanks
to Dr. Gitte Olsen and Dr. Niels Smedegaard and on behalf of them to all the
nurses and doctors of the hematology departments of Aarhus Universitet-
shospital and Rigshospitalet of Copenhagen, who with their efforts and cares
saved my life and made possible this dream to come true. I am also thankful
with Inger Malgår and Jørgen Langdal Nielsen who offered me their friend-
ship and support during my stay in Aalborg.
Last but not least, I would like to give my deepest gratitude to my family,
and of course to my friend my partner, my colleague, my inspiration, who in
any role she plays she is the greatest blessing in my life; to the love of my life
Adriana Carolina Luna.
To God thanks because his times are perfect, and for making a perfect
coordination of my life which makes possible this dream to come true.
Nelson Leonardo Díaz Aldana.

Aalborg University, December 13, 2016
xiv
Contents
Curriculum Vitae iii
Abstract v
Resumé vii
Thesis Details and Publications ix
Acknowledgement xiii
List of Figures xvii

List of Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xvii
I Introduction 1
Introduction 3
1 Energy Storage Systems in Microgrids . . . . . . . . . . . . . . . 3
1.1 Grid-Connected and Islanded Operation . . . . . . . . . 4
1.2 Energy Storage Systems in Islanded Microgrids . . . . . 7
1.3 Charge/Discharge Characteristic of Batteries . . . . . . . 11
1.4 Aggregated or Distributed ESSs-Motivation . . . . . . . 15
2 Coordination of Islanded Microgrids Based on Distributed RESs
and ESSs-Motivation . . . . . . . . . . . . . . . . . . . . . . . . . 18
2.1 Operation of an Islanded Microgrid Based on RESs and
ESSs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
2.2 Coordination Architecture . . . . . . . . . . . . . . . . . . 22
2.3 Cooperative Behaviours . . . . . . . . . . . . . . . . . . . 27
3 Thesis Contribution . . . . . . . . . . . . . . . . . . . . . . . . . . 31
4 Thesis Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
5 Thesis Outline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
xv
Contents
II Paper 1 37
III Paper 2 41
IV Paper 3 45
V Paper 4 49
VI Paper 5 53
VII Paper 6 57
VIII Concluding remarks 61

Conclusion 63
1 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
2 Contributions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
3 Future Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
Literature List 67
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
xvi
List of Figures
List of Figures
1 Microgrid scheme with one point of common coupling with
the utility grid. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2 Equivalent circuital scheme of MGs: (a) Grid-Connected, (b)
Islanded Single Master, (c) Islanded Multi Master. . . . . . . . . 6
3 Charge/Discharge curve of a Cell. . . . . . . . . . . . . . . . . . 12
4 Charging/Discharging Characteristic of Lead-acid batteries. . . 13
5 Charging/Discharging Characteristic of Lithium-ion batteries. . 15
6 Microgrid with aggregated ESSs an loads. . . . . . . . . . . . . . 16
7 Microgrid based on distributed active generators. . . . . . . . . 17
8 Microgrid based on distributed resources with coordinated ope-
ration. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
9 MG based on ESSs and RESs: (a) Basic scheme, (b) Equivalent
circuit. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
10 Transition between operation modes for proper charge of Lead-
acid batteries in a MG. . . . . . . . . . . . . . . . . . . . . . . . . 21
11 Active Power curtailment of generation from RES for avoiding
overcharge of the ESS. . . . . . . . . . . . . . . . . . . . . . . . . 22
12 Islanded MG control schemes: (a) Centralized, (b) Distributed. 26
13 Comparative curve between SoC profiles with and without
equalization [1]. Dashed lines show the SoC profiles without
equalization and the continuous lines shows the SoC profile by
performing the SoC equalization. . . . . . . . . . . . . . . . . . . 29
14 SoC profiles for an islanded MG with: (a) aggregated ESS with
a capacity of 15 Ah, (b) distributed ESSs with capacities of 10
Ah and 5 Ah and SoC equalization. The figures were obtained
from the work presented in [2]. . . . . . . . . . . . . . . . . . . . 30
15 Comparison of cycle depth for an islanded MG: (a) without
SoC equalization (b) with SoC equalization. The figures were
obtained from the work presented in [2]. . . . . . . . . . . . . . 30
xvii
List of Figures
16 Comparison between charging current behaviours in distributed

ESSs: (a) without SoC equalization (b) with SoC equalization.
The figures wer obtained from the work presented in [3]. . . . . 30
17 Operation of RESs as gird-interactive units (off-MPPT mode).
Here, the power generation is adjusted to match the power
consumption. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
xviii
Part I
Introduction
1
Introduction
1 Energy Storage Systems in Microgrids

A generalized rising of modern living standards, not only for households
but also in commercial sectors, buildings, and industry had led to a growing
demand for energy resources by the end-users [4]. It is estimated that by 2030
the energy consumption will be double and consequently the CO2 emission
can be increased up to a 75% due to the use of conventional energy generation
systems based on fossil fuels [5]. In this scenario, the capability, reliability
and efficiency of the conventional central generators, connected through long
distance transition lines until the end-user, can be seriously compromised.
This fact combined with environmental effects caused by CO2 emissions have
demanded changes in the way that new energy sources are integrated.
The integration of distributed generation units close to the end-user has
emerged as an alternative to counteract the growing demand more efficiently,
thanks to the reduction in losses compared to conventional utility infrastruc-
ture. On top of that, government policies, combined with personal environ-
mental commitments of private consumers, are promoting the integration of
renewable energy sources (RESs) such as wind, hydro, solar and biomass [6].
Particularly, variable RESs such as wind turbines (WT) and solar photovoltaic
(PV) generation units are the mainstays in Europe and in the United States for
the integration and increasing penetration of renewable energy sources [7, 8].
This fact has meant a change of paradigm for energy consumers who envi-
sion the possibility of reducing the dependence on the main grid and ensure
self-energy consumption, by relying only on their own local energy resources
“prosumer” [9]. However, due to the intermittent nature of variable RESs, it
is not possible to ensure that the local generators can match the local de-
mand by themselves. Therefore, Energy storage systems (ESSs) are used for
smoothing the variability of the renewable energy resource, either by absorb-
ing the excess of generation from RESs or by supplying energy when the gen-
eration from RESs is not enough to meet minimum load requirements [10].
Even in large-scale integration of variable RESs, the ESSs are useful for miti-
gating the impacts caused by variable nature of RESs in operation and plan-
3
Microgrid
Local Load Wind
Microturbine
Photovoltaic
Generation
Microgrid PCC
Central Switch
Control
Wind
Generation Batteries Power
Electronics Photovoltaic
Interface Generation Main
Utility Grid
Energy Flywheel Electric Transmission Power Line
Storage Vehicle AC Distribution Line
DC Distribution Line
Energy Communication Link
Storage
Fig. 1: Microgrid scheme with one point of common coupling with the utility grid.
ning [11].
The local power system composed of loads and distributed energy re-
sources (DER) such as generators and energy storage systems, can be mana-
ged and coordinated as a single entity in order to ensure on-site generation
and consumption. Here the microgrid (MG) concept emerges to define small-
scale power systems based on local resources.
A MG can operate completely in grid-connected or islanded mode and
as well as in dual mode by means of seamless transition between them [5].
Typically, the MG is integrated to the main grid by a common coupling point
(PCC) connected throughout a static switch, therefore, it can be seen as a
unified entity from the point of view of the distribution system as shown in
Fig. 1 [12]. Also, a MG can operate, either in ac or dc configurations based
on specific requirements of the DERs and loads, or it can be configured as
hybrid ac/dc system by means of proper power electronic interfaces [13, 14].
1.1 Grid-Connected and Islanded Operation

A MG continuously connected to the main grid can import or export energy
from or toward the main grid, for balancing supply and demand. Since
one of the main objective of a microgrid is ensuring self-consumption and
generation based on its own local resources, its interaction with the main
grid differ from the traditional approach oriented to maximize the energy
exported to the main grid [9]. The conventional strategy relies on the use
of ESSs for storing the energy during the time where the generation exceeds
the demand and releasing the stored energy when generation is lower than
4
1. Energy Storage Systems in Microgrids
consumption [15]. Because of that, the participation of microgrids on the

energy market is commonly limited at the retail level [16].
The increasing penetration of small-scale microgrids based on variable
RESs implies additional technical challenges for the Distribution System Ope-
rator (DSO). Since the distribution systems will be exposed to more variabil-
ity in the net load, it is necessary to include additional support schemes in
order to ensure appropriate frequency and voltage regulation [17, 18]. The
cost derived from the inclusion of new support schemes or even for repla-
cing old infrastructure, in accordance with new requirements, is passed in
a grand extent, to the end-consumers through additional surcharges in the
retail prices [18, 19].
Apart from that, MGs are mainly connected to the utility grid through
the low-voltage (LV) distribution system, therefore, the injection of energy to
the main grid may cause significant variation in the voltage magnitude along
the LV feeder and consequent deterioration of voltage quality [15]. To solve
this problem different actions have been proposed in order to limit the active
power feed-in to the main grid such as active power curtailment of generation
from RESs [15]. This fact, in addition to recent reductions in feed-in tariffs,
has promoted a more independent operation of microgrids [9], [18].
Nevertheless, the MGs may always appeal to the utility grid as a backup
solution, in the cases where the storage unit is not able to provide or store
more energy, or under unforeseen situations in which one or more of the
distributed energy resources (DERs) have to be disconnected [5]. Also, the
main grid can behave as an infinite energy buffer in order to balance short-
time mismatches between generation and consumption, which are common
in power systems with variable RESs [20, 21]. In this way, load demand and
power balance can be met all the time regardless of different operating con-
ditions and variability of RESs. As a consequence, the reliability of the lo-
cal power system is enhanced [22]. Here, the concept of connected island
mode appears for defining the growing trend for grid-connected microgrids
in which importing/exporting energy with the main grid is avoided as much
as possible [23, 24].
Another characteristic of grid-connected MGs is that the main grid be-
haves as a master unit by defining the references for voltage amplitude and
frequency in the PCC. Meanwhile, the DERs are mainly grid-following units
synchronized in phase and frequency with the main grid and controlled in
current control mode (CCM) for injecting or absorbing active/reactive power
(Fig. 2(a)) [25]. This fact, added with the aforementioned backup and buffer-
ing support provided by the main grid, facilitates the management of the
DERs. The power references for grid-following units can be dispatched by
a high-level power management strategy, deployed in the microgrid central
control as shown in Fig. 1 or they can be locally calculated according to spe-
cific power profiles (i.e. maximum power point tracking (MPPT) in variable
5
YLineGrid
Main  Grid
PCC
YLine1 YLine 2 YLOAD YLine3 YLine 4
ICPS1
ICPS 2
ICPL1
ICPL2
YCPL1 YCPS1 YCPS 2 YCPL 2
Grid-Following Grid-Following LOAD Grid-Following Grid-Following
Microgrid
(a)
YLine1 PCC Microgrid
Grid-Forming
YLine 2 YLOAD YLine3 YLine 4

ICPS1
ICPS 2
ICPL2
YCPS1 YCPS 2 YCPL 2
Grid-Following LOAD Grid-Following Grid-Following
(b)
YLine1 Microgrid YLine 4

PCC
Grid-Forming
Grid-Forming
YLine 2 YLOAD YLine3

ICPS1
ICPL2
YCPS1 YCPL 2
Grid-Following LOAD Grid-Following
(c)
Fig. 2: Equivalent circuital scheme of MGs: (a) Grid-Connected, (b) Islanded Single Master, (c)
Islanded Multi Master.
RESs or regulated process of charge in ESSs). Within the current trend in

MGs, the management system aims for minimizing the power exchange with
the main grid and maximizing the use of the local energy resources. Also,
another pre-defined optimization objectives can be achieved [23–25].
Islanded operation may appear for several reasons such as a fault in the
main grid, intentional disconnection due to maintenance needs or when eco-
nomic and geographic issues do not allow an interconnection with the main
grid (i.e. remote areas) [26, 27]. Particularly, islanded microgrids based on
variable RESs have been installed worldwide for electrification of remote vil-
lages in the developing world [28]. The access to electrical energy means an
important advance in the development of poor communities and represents
a technological challenge for the deployment of islanded MGs, by consider-
ing that still around 17% of the global population remain without access to
electricity [29].
6
However, the management of DERs in islanded microgrids involve addi-

tional technical challenges, since the voltage and frequency are not imposed
anymore by the main grid and should be generated internally by the DERs.
Once the microgrid is disconnected from the main grid, the responsibility of
forming the grid in the islanded power system (grid-forming role), should
be assumed by at least one of the DERs, which will operate under voltage
control mode (VCM) [21, 30]. The grid-forming unit will behave as an energy
buffer in the islanded power system in order to ensure the power balance and
to avoid losing the control over frequency and voltage in the common bus.
The grid-forming role can be assumed by a single unit in the islanded MG
(single master configuration as shown in Fig. 2(b)), or this role can be shared
by multiple DERs (multi-master configuration as shown in Fig. 2(c)) [31].
Since the grid-forming units are responsible for managing the power ba-
lance in the islanded microgrid, they should have enough capability and an
adequate reserve for injecting or absorbing energy. Because of that, the ESSs
are commonly used as the grid-forming units [25], [31]. In general, ESSs can
provide fast response in terms of power injection and they can ensure either
short-term or long-term balancing and power dispatch.
1.2 Energy Storage Systems in Islanded Microgrids

For microgrids based on small-power variable RESs the energy storage sys-
tem should be scalable and modular in order to be sized properly in accor-
dance to different characteristics of the distributed generators and load re-
quirements. Additionally, several characteristics such as life-time, cycle life,
power rates, energy density, availability, capital cost, etc., should be conside-
red for an adequate selection of the ESSs. In [11] is suggested that superca-
pacitors, flywheels, super conducting magnetic energy storage (SMES), and
batteries are the most suitable storage technologies for applications based on
variable RESs. Here is a brief summary about the characteristics of these
storage technologies.
• Supercapacitors: They are electrical energy storage systems character-

ized by high power density (10.000 W/kg), high cycling (> 100.000)
and long-life (8-10 years). They are preferred for fast cycling appli-
cations because the can be charged and discharged faster than others
ESSs. However, they have low energy density, which makes them not
suitable for long-term balancing [32]. On the contrary, they are ideal for
providing a fast power injection, which is useful for smoothing transient
responses and sudden power shortage caused by changes in the ope-
ration mode of the microgrid or due to the variability of RESs. They
are also useful in cases when the load needs high instantaneous current
(i.e for starting an electric motor). They are commonly used connected
7
directly to de dc-link of the variable RESs operating as an energy buffer
for smoothing and filtering the generation from variable RESs. How-
ever, supercapacitors have larger cost per kW than many others electric
energy storage technologies [11], [21], [33].
• Flywheels: They are mechanical ESSs based on a rotating mass, which

stores kinetic energy in accordance to changes in the rotational velocity
of the mass and releases the stored energy via an electric motor. They
are characterized by long life (15 to 20 years), high cycling (10.000 to
100.000), high power density and small energy density. They can pro-
vide very high peak power, therefore they can be used for mitigating
the fast ramping of variable RESs and for fast charging of electric vehi-
cles [34]. The main disadvantage is the highest cost per kWh, because
of that, they have had a small penetration in the market [11], [33].
• Super Conducting Magnetic Energy Storage (SMES): It is a device that

storage energy in form of a magnetic field created by a direct current
flowing through a superconducting coil. SMESs are relatively new tech-
nology, characterized by long life (>20 years), high cycling (>10.000),
high power density with a very fast dynamic response, but also small
energy density. For the integration of variable RESs, the SMESs can
be used in the same applications than supercapacitors and flywheels.
SMESs have demonstrated to be effective for short duration power qu-
ality applications and reactive power support. However, they can be
extremely expensive due to the cooling system required for keeping the
superconductivity of the coil [11], [33].
• Batteries: They are devices that convert chemical energy into electrical
energy by means of an internal chemical reaction. Batteries are catego-
rized between primary and secondary batteries. The primary batteries
can be used only once and they can not be recharged. Meanwhile, in
secondary batteries, the electrochemical reaction can be reversed and
the batteries can be recharged [35]. Therefore, secondary batteries are
ideal for applications with variable RESs since they can be charged and
discharged for smoothing the uncertain generation profiles of RESs. In
general, secondary or rechargeable batteries are characterized by high
energy density -compared with the devices previously introduced- and
good capability in terms of fast power injection [21], [33]. Therefore,
batteries can be considered for medium to long-term applications, ca-
pable of supporting a load from minutes to hours [32], [36], which
makes them desirable for Uninterruptible Power Supply systems or for
islanded MGs. One of the main drawbacks of batteries is that they
have relatively low durability in terms of cycle life, compared with the
storage technologies aforementioned [37]. A wide range of recharge-
8
able batteries technologies can be utilized in conjunction with variable

RESs [15]. Their main differences lay on the manufacturing technology
and materials.
– Lead-Acid Batteries: They are characterized by their low cost, high

reliability and efficiency. However, they have poor cycle-life (500-
2.000) and short lifetime (5-15 years) compared with other tech-
nologies. There are basically two types of Lead-acid batteries:
vented and valve regulated lead-acid (VRLA). Vented batteries are
less expensive and have larger cycle-life than VRLA batteries, but,
they require periodic maintenance and special conditions for trans-
portation and installation. On the contrary, VRLA batteries are
designed to be maintenance-free, they are smaller and lighter than
vented batteries. Additionally, VRLA batteries can be installed and
transported without considering vertical orientation, which allows
an easy transportation, installation, and storage. These character-
istics, make VRLA batteries more suitable for applications in re-
mote areas [38]. Among other applications Lead-acid batteries are
commonly used in automotive applications, data centers, telecom-
munications systems, UPS and another emergency and standby
power systems.
– Nickel-Based Batteries: There are mainly three types of Nickel-based
batteries, which differ one to the other by the material used for
the negative electrode; cadmium hydroxide (NiCd), metal alloy
(NiMH) and zinc hydroxide (NiZn). They are characterized by
a higher energy density and in some of them a larger cycle-life
than Lead-acid batteries but at the expense of higher cost (up to
10 times). Additionally, the efficiency of Nickel-based batteries is
lower than the efficiency of Lead-acid batteries. Among the nickel-
based batteries, NiCd is the most used for industry applications
especially for UPS [11], [36]. However, due to environmental is-
sues caused by the toxic cadmium, the use of NiCd has been de-
clined or even banned in some countries [11]. This problem of us-
ing toxic metals has been solved with the use of NiMH batteries,
however, the cycle-life of NiMH batteries is smaller (comparable
with the cycle-life of Lead-acid batteries). The main disadvantage
of Nickel-based batteries for the integration with variable RESs is
the memory effect [11]. The memory effect means that the battery
loses capacity when it is not fully discharged [35].
– Sodium-Sulphur Batteries (NaS): This is a promising emerging tech-
nology in the last decade. NaS batteries are characterized by high
efficiency and higher power density. They are economically and
environmentally feasible because they are made of inexpensive
9
and non-toxic materials [36]. Their main drawback is that the
should be operated at high temperatures (300 to 350 degrees Cel-
sius) which imply additional costs [11]. Therefore, this technology
is more suitable for large-scale stationary applications and not for
small-scale power systems such as microgrids [33], [36].
– Lithium-ion Batteries: This is an outstanding technology compared
with the others technologies in terms of many of its different fea-
tures such as higher energy and power density than Nickel-based
and Lead-acid batteries, higher efficiency than Nickel-based, and
higher cycle life than Lead-acid batteries. Additionally, Li-Ion bat-
teries do not have the problems with the memory effect, have low
self-discharge and are less toxic than Lead-acid and NiCd. The
main drawback of Li-Ion batteries is their high cost and the lim-
ited reserves of lithium-ion [11]. Additionally, Li-Ion batteries are
very strict on their limit of charge, they cannot be overcharged
because safety will be seriously compromised.
– Flow Batteries: This technology is characterized for storing at least
one of its liquid electrolytes on external tanks, from where the elec-
trolytes is pumped to a reactor in order to obtain electric current
from a reversible chemical reaction. The most popular technolo-
gies are the Zinc-Bromine (ZBR) batteries and the Vanadium redox
batteries (VRB). One of the main advantages of this kind of batter-
ies is that the stored energy is decoupled from the rated power
capacity, since the rated power depends on the size of the reactor
while the energy density depends on of the size of the auxiliary
tanks [33]. This fact makes the flow batteries more flexible and
easy to fit and customize for a wide range of applications. One of
the main problems with this technology is its low energy density,
therefore, it is required a large amount of space for applications
that require long-term support.
For small power systems where the ESSs play an important role, the key
element is the autonomy level provided by the ESSs. Based only on technical
reasons Lithium-ion batteries have the best performance in terms of lifetime,
energy density, power density, self-discharge, efficiency and become the best
candidate for small to medium power applications such as islanded MGs
based on variable RESs. However, whether the cost is a sensitive issue to
consider (for instance in remote villages in developing countries) Lead-acid
batteries and particularly VRLA batteries have the best compromise between
performance and cost [15], [37]. Because of that, Lead-acid batteries still
dominate the market for applications in islanded power systems and remote
villages, since in addition to reduced costs they also offer good availability
and easy transportation [26], [29], [38].
10
Given the above points, ESSs based on batteries, and more specifically
Lead-acid and Lithium-ion batteries will be considered in this work for ap-
plications in islanded MGs by considering specific limitations for charging
within a safe operating window.
1.3 Charge/Discharge Characteristic of Batteries

The name “battery” is often used for the basic unit where the electrochemi-
cal reaction takes place. To be more precise, the basic unit is the cell, and a
battery by itself is composed of a group of cells connected in series or par-
allel depending on the desired voltage and capacity [35]. In general, a basic
electrochemical cell is composed of two electrodes, one positive (cathode)
and one negative (anode), which are plunged into an electrolyte. The pro-
cess of charge and discharge implies movement of ions between cathode and
anode. During the process of discharge, the anode provides electrons and
suffer oxidation, while the cathode receives the electrons and it is reduced.
The process of charge reverses the movement of ions, which is facilitated by
the electrolyte [36]. Because of the changes in the electrochemical reactions,
the voltage of the cell and its internal resistance will vary during the process
of charge and discharge. This fact is due mainly to three factors such as:
(i) the state of charge (SoC), which is an indication of the amount of energy
stored in the cell, (ii) the current flowing through the cell, and (iii) the tem-
perature. The variations in the voltage of the cell differ from one technology
to other depending on the material used in the manufacturing process of the
cell. Therefore, different and particular signatures of the voltage vs. SoC can
be found for all the electrochemical batteries (i.e. the discharge signature in
Li-Ion is flatter than the Lead acid signature) [39]. Fig. 3 shows a typical
voltage vs. SoC curve for an electrochemical battery.
An ideal battery will have a charge/discharge curve extremely flat, how-
ever, in real batteries, the curve will have a non-linear behaviour as is shown
in Fig. 3. When the batteries are approaching their states of full charge or
full discharge their internal resistance tends to increase due to the polariza-
tion of the active materials, as a consequence, the curve shows a more sloping
profile [35]. Beyond certain values, the batteries cannot continue storing or
delivering energy at the risk of being degraded or being under unsafe opera-
ting conditions. A battery can operate within a safe operation region in which
the voltage signature is almost linear. The range of safe operation window
may involve potential differences from Li-Ion to Lead-acid, but a safe SoC
window is typically established between 20 % to 90 % [15], [38], [40]. These
values roughly correspond to the points where the voltage signature exhibits
a well-defined knee as can be seen in Fig 3. However, the SoC value could be
not the best option in order to define the operation limits, since the voltage
signature is also affected by the battery current and temperature. When the
11
27-11-2016 - 04-12-2016 27-11-2016 - 04-12-2016 27-11-2016 - 04-12-2016
Overcharge Safe Operation Window Over-discharge

Voltage
Vr
Battery
Battery Voltage
EOD
90 20
Charge SoC (% ) Discharge
SoC (%)
Fig. 3: Charge/Discharge curve of a Cell.
Table 1: Summary Table of Typical Voltage Values per Cell
Charge Cut-off Cell Voltage Discharge Cut-off

Voltage Nominal Voltage
Lead-acid 2.4 V 2V 1.75 V
Li-Ion 4.20 V 4.2 V 2.5-3.0 V
battery is being charged the current flow causes an increment in the voltage
and when the battery is being discharged the current causes a drop in the
voltage. Additionally, the temperature affects the internal resistance of the
cell [39].
Particularly, Lead-acid and Lithium-ion batteries are voltage-limiting de-
vices whose threshold values are determined specifically by the manufactur-
ers. Table 1 summarizes typical values defined by the manufacturers for the
charge and discharge cut-off values for both Li-Ion and Lead-acid cells. The
maximum value that a cell is allowed to reach is commonly known as the
regulation voltage (Vr ) and the minimum value is known as the end of dis-
charge value (EOD) or low-voltage disconnect (LVD), since beyond this point
the load has to be disconnected from the battery in order to avoid further dis-
charge [38]. Despite the similarities between Lead-acid and LI-Ion batteries,
there are some particular requirements that should be taken into account in
the process of charge, in order to enhance the performance and lifespan of
the batteries.
Charging Lead-acid
Lead acid batteries require a process of charge of at least two stages (com-
monly three), since either overcharge and undercharge may compromise the
lifespan and performance of the battery [41]. As stated in [38] “Appropri-
12
SoC (% )
SoC
Battery 100
Voltage
V_r
Battery
Current 80
Battery Voltage
Battery Current
50
0
EOD
20
0
0 Time 1 Time2 Time 3
Fig. 4: Charging/Discharging Characteristic of Lead-acid batteries.
ate charging is critically important to the life and performance of lead-acid

batteries, especially VRLA batteries”.
For applications within an islanded MG based on variable RESs the two
stages are: 1). A current-limited stage in which the battery as part of an ESSs
operating in the grid-forming role will absorb the surplus of energy in the
local grid. 2). A constant-voltage stage in which the battery voltage should
be kept constant by means of a topping charge in which the battery current
will begin to taper, while the battery continues being charged until saturation
[38], [41–43]. The second stage should start once the battery voltage reaches
the regulation value Vr .
A third stage denoted as float charge is recommended for Lead-acid bat-
teries when they remain on standby for long periods (more than 48 hours)
[35]. This stage basically compensates the self-discharge of a lead-acid bat-
tery and keeps the SoC [41]. This stage can be omitted when the battery
is not going to be in standby for long periods as usually happen in remote
power systems (islanded MGs), where one or two charge cycles occur in a
day [29], [43]. Fig. 4 shows the expected charging profile for a Lead-acid
battery.
In Fig. 4, from 0 to Time 1 the charger operates in stage 1 until the battery
voltage reaches the regulation value (Vr ) at Time 1. From Time 1 to Time 2
the charger operates in stage 2.
To sum up, in the process of charge there are important factors recom-
mended by the manufacturers that should be taken into account for ensuring
a proper performance and avoiding limiting the life of Lead-acid batteries as
is stated in the references [35], [38], [41–43].
• In the first stage of charge it is ensured about an 80 % of the total charge,

while the second stage complements the remaining charge.
13
• It is important to fully charge the battery between discharge cycles in
order to maximize the capacity and avoid fast degradation.
• The topping charge (constant-voltage charging) is the most effective

method for ensuring full charge of Lead-acid batteries.
• Since the Lead-acid batteries are voltage-limiting devices, the manu-

facturers recommend a charge controller with a constant set-point (the
voltage regulation Vr ), which is determined by the manufacturers.
• A seamless transition is required between the current-limited stage and

the constant-voltage stage.
• Since the process of charge is inefficient it is necessary to charge the

battery with more that 100 % of SoC (105 % to 110 %).
Charging Lithium-ion
Lithium ion batteries are also voltage-limiting devices but unlike the Lead-
acid batteries, the voltage per cell is higher as can be seen in Table 1. On
top of that, Lithium-ion batteries are less tolerant to accept overcharge than
Lead-acid batteries, since they are very strict in their maximum voltage limits.
Any operation beyond the limit compromises seriously their safety [41]. In
counterpart, Lithium-ion batteries do not need to be fully charged. In fact, a
partial charge is more recommended that full charge, and as a consequence,
the saturation stage can be discarded. Furthermore, any reduction in the peak
charge voltage can be reflected in a larger cycle-life. However, this decision
is always a trade-off, because the voltage reduction reduces the capacity.
Since the saturation stage is not required, the controller for the charger can
be simpler in comparison with the charge control of Lead-acid batteries. Ac-
tually, it is recommended to cut-off the battery current and keep the battery
voltage in the regulation value as short time as possible. This fact causes that
the battery voltage begins to droop, and the voltage stress is reduced [41].
Fig. 5 shows the charging curve that can be applied to Lithium-ion batteries,
by taking into account the aforementioned considerations.
In Fig. 5, at Time 1 the battery voltage reaches its maximum limit (Vr )
and the battery current is cut-off. The following list summarizes some of the
main features to take into account for charging Li-Ion batteries.
• They cannot be operated beyond the regulation value.
• Partial charge is better than full charge.
• Partial charge does not cause additional degradation.
• A constant-voltage controller is not required.
14
SoC (% )
100
V_r
Battery Voltage
Battery Battery
SoC
Current Voltage 80
Battery Current
EOD 50
0
20
0
0 Time 1 Time2 Time 3
Fig. 5: Charging/Discharging Characteristic of Lithium-ion batteries.
• A smaller regulation voltage means less stress and longer life but is
reflected in the battery capacity.
Discharging
In addition to a proper charge procedure, over-discharge should be avoided
for preventing degradation of the storage unit. For both Lead-acid or Li-
ion, deep discharge wears the battery down. Because of that, the depth of
discharge (DoD) of a battery should be limited to safe valued (DoD(%)=100-
SoC(%)). For Li-Ion and Lead-acid batteries, it is acceptable a maximum DoD
of 80 %, beyond that limit the operation of the battery is not recommended
because of wear. Similarly, the SOC or the DoD values are not the best indica-
tions to end up the discharge of the batteries. As aforementioned, the voltage
signature can also be affected by other parameters such as the current and
temperature. Therefore, the EOD should be the value to be taken into ac-
count to cut-off the battery discharge current. The EOD should be decided
based on the manufacturer’s guidance and the battery control should also
incorporate proper actions to avoid deeper discharge [38]. To illustrate, Figs.
4 and 5 show the expected behaviours for a deep discharge cycle (from Time
2 to Time 3). At Time 3, the batteries reach their corresponding EOD value
and the battery current is cut-off for avoiding further discharge.
1.4 Aggregated or Distributed ESSs-Motivation

One of the biggest discussions about the deployment of ESSs is whether they
should be distributed or aggregated. The traditional approach relies on a
holistic vision of the power system where the power balance service pro-
vided by the ESSs benefits the power system as a whole. Therefore, any
15
RES1 AGGREGATED LOAD
RES2
AGGREGATED ESS Microgrid
Fig. 6: Microgrid with aggregated ESSs an loads.
unbalance between generation and consumption caused by the variability of

RESs can be managed by a single or a small number of larger ESSs (Fig. 6).
The aggregated approach is characterized by simplified schemes of control
and coordination with other distributed generators and loads. Also the ag-
gregated approach ensures an efficient integration of variable RESs [17], [44].
However, aggregated approaches are frequently limited by the space, due to
the requirement of bulked installations. On top of that, they are not easily
scalable and reconfigurable for the integration of additional generation units
and increasing load requirements.
On the other hand, distributed approach dominates the current trend in
the integration of ESSs. They are characterized by small size distributed
storage units, which offer a flexible integration in terms of location restric-
tions, since they can be connected either near generation or near consump-
tion based on optimal considerations [33], [45], [46]. Also, they are flexible in
terms of easy dimensioning for various size requirements. Apart from that,
distributed ESSs are scalable and typically require reduced cost associated
with additional integration, replacement, maintenance and operation [11].
For the integration of distributed RESs, it is commonly suggested that the
storage units should be located on the same node and operated in conjunction
to individual RESs and local loads. In this way, it is possible to enhance the
voltage and power quality at the point of common connection [11], [44]. On
top of that, the ESSs can be optimally sized for particular conditions of gen-
eration and consumption, in which the couple formed by an ESS and a RES
(ESS+RES) can behave either as a baseload generator with constant output
power or a load-following generator looking for ensuring on-site generation
and consumption [47]. Here, is where the concept of the active generator
has emerged as an alternative in which the integration of variable energy re-
source is considered together with its own storage system and load control
in order to ensure a more stable power generation profile from RESs [48–50].
16
Microgrid
RES1 ESS1 LOAD1 RES2 ESS2 LOAD2
RES3 ESS3 LOAD3 LOAD4 ESS4 RES4
Fig. 7: Microgrid based on distributed active generators.
Fig. 7 shows a typical configuration of a MG based on active generators in

which the variable nature of each RES is smoothed by its own ESS. An active
generator can participate as a dispatchable source within a MG or alterna-
tively participate in the common bus regulation and supporting the power
balance of the whole microgrid.
The main drawback of a fully distributed approach based on active gen-
erators is, in fact, the elimination of the benefits derived from an aggregated
operation such as a more efficient use of the distributed resources [17]. To
be more precise, by considering individual and different consumption and
generation profiles, if the active generator (RES+ESS) is configured for sup-
porting only local consumption, it is possible to have the case in which si-
multaneously an ESS could be being charged while other could be being
discharged and the power generation from one RES could be being wasted
at the time it is required by another load.
In light of the above, it is desirable to combine the flexibility, scalability
and optimal sizing provided by a distributed approach with the efficient use
of the distributed resources derived from an aggregated approach. For that
reason, an appropriated management and coordination strategy that allows
distributed resources and loads cooperate between them, and behave as an
aggregated unit can be the best option for achieving high flexibility and an
efficient operation in a microgrid based on distributed ESSs (Fig. 8).
Alternatively, the coordination of the distributed resources can ensure a
reliable operation of the islanded MG by means of specific control actions
in each of the distributed units, which consider not only its own particu-
lar operating conditions but also the limitations and operating conditions of
the other DERs and the whole MG [51]. Simple actions such as reducing
or increasing the generation or adjusting the load demand in order to keep
the power balance can be addressed for a proper coordination. Likewise,
changing the operation mode for assuming the regulation of the common
17
Microgrid
RES2
RES1 ESS1 ESS ESS2
Coordination
Aggregated
Load
LOAD1 LOAD2
Fig. 8: Microgrid based on distributed resources with coordinated operation.
bus whenever needed can also be considered as cooperative control actions.
2 Coordination of Islanded Microgrids Based on

Distributed RESs and ESSs-Motivation
2.1 Operation of an Islanded Microgrid Based on RESs and
ESSs
An islanded MG is a complex system composed of heterogeneous DERs
which are integrated by means of power electronics interfaces with their own
local controllers (LC) as is shown in the simplified scheme of Fig. 9a. The
role that each distributed unit assumes determines its control mode and the
way in which it behaves determines its level of interaction with other units
inside the MG.
Two main roles (control modes) can be defined for the operation of dis-
tributed generators in a islanded MG; grid-forming or grid-following. The
grid-forming units are mainly responsible for controlling the voltage and fre-
quency, while the grid-following units are mainly responsible for injecting
or absorbing active/reactive power. The role of a distributed generator can
be assumed by means of grid-interactive or grid-noninteractive behaviours
(control methods) as proposed in [25]. A grid-interactive behaviour means
that the distributed generator participates actively in the regulation and con-
trol of the whole MG, while a grid-noninteractive operation means that the
distributed generator only takes care of its own particular control objectives
regardless the operation of the whole MG. The role of a distributed generator
takes an important place in the coordination of the MG, while the behaviour
18
2. Coordination of Islanded Microgrids Based on Distributed RESs and
ESSs-Motivation
PCC
Power LC LC Power
Converter Converter
MPPT
PV-RES LOAD Battery-ESS

(a)
PCC
CCM VCM
RES LOAD ESS

Grid-Following Grid-Forming
(b)
Fig. 9: MG based on ESSs and RESs: (a) Basic scheme, (b) Equivalent circuit.
determines the level of cooperation of the distributed generator in the MG.

For instance, when the ESSs are operated under a partial state of charge
(PSoC), they can assume the grid-forming role with a grid-interactive be-
haviour, since they are able to absorb or supply power in order to compensate
any unbalance between generation and consumption, while they are regulat-
ing the common bus. Because of this feature, provided by the storage sys-
tems, they are commonly assigned to be the master units, being responsible
for forming bus in the islanded power system (Fig. 9b) [31, 52]. Meanwhile,
it is expected to obtain from RESs the maximum amount of available energy
in order to make an efficient use of the renewable resource. Because of that,
the operation of variable RESs, such as WT and PV, is preferred under a
maximum point tracking algorithm (MPPT) [53–55]. RESs operating under
MPPT are a clear example of grid-following units with a grid-noninteractive
behaviour, since the amount of power fed to power system grid depends
more on weather conditions rather than on particular requirements of the
local power system [25].
The operation of an islanded MG under a partial-state-of-charge (PSoC)
is also a management strategy that has been commonly addressed in order to
maintain the operation mode of the MG as shown in Fig. 9b and avoid the use
of additional coordination actions [52]. However, the PSoC strategy requires
the use of batteries with oversized capacity for keeping the SoC value as far as
possible from the safe operation limits (Fig. 3). As a consequence, the initial
capital cost of the battery is increased. On top of that, when this strategy
is applied to Lead-acid batteries, they will require the periodic schedule of
19
full charge cycles (saturation), for avoiding the degradation of the Lead-acid
batteries. Apart from the cost, the saturation of the battery by itself is not
a problem, but this procedure implies that the battery should be put out of
service for at least 12 hours and up to 48 hours depending on the capacity of
the battery. What is more, in remote villages is not possible to ensure that the
amount of energy required for fully charge a large battery is available [41].
One alternative for avoiding a periodic fully saturated charge is the use of
Lithium-Ion batteries, which can be operated with partial charge without
any problem, but the cost of the battery becomes a bigger problem in cost
sensitive application such as remote villages, by considering that the prices
of Lithium-ion batteries is more than double the cost of Lead-acid batteries.
Even for an optimally sized MGs, the PSoC operation mode cannot al-
ways be ensured due to the unpredictable nature of variable RESs [56]. It
is possible to have the case in which the storage unit gets fully charged and
the generation is bigger than consumption, or conversely, the case in which
the amount of power generation is not able to satisfy the load requirements
and then the ESSs could be fully discharged. The appropriate use of the
energy storage system takes a significant relevance for islanded MGs with
ESSs based on batteries, since the well-being and lifespan of batteries rely on
specific and precise constraints for charging and discharging as was exposed
before. This is not a minor issue by regarding the cost associated with early
replacement and maintenance of batteries [57, 58]. In light of the above, a
basic set of rules can be defined to summarize the main aims to be taken into
account for a reliable operation of an islanded MG based on batteries and
RESs.
1. The power balance and regulation of the local power system should be
ensured under all the operating conditions.
2. Overcharge and deep discharge should be avoided by considering spe-

cific characteristics of the storage technology.
3. The RESs are more likely to operate under MPPT algorithms in order
to obtain from them their maximum energy.
4. The load should be fed whenever possible.
With respect the aforementioned goals, the first two are of strictly en-
forced, while the last two are more flexible and could be sacrificed at the
expense of accomplishing the first two. In other words, actions as active
power curtailment of the generation from RESs and load-shedding can be
considered within a coordination strategy in order to ensure a reliable and
safe operation of the power system [59–61].
20
ESSs-Motivation
MODE II
PCC
Vbat Vr CCM CCM

COORDINATION
RES LOAD ESS
Grid-Following Grid-Following
MODE I MODE III
PCC PCC
Coordination
CCM VCM VCM CCM
Strategy
RES LOAD ESS RES LOAD ESS
Grid-Following Grid-Forming Grid-Forming Grid-Following
Fig. 10: Transition between operation modes for proper charge of Lead-acid batteries in a MG.
Particular consideration for the operation with Lead-acid Batteries

Since Lead-acid batteries require a saturation stage for the charging process,
it is needed a change in the control mode of the ESS once the battery voltage
(Vbat ) reaches the regulation value Vr . For that reason, the ESS needs to
change its control mode from VCM to CCM in order to regulate the amount
of current drawn by the battery and keep the battery voltage in the regulation
value. This change in the control mode implies also a change in the role of
the ESSs and the way they are interacting in the microgrid. In this stage, the
ESSs will become grid-following units with a noninteractive operation, since
the ESSs leave the tasks related to the regulation of the common bus and
the power balance of the microgrid to assume the regulation of the battery
voltage.
Any change in the operation mode of each unit is reflected in a change in
the topological operation of the whole MG as can be seen in Fig. 10, where the
MG changes from (MODE I) to (MODE II) [62, 63]. In (MODE II) there is not
regulation of the common bus, then a coordination action must be enabled
to assign responsibility for regulating the common bus to another unit with
enough capacity. In the case of the MG shown in Fig. 9 the grid-forming role
can be assumed by the RESs as can be seen in the (MODE III) of Fig. 10. The
main task of a coordination strategy will be to ensure a seamless transition
between operation modes by considering particular operating conditions of
each distributed unit and the whole MG.
Particular consideration for the operation with Lithium-ion Batteries

Lithium-ion batteries are also a voltage limiting device in which the process
of charge should be terminated once the battery voltage reaches a cut-off
value. Unlike the process of charging Lead-acid batteries, the Li-Ion batteries
21
Maximum Power Point
PCC
Load Power comsumption Power
adjustment
P(W)
CCM VCM
Equilibrium point
PRES  PLOAD
RES LOAD ESS
Grid-Following Grid-Forming
VPV (V)
Fig. 11: Active Power curtailment of generation from RES for avoiding overcharge of the ESS.
do not need the stage of topping charge (saturation). Indeed, it is recom-

mended to avoid the saturation stage for reducing the stress of the battery
and also operation under partial charge is preferred for preserving the lifes-
pan of the batteries [41].
For avoiding overcharge, the battery current should be set to zero once the
voltage limit is reached. In this case, a similar strategy to the one explained
before based on changes in operation modes for Lead-acid batteries can be
used, but by setting the reference of the battery current at zero.
Since the saturation stage can be avoided, it is no necessary to change
the operation mode of the ESSs. Alternatively, the power generation of the
other distributed units can be adjusted until the power balance in the system
is achieved and the power drawn for the ESSs becomes equal to zero. To
clarify this scenario, the RESs can keep its role of grid-following unit, since
the ESSs would continue as the grid-forming unit, but the behaviour of the
RESs can change for grid-noninteractive operation (MPPT operation) to a
grid-interactive operation (off-MPPT mode) in which the power injected to
the MG can be dispatched or adjusted. This behaviour is illustrated in Fig.
11 for the MG shown in Fig. 9a. The main task of a coordination strategy will
be to ensure a proper coordination and changes in the behaviour of the RESs
by considering the amount of available power and the power requirements of
the whole MG.
The strategy presented in Fig. 11 can also be deployed in MG with Lead-
acid batteries. However, periodic cycles of saturated charge should be sche-
duled.
2.2 Coordination Architecture

The coordination architecture for distributed resources in an islanded mi-
crogrid corresponds to the set of actions, roles, and behaviours that the dis-
tributed resources should assume in order to ensure the fulfilment of com-
mon aims. The roles and responsibilities of the distributed resources are
22
ESSs-Motivation
assigned based on their particular operating conditions, capabilities, and lim-

itations and based on a collective perspective in which the requirements and
operating conditions of the whole microgrid should be considered [64].
The coordination architecture can be deployed by means of decision-
making (DM) mechanism, which execute automatically a computer program
for adapting the operation of the distributed resources in accordance with
changing circumstances. The DM uses the information obtained from the en-
vironment (operating conditions of the whole MG and particular operating
conditions of each distributed unit), to make the decisions and request spe-
cific actions from the distributed resources [65]. The decision-making process
is performed by an inference engine, which contains assertions and condi-
tions about the expected operation of the microgrid. From a metaphorical
point of view, this is equivalent to the process performed by an expert hu-
man being. Because of that, a knowledge-based or expert system can be con-
sidered for managing the complexity of a MG and define the coordination
architecture from a higher level of abstraction.
Decision-Making based on Expert Systems

An expert system is composed of a knowledge base and an inference engine.
The knowledge base contains facts to be considered for a proper operation
of a MG. This is all the information concerning the common aims of the
MG and the particular limitation and requirements of the DERs (i.e. the
constraints and requirements for charging/discharging battery-based ESSs).
On the other hand, the inference engine contains all the information about
the conditions and actions that should be performed by the DM units, which
usually can be represented by means of IF-THEN rules (i.e. If the voltage in
an ESSs based on VRLA batteries has reached the regulation value THEN the
ESSs should operate in CCM).
From the approach proposed in this project, the different stages for charg-
ing and discharging properly the ESSs based on batteries will determine the
different operation modes and the coordination of the whole MG. In this
sense, all the knowledge and information exposed before, which is sum-
marized in subsection 2.1 will define the conception of the expert system
responsible for the coordination of the microgrid.
A stable and safe operation of a microgrid can be defined within a lim-
ited set of operation modes. The microgrid and the DERs inside the MG can
continue doing the same action or performing the same role until the opera-
ting conditions make them change [3]. For instance, in a first stage the ESSs
may operate within the safe operation window (Fig. 3) and keep the grid-
forming role in a continuous way. However, if the maximum and minimum
limits of charge are reached, it is required to define some coordinated actions
which lead to changes in the operation mode of the MG for avoiding over-
23
charge or over-discharge and keep the power balance in the MG. The MG
can continue operating at this stage until there is a change in its operating
conditions. In this sense, finite state machines can provide a simple and in-
tuitive mechanism for the deployment of the coordination architecture based
on DM units, since the operation mode of the DERs and the whole micro-
grid can be bounded within a finite number of operation modes, which are
mainly determined by the different stages for a proper charge of ESSs based
on batteries [65], [3].
The use of finite state machines allows to keep the operation of the MG
within a known set of states, and the transitions between operation modes
for a proper coordination can be defined in accordance with a simple and
intuitive rule base, which considers the different stages for charging battery-
based ESSs. This fact allows to addressing the coordination problem of the
islanded MG from a higher level of abstraction by means of simple rules as an
expert does and in this way hide details that are not important for coordina-
tion purposes. For instance, by assuming that the distributed resources oper-
ate properly in accordance with their roles and control operation modes, and
the coordinated actions can be managed properly at each unit. Nevertheless,
it is important to understand how the controllers at the levels of abstraction
immediately above and below operate in order to ensure a proper integration
of the coordination strategy [66]. In the particular case of a MG composed
by distributed generators, it is important to be acquainted about the control
loops and the whole control structures involved in the MG in order to ensure
a proper matching between coordination commands and control signals.
Coordination Within a Hierarchical Control Frame for Microgrids

In order to standardize the control of microgrid and managing its complexity
from different levels of abstraction, a hierarchical control approach has been
defined in four levels typically [13], [51].
The first level deals with inner control loops responsible for regulating
current and/or voltage based on local measurements and ensuring the local
stability of the power electronic interface. The inner control defines the role of
the distributed generation in the MG (grid-forming or grid-following) [67].
The second level is termed as primary control which is also a local stage
that allows power electronics converters to operate independently but at the
same time may allow a cooperative interaction between them. A good exam-
ple of primary controllers enabling cooperative operations in MGs are the
droop control loops, which allows the parallel connected grid-forming unit
to share autonomously the power balance in the local grid at the expenses of
sacrificing the voltage and frequency regulation [68]. The third level (secon-
dary control) ensures accurate regulation of the global system variables by
restoring any deviation caused by the primary controllers (i.e. common bus
24
ESSs-Motivation
voltage/frequency, reactive power sharing, power quality in certain points of

the microgrid, etc.) [69–71]. Finally, the tertiary control commonly operates
as an energy management system (EMS) who defines power references for
dispatchable sources by considering an optimized power flow between the
MG and the main grid and based on specific interests related to efficiency
and economic operation of the MG [72]. Commonly, secondary and tertiary
control require global information to perform their task, because of that they
require the deployment of additional communication infrastructure.
The definition of the of the hierarchical levels is more related to specific
tasks and the bandwidth of each control level, rather than the way in which
the task are performed. That is, the tasks can be performed by a central
controller, can be decentralized among all the distributed units, or they can
be based on hybrid centralized/decentralized solutions [73].
Currently, the dominant trend is the decentralized approach, especially
for tasks which require a higher speed (wider bandwidth) such as the inner
control loops and the primary controllers. Because of that, the decentralized
or even hybrid approach is becoming almost mandatory in order to reduce
the dependence of high bandwidth communication [73–75].
Since the inner control defines the role of the distributed resources (con-
trol modes) and the primary control allows a cooperative interaction between
units (i.e. droop control loops) these two levels should be considered for the
definition of a coordinated strategy. By acting on the first level it is possible
to modify the role of each DER in the microgrid. Meanwhile acting on the
second stage will allow defining cooperative behaviours for the achievement
of particular purposes. A coordination architecture based on roles and be-
haviours may operate independently and in parallel to the secondary and
tertiary control layers as is shown in [24, 76].
Any, coordination actions or cooperative behaviour (including adaptive
adjustment of primary controllers) can be performed either by centralized
or distributed DM mechanism. While, centralized architectures, as shown
in Fig. 12a, are preferred for small-scale and local area power system with
reliable communication systems [2, 76]. Decentralized and specifically dis-
tributed approaches, as shown in Fig. 12b, are preferred in applications that
require high degrees of scalability and reconfiguration and for MG imple-
mented over wide areas where additional communication infrastructure is
not always suitable [3, 75, 77].
Background in Coordination Strategies for Microgrids

Different coordination strategies have been proposed for managing the com-
plexity in the operation of islanded microgrids. Approaches based on central
control units take advantages of the global perception of the MG, which is
provided by a dedicated communication network with all the distributed
25
Microgrid Central Controller
&
Coordination (DM)
Communication Channel
RES1 ESS1 LOAD RES2 ESS2
MPPT LOAD MPPT

Control Primary &
and
LC LC Metering LC LC Inner-control
Power
Convertertion
Stage
PCC
(a)
RES1 ESS1 LOAD RES2 ESS2
MPPT
DM MPPT
LOAD
Control
Primary &
DM LC DM LC and DM LC DM LC Inner-control
Metering
Power
Convertertion
Stage
PCC
(b)
Fig. 12: Islanded MG control schemes: (a) Centralized, (b) Distributed.
units, in order to provide a stable, multi-objective, and optimized coordina-

tion [31, 78, 79]. In [60] the coordination is managed by a secondary cen-
tral controller which adjust the power generation from dispatchable micro-
sources in order to ensure the power balance in the islanded system. Mean-
while, a single ESS based on batteries is responsible for forming the power
grid and compensating any transitory mismatch between generation and con-
sumption. As a result, the operation of the ESS is kept within a partial SoC
value (≈ 50%).
On the other hand, distributed approaches use the values of the voltage
and frequency in the islanded power grid for enabling a kind of communi-
cation between distributed units without the use of dedicated communica-
tion channels. For ac microgrids the most common approaches use multi-
segments adaptive power/frequency curves such as in [80–84]. In [80–82] the
interaction between battery-based ESSs and a RESs is determined by the tran-
sition between voltage control mode and power control mode (PCM) in the
primary controllers. To do that, the droop control loops include integral ac-
tions instead of simple droop with proportional loops, which allow the units
to activate a PCM operation under specific circumstances defined by the coor-
dination strategy. For instance, in [80,81] the power/frequency curves are de-
fined based on the relation between generation and consumption, rather that
the safe SoC limits. Unlike, the work presented in [82] determines the tran-
26
ESSs-Motivation
sitions between PCM and VCM based on SoC threshold values. In [83, 84],
there are not changes in control modes, since the ESS will always operate in
VCM as grid-forming units. Once the threshold values are reached, the ESS
generates a change in the pattern of frequency, which is recognized by the
RESs and it is used as information for adjusting the power generation bellow
the maximum power point. Similar approaches but adjusting the voltage in
the common bus have been proposed for dc MGs such as in [85].
The coordination problem in islanded microgrids based on variable RESs
and ESS based on batteries is still open to research, since the coordination
methods which consider complete charge profiles for ESS has been seldom
discussed. On top of that, little work so far has been found for the coordi-
nation of distributed ESSs in islanded microgrids. The proposal in this work
tries to consider the limitations and charging requirements for ESSs based
on batteries and how the charging process can harmonize with a reliable
operation of the microgrid even for microgrids with distributed ESSs.
2.3 Cooperative Behaviours

While a pure coordination approach is responsible for arranging the roles
and control modes of the distributed resources in a microgrid, a cooperative
approach will determine the degree of participation and efforts provided by
the distributed resources in order to achieve goals of common interest. Co-
operative behaviours can be defined either in grid-forming units or in grid-
following units operating under grid-interactive control methods. A clear
example is the use of adaptive droop control loops for adjusting the amount
of power shared for each grid-forming unit, in order to achieve common goals
by considering particular capacities of each unit [2, 3, 76, 86]. Additionally,
grid-following units can also perform cooperative behaviours by adjusting
the amount of power injected or absorbed, in order to balance the power of
the whole microgrid [2].
Two kinds of cooperative behaviours are defined in this project. The first
one is oriented to the equalization of the state of charge of distributed energy
storage systems, and the second one is the adjustment of the power delivered
by RESs in a proportional ratio to their maximum power availability.
SoC equalization
Within a battery string, it is impossible to ensure that all the cells behave
identically due to differences in the manufacturing process. Because of that,
some cells can be overcharged while others can be uncharged during the
charging process. As a consequence, some cells in a battery array may operate
out of the safe operation window shown in Fig. 3. This fact causes uneven
degradation of the cells what in turn is reflected in a faster degradation of
27
the whole battery array [87]. In order to overcome this problem, the battery
manufacturers offer some kind of cell balancing or circuits which perform
the SoC equalization process [38], [41].
The same principle of SoC equalization between cells in a battery array
can be applied for distributed energy storage systems based on batteries. In
this case, advantages such as avoiding the uneven degradation of the battery
arrays can be extrapolated for a distributed battery-based ESSs.
The SOC equalization problem between distributed energy storage sys-
tems has been widely addressed recently in the literature by means of ad-
justments in the power-sharing control loops. Some strategies are based on
purely distributed approaches which use adaptive curves such as in [77, 88,
89], others rely on centralized approaches such as in [90, 91], and some oth-
ers approaches rely on decentralized software agents which achieve the SoC
equalization based on average consensus algorithms such as in [92–95]. The
vast majority of the aforementioned approaches fall into the error of consider-
ing the SoC equalization as an independent issue, without any contextualiza-
tion for their applicability in a real MG. This issue has rarely been considered
as part of a complete coordination strategy for MGs.
When the SoC equalization is integrated as a cooperative behaviour into
a coordination strategy, some additional advantages become evident, which
enhanced the overall behaviour of a coordinated MG as is proposed in [1–3,
30, 76, 86, 96]. Among others, Fig. 13 illustrates some of the main advantages
of performing the SoC equalization as part of a coordination strategy. The
following list summarizes the main advantages.
• Unification of the discharge/charge profiles, which allows dealing with

the distributed ESSs as a single ESS [2]. Fig. 14 shows a clear example
in which the SoC profiles of distributed ESSs with different capacities
(5 and 10 Ah) behave like the SoC profile of an aggregated ESS with a
total capacity of 10 Ah.
• Avoids uneven degradation of the distributed energy storages caused

by unequal stress factors:
– Cycle Depth as can be seen in Fig. 15.

– DoD as can be seen in Figs. 13 and 15.
– Charging/Discharging rate as can be seen in Fig. 16. for instance
in an islanded MG composed by two distributed ESSs one of them
changes to a topping charge stage and the other assumes com-
pletely the power balance in the islanded MG.
• Overall reduction in the DoD, which is reflected in larger cycle-life for

the whole distributed ESSs.
28
ESSs-Motivation
Fig. 13: Comparative curve between SoC profiles with and without equalization [1]. Dashed
lines show the SoC profiles without equalization and the continuous lines shows the SoC profile
by performing the SoC equalization.
• Faster overall charge of the distributed ESSs.
• Smooth transition between operation modes for the MG [1].
• Simultaneous changes between stages of charge, which is reflected in

and easier operation and definition of the coordination strategy.
Proportional Active Power Curtailment

Another fact to be considered as a cooperative behaviour is the operation of
RESs as grid-interactive units in the case when the generation is larger than
consumption and the energy storage system is not able to store more energy.
Therefore, the generation from RESs should be adjusted to match the power
consumption as is shown in Fig. 17. This action allows to keep the power
balance and avoid that the batteries get overcharged. Here, the problem to
solve is determining the power-sharing proportion of each RES.
By considering variable RESs, an equal power sharing is not always the
best option, since the generation capacity of each unit may vary from one to
other. A possible solution is to adjust the power contribution of each unit
as a fraction of its maximum available power (PRES = βPRES MPPT), where
(0 ≤ β ≤ 1).
The coordination strategies proposed in this project include cooperative
approaches, in which the basic criteria for determining the amount of power
shared of each RES is based on a proportional relation with its maximum
29
SoCBat2
SoCBat (C=5Ah)
(C=15Ah) SoCBat1
(C=10Ah)
(a) (b)
Fig. 14: SoC profiles for an islanded MG with: (a) aggregated ESS with a capacity of 15 Ah,
(b) distributed ESSs with capacities of 10 Ah and 5 Ah and SoC equalization. The figures were
obtained from the work presented in [2].
Cycle Depth=39%
Cycle Depth=56%
Cycle Depth=27%
SoCBat2
SoCBat2
SoCBat1 SoCBat1
(a) (b)
Fig. 15: Comparison of cycle depth for an islanded MG: (a) without SoC equalization (b) with
SoC equalization. The figures were obtained from the work presented in [2].
Imax
Current (A)
Imax
Current (A)
0
Time
Imax
0 (a)Time
Current (A)Current (A)
Imax
0
Time
0
Time
(b)
Fig. 16: Comparison between charging current behaviours in distributed ESSs: (a) without SoC
equalization (b) with SoC equalization. The figures wer obtained from the work presented in [3].
30
3. Thesis Contribution
Maximum Power Point
Adjusted Output Power

PRES1
P (W)
Wind Load Power comsumption

Turbine
Turbine speed, h , (rad/sec)
Maximum Power Point
+
P(W)
PRES 2
P(W)
Time (sec)
Adjusted Output Power
Photovoltaic
Generation
VPV (V)
Fig. 17: Operation of RESs as gird-interactive units (off-MPPT mode). Here, the power genera-
tion is adjusted to match the power consumption.
power point. That is, the power generator with more power available will
contribute with more power than the others and so on [2, 3, 76, 96].
Another possibility is to determine the sharing proportion of each unit
based on improving the overall system efficiency in the energy conversion
process. It is well known that the efficiency of the power conversion stage
is reduced when it operates at low power rates [97]. Therefore, the sharing
proportion can be determined by optimal overall system efficiency as is pro-
posed in [98]. This strategy has not been considered in this thesis but can be
considered as a part of future work.
3 Thesis Contribution
This thesis addresses the coordination of distributed generation units for
small-scale islanded microgrids based on variable renewable energy sources
and battery-based energy storage systems. The coordination strategies look
for ensuring the reliable and safe operation of the islanded power system by
assigning specific responsibilities and actions to the distributed resources in
accordance with their particular limitations and capabilities.
The coordination strategies proposed in this thesis are based on specific
requirements and characteristics for charging and discharging battery-based
energy storage systems. In this sense, three basic stages for the operation
of the ESSs are considered as the base for the definition of coordinated ac-
tions within the islanded microgrid: (i) The ESS is operating under a partial
SoC, therefore, it is able to be charged or discharged in accordance with
the requirement of the microgrid. (ii) The ESS reaches its maximum limit
31
for storing energy, therefore, the generation should be adjusted for ensur-
ing the power balance in the microgrid. Meanwhile, proper charge profiles
are ensured based on particular requirements of the battery-based ESS. (iii)
The ESS reaches its limit for delivering energy to the microgrid, therefore,
the load consumption of the microgrid should be adjusted in order to avoid
further discharge of the storage unit beyond safe operating limits.
Given the above points, the operation of the microgrid can be conside-
red within a finite number of operation modes or states in a system level,
which are determined by specific combinations of operation modes of the
distributed resource [3]. As a consequence, the coordination strategies are
deployed by means of decision-making units based on finite state machines,
which can easily synthesize the required charge profiles for battery- based
ESSs.
Finite state machines are probably the most used tool for coordinating
multi-robot systems, since they provide a simple way of managing the com-
plexity of the system. This thesis adopts concepts used for coordination ar-
chitecture of multi-robot systems such as robot soccer teams. The terms of
roles and behaviours have been adopted for specifying the control modes,
and level of interaction of each distributed unit in the operation of the micro-
grid respectively. In fact, a coordination architecture based on a metaphoric
adaptation of coordination strategies previously proposed for robot-soccer
scenarios has been proposed in this thesis.
For ensuring a stable operation of the islanded microgrid under the dif-
ferent stages or operation modes considered within the coordination archi-
tecture, a small-signal stability analysis has been performed for islanded ac
microgrids by considering different combinations of operation modes at the
distributed energy resources. The proposed analysis considers the effect of
constant power loads and constant power sources in the dynamic perfor-
mance of a microgrid and allows to determine the values of the droop coef-
ficients for enabling the power sharing between grid-forming units based on
stability constraints.
Thanks to the stability analysis, it is possible to determine safe ranges
for adjusting the values in the droop coefficients, which enables the imple-
mentation of adaptive power sharing functions. The adaptive functions allow
differentiating the amount of power shared by each unit in accordance with
its capacity or based on the level of cooperation determined by the cooper-
ative strategy. This fact enables the definition of cooperative behaviours by
means of adaptive functions within the coordination strategy such as: (i) SoC
equalization. (ii) uniformity in the SoC profiles between distributed battery-
based ESSs. (iii) proportional power contribution from variable RESs.
The use of cooperative behaviours for achieving SoC equalization and the
uniformity in the SoC profiles allow treating distributed storage systems like
a single aggregated storage system, from the point of view of coordination
32
4. Thesis Objectives
strategy. This because the processes of charging and discharging are car-
ried out almost simultaneously. This fact allows an easy definition of unified
actions for coordination in the microgrid and is reflected in uniform degra-
dation and stress for the distributed ESSs.
Expert systems have been used not only for defining the rules and condi-
tions for the finite state machines in the decision-making mechanisms. Ad-
ditionally, knowledge-based fuzzy inference systems have been proposed for
obtaining the adaptive functions which allow the SoC equalization in dis-
tributed coordination strategies. The fuzzy system provides an intuitive and
qualitative way of defining the adaptive curves for cooperative behaviour,
for instance, the adjustment of the droop control loops in accordance to par-
ticular operating conditions in each unit. Since the fuzzy system is based
on qualitative knowledge, it can be easily adapted for ac or dc microgrids.
What is more, a fuzzy inference system is able to manage different control
objectives simultaneously as is shown in [86].
In this thesis different centralized and distributed strategies have been
proposed for a proper coordination of islanded microgrids based on variable
RESs and distributed battery-based ESSs. The proposed architectures can be
easily adapted for ac and dc microgrids since the principle of the architecture
is based on qualitative knowledge.
The proposed coordination architectures have been validated by either:
simulations or experiments. A hardware-in-the-loop (HiL) platform based
on Simulink models and dSPACE are chosen as the simulation environment
for dc MGs. Experimental setups in the microgrid laboratory are used to
verify the coordination architectures in islanded ac microgrids.
4 Thesis Objectives
The research objectives of this project are listed below:
• To investigate the feasibility and advantages of integrating different dis-

tributed energy storage systems in combination with distributed renew-
able energy sources for small-scale islanded microgrids.
• To define suitable power conversion stages and inner control loops for
integrating battery-based distributed energy storage devices to islanded
microgrids.
• To propose control strategies for coordinated operation of distributed

energy storage systems based on batteries and variable renewable energy
sources in islanded ac and dc microgrids.
• To develop centralized and decentralized control strategies as well as

energy management algorithms for achieving stored energy balance
33
among distributed energy storage units.
• To develop different scalable and expandable coordination strategies
for islanded microgrids based on distributed energy storage systems
and variable renewable energy sources, taking into consideration both
the storage and the grid side (energy storage limitation and microgrid
stability).
• To integrate the coordination strategies to the hierarchical control struc-
ture for microgrids.
• To develop coordination and management algorithms according to the
power limitation of each power electronics converter and capacity of
charge of each energy storage unit. The proposed strategies will con-
sider the particular characteristics of each ESS and their requirements
for the proper process of charge/discharge.
5 Thesis Outline
This thesis is organized as follows:
Chapter II: presents the first paper, published in IEEE Transactions on

Smart Grid. This paper proposes a distributed coordination strategy for a
low voltage islanded dc microgrid, which considers changes between ope-
ration modes in the distributed energy resources for ensuring appropriate
charge profiles for distributed energy storage systems based on lead-acid
batteries. The proposed coordination strategy rely on a multi-objective fuzzy
logic based adaptive function for adjusting the value of the virtual resistance
in the droop control loops in order to balance the stored energy between dis-
tributed battery-based energy storage systems, and at the same time reduce
the voltage deviation at the common dc bus. The units are self-controlled by
using local variables only, hence, the microgrid can operate without relying
on communication systems. Results from hardware in the loop (HiL) show
the feasibility of the proposed method. The proposed coordination strategy
is tested in the microgrid laboratory with dSPACE based HiL setups.
Chapter III: contains the second paper, presented in IEEE Energy Con-
version Congress and Exposition (ECCE) 2015. This paper proposes a small-
signal stability analysis for the use of adaptive droop control loops in is-
landed ac microgrid composed of distributed energy storage systems and
variable renewable energy sources. The performed analysis considers dif-
ferent stages and topological circuit operation modes of the microgrid, for
providing the appropriate charge profiles to distributed energy storage sys-
tems based on lead-acid batteries in ac microgrids. The proposed analysis
34
5. Thesis Outline
allows the adequate selection of values and range for adjusting the droop co-
efficients based on adaptive functions, in order to ensure the stability of the
microgrid under different topological operation modes.
Chapter IV: presents the third paper, which has been submitted to IEEE
Transactions on Power Electronics. This paper proposes a distributed ar-
chitecture for assigning roles to distributed units in islanded ac microgrids.
The distributed units are self-controlled and coordinated by means of dis-
tributed decision-making units which rely only on limited information about
the global state of the microgrid, provided by a bus-signalling method with-
out the use of additional communication channels. The proposed strategy
considers different stages for charging properly storage units based on lead-
acid batteries and includes a distributed method based on a fuzzy infer-
ence system for equalization of the state of charge (SoC) between distributed
energy storage systems. The SoC equalization allows a smooth and uniform
transition between charging stages for the distributed ESSs. The proposed
control strategy is tested experimentally in the microgrid laboratory.
Chapter V: presents the fourth paper, accepted for publication in the IEEE
Transactions on Power Electronics. This paper proposes a centralized con-
trol architecture based on tactics, roles, and behaviours for a coordinated
operation of islanded ac microgrids. The architecture is inspired by a robot-
soccer coordination strategy, which is used from a metaphoric point of view
in order to structure the coordination of the distributed energy resources for
ensuring a reliable operation of the islanded system and appropriate charge
profiles for distributed energy storage systems based on lead-acid batteries.
The proposed architecture takes advantages of the global perception pro-
vided by the centralized control agent to allow greater flexibility in operating
modes and transitions between different operating conditions. The proposed
control strategy is tested experimentally in the microgrid laboratory.
Chapter VI: presents the fifth paper, presented at the IEEE First Interna-
tional Conference on DC Microgrids (ICDCM) 2015. This paper presents a
centralized strategy for equalizing the state of charge of distributed energy
storage systems in an islanded dc microgrid. On top of that, the proposed
strategy allows uniform SoC profiles between distributed battery-based ESSs
based on a proportional power sharing in accordance with the capacity of
each storage system. The proposed strategy is based on a simple algorithm
denoted as equalization algorithm, which modifies the charge or discharge
rate by weighting the virtual resistor of local droop control loops at each dis-
tributed energy storage system. The proposed strategy can be used as an
additional function within a coordination strategy Hardware in the loop re-
sults show the feasibility of the proposed method. The proposed coordination
35
strategy is tested in the microgrid laboratory with dSPACE based HiL setups.
Chapter VII: contains the sixth paper, accepted for publication in the
IEEE Transactions on Power Electronics. This paper adapts the equalization
method presented in the previous section to be implemented in an islanded
ac microgrid. The equalization method is incorporated within a complete
coordination strategy, which relies on a central control unit which works as
a supervisory stage for avoiding overcharge or over-discharge on battery-
based ESSs. The proposed strategy provides a more suitable charge profile
for Li-Ion batteries. Experimental results validate the proposed coordination
strategy.
Chapter VIII: contains the conclusion and summarizes the main contribu-
tions. Additionally, this part presents perspectives for future research.
36
Part II
Paper 1
37
Intelligent Distributed Generation and Storage
Units for DC Microgrids - A New Concept on
Cooperative Control without Communications
Beyond Droop Control
Nelson L. Díaz, Tomislav Dragičević, Juan C. Vasquez,

and Josep M. Guerrero
Abstract
Low voltage dc microgrids have been widely used for supplying critical loads, such
as data centers and remote communication stations. Consequently, it is important
to ensure redundancy and enough energy capacity in order to support possible incre-
ments in load consumption. This is achieved by means of expansion of the energy
storage system by adding extra distributed energy storage units. However, using
distributed energy storage units adds more challenges in microgrids control, since
stored energy should be balanced in order to avoid deep discharge or over-charge in
one of the energy storage units. Typically, voltage droop loops are used for inter-
connecting several different units in parallel to a microgrid. This paper proposes a
new decentralized strategy based on fuzzy logic that ensures stored energy balance,
for a low voltage dc microgrid with distributed battery energy storage systems, by
modifying the virtual resistances of the droop controllers in accordance with the state
of charge of each energy storage unit. Additionally, the virtual resistance is adjusted
in order to reduce the voltage deviation at the common dc bus. The units are self-
controlled by using local variables only, hence, the microgrid can operate without
relying on communication systems. Hardware in the loop results show the feasibility
of the proposed method.
The paper has been published in the

IEEE Transactions on Smart Grid Vol. 5, No. 5, pp. 2476-2485, 2014.
39
40
Part III
Paper 2
41
Stability Analysis for Isolated AC Microgrids
Based on PV-Active Generators
Nelson L. Díaz, Ernane Alves Coelho, Juan C. Vasquez,

Abstract
The current trend in isolated microgrids is oriented to distributed renewable energy
generators, such as photovoltaic (PV) generators and their corresponding distributed
energy storage systems (ESS) as an unit denoted as active generator (PV+ESS).
In an isolated microgrid, every distributed generator (ESS or PV) may operate in
two operation modes (grid-forming or grid-following) in accordance to specific op-
erational conditions. Nevertheless, at least one distributed generator should assume
the regulation of the common bus, and seamless transition between operation modes
should be ensured for all the distributed generators. In particular, ESS’s based on
batteries require at least two different mode of charge. As consequence, the topologi-
cal operation mode of the microgrid is affected by the changes of the operation mode
of each distributed generator. Typically, droop control loops are used for intercon-
necting several different distributed generators in parallel to a common bus, whose
parameters determine the stability and damping of the microgrid operation. In this
paper, a small-signal stability analysis is applied to an isolated AC microgrid com-
posed of (PV+ESS) active generators, regarding three different operation modes of
the microgrid. The proposed analysis allows the designers an adequate selection of
the droop coefficient in order to ensure the stability of the microgrid under different
topological operation modes.

Proceeding of 2015 IEEE First International Conference on DC Microgrids
(ECCE 2015), pp. 4214-4221, 2015.
43
44
Part IV
Paper 3
45
A Communication-less Distributed Control
Architecture for Islanded Microgrids with
Renewable Generation and Storage
Nelson L. Díaz, Juan C. Vasquez, and Josep M. Guerrero
Abstract
For reliable operation of an islanded microgrid, at least one of its distributed re-
sources should assume the responsibility of forming the off-grid power system. This
responsibility is usually assumed by energy storage systems based on their capability
of compensating the unbalance between generation and consumption. However, the
storage units may lose their capability of regulating the power grid when maximum
and minimum limits of charge are reached. Under these conditions, the regulation of
the power grid may be assumed by another unit with enough capability or the power
balance should be adjusted coordinately. This paper proposes a coordination archi-
tecture for islanded ac microgrids, which considers the appropriate charge profiles
for battery-based energy storage systems. The architecture is based on distributed
decision-making mechanisms, which use only local measurements for determining
the operation mode of each unit independently. The coordination relies on a bus-
signalling method, which enables the distributed units to have a global perception
about the operation of the microgrid, without any communication infrastructure.
The proposed architecture includes cooperative operation between distributed energy
storage systems for equalization of charge, which allows a uniform transition between
appropriate charging stages. Experimental results validate the proposed strategy un-
der different operational conditions in the microgrid.
The paper has been submitted to the

IEEE Transactions on Power Electronics 2016.
47
48
Part V
Paper 4
49
A Robot-Soccer-Coordination Inspired Control
Architecture Applied to Islanded Microgrids
Nelson L. Díaz, Jose G. Guarnizo, Martin Mellado, Juan C. Vasquez,

Abstract
Nowadays islanded microgrids present a high interest due to the increasing penetra-
tion of renewable energy resources, especially in remote areas, or for improving the
local energy reliability. A microgrid can operate in grid-connected or islanded mode,
being necessary the use of energy storage systems under islanded operation, in or-
der to ensure the generation/consumption power balance and smooth uncertainties in
the dynamics of the renewable energy sources. Particularly, in islanded operation at
least one of the distributed energy resources should assume the regulation of the com-
mon bus. In a microgrid, every distributed energy resource may be able to cooperate
with the grid regulation in accordance to its particular operational conditions. In
this sense, a centralized unit with a global perception of the load demand, the power
provided by the renewable energy sources, and the storage capacity of the energy
storage systems, may ensure proper and reliable operation of the microgrid. This
paper proposes a structured architecture based on tactics, roles and behaviors for a
coordinated operation of islanded microgrids. The architecture is inspired on a robot
soccer strategy with global perception and centralized control, which determines the
changes among operation modes for the distributed energy resources in an islanded
ac microgrid.
The paper has been accepted for publication in the

IEEE Transactions on Power Electronics, Early Access,
DOI. 10.1109/TPEL.2016.2606653, 2016.
51
52
Part VI
Paper 5
53
Energy Management System with Equalization
Algorithm for Distributed Energy Storage Systems
in PV-Active Generator Based Low Voltage DC
Microgrids
Nelson L. Díaz, Adriana C. Luna, Juan C. Vasquez, and Josep M. Guerrero
Abstract
This paper presents a centralized strategy for equalizing the state of charge of dis-
tributed energy storage systems in an islanded DC microgrid. The proposed strat-
egy is based on a simple algorithm called equalization algorithm, which modifies the
charge or discharge rate by weighting the virtual resistor of local droop control loops
at each distributed energy storage system. The proposed strategy, can be used as an
additional function of the microgrid energy management system where the state of
charge of distributed ESS is equalized within a determined window of time. Finally,
real-time simulation results of a low voltage DC microgrid are presented in order to
verify the performance of the proposed approach.

Proceeding of 2015 IEEE First International Conference on DC Microgrids
(ICDCM 2015), pp. 293-298, 2015.
55
56
Part VII
Paper 6
57
Centralized Control Architecture for Coordination
of Distributed Renewable Generation and Energy
Storage in Islanded AC Microgrids
Nelson L. Díaz, Adriana C. Luna, Juan C. Vasquez, and Josep M. Guerrero
Abstract
The coordinated operation of distributed energy resources such as storage and gen-
eration units and also loads is required for the reliable operation of an islanded mi-
crogrid. Since in islanded microgrids the storage units are commonly responsible for
regulating the voltage amplitude and frequency in the local power system, the coordi-
nation should consider safe operating limits for the stored energy, which prevents fast
degradation or damage to the storage units. This paper proposes a centralized control
architecture, applicable for local area power systems such as a small-scale microgrid.
The centralized architecture is based on three supervisory control tasks which con-
sider: active power curtailment of generation for avoiding overcharge of the storage
units, load shedding actions for preventing deep discharge of the storage units, and
equalization of the state of charge among distributed storage systems for avoiding
uneven degradation. The proposed equalization method has proved to be effective for
equalizing the state of charge of distributed energy storage systems and for ensuring
uniform charge/discharge ratios regardless of differences in the capacity of the storage
units. Additionally, the strategy is complemented with an optimal scheduling of load
connection, which minimizes the connection and disconnection cycles of the loads
within a time horizon of 24 hours. The proposed architecture is verified experimen-
tally in a lab-scale prototype of a microgrid, which has real communication between
the microgrid and the central controller.
The paper has been accepted for publication in the

IEEE Transactions on Power Electronics, Early Access, DOI.
10.1109/TPEL.2016.2606653, 2016.
59
60
Part VIII
Concluding remarks
61
Conclusion
1 Summary
This thesis proposes different coordination architectures for small-scale is-
landed microgrids based variable renewable energy sources and energy stor-
age systems based on batteries. These kinds of microgrids are highly used
for the electrification of non-interconnected and remote villages or commu-
nication stations. Due to the variability and unpredictable behaviour of the
primary energy resources, the islanded power system relies on energy stor-
age systems for ensuring a reliable operation. The Energy storage systems
can provide the required energy backup and energy buffer for supporting the
variations on the primary energy resources and provides enough autonomy
to the islanded power system. Particularly, battery-based energy storage sys-
tems continue dominating the applications in islanded systems, thanks they
offer the best compromise between lifetime, energy density, power density,
availability, and cost.
The proposed coordination architectures enable an appropriate interac-
tion of the energy storage systems based on batteries with other distributed
generators within an islanded power system, by considering the limitations,
particular requirements, and appropriate charging stages of the storage units.
As a matter of fact, the coordination strategies are based on charging profiles,
which ensures a proper charging process of the energy storage system.
The proposed coordination architectures can be easily deployed and adap-
ted for either ac or dc microgrids, since are based on expert systems which
manage the complexity derived from the interaction of different kinds of dis-
tributed units from a higher level of abstraction. The distributed units in the
microgrid are able to operate independently and without any coordination
by means of a proper definition of inner and primary controllers. Therefore,
the coordination strategies are properly dimensioned and integrated to the
hierarchical control structure defined for microgrids. In this sense, by acting
over the first level (inner control loops, in the hierarchical control structure),
it is possible to modify the role (control mode) of the distributed units in
order to achieve different stages for charging the energy storage system and
63
assigning the responsibility of forming the local power grid to a distributed
unit with enough capability. Meanwhile, adjustments in the primary con-
trollers (second level) allow the definition of cooperative behaviours between
distributed units.
Conventional droop control loops have been considered for the primary
controllers. Therefore, the adjustment derived from cooperative functions
defined within the coordination strategy are properly scaled by considering
the stability issues and constraints. To do that, small-signal models of the
islanded microgrid have been obtained and studied, which consider differ-
ent configurations and interactions between grid-forming units operating in
voltage control mode, and grid-following units operating as constant power
sources or loads.
One of the cooperative operations explored within this project is the equa-
lization of the state of charge of distributed energy storage systems. The SoC
equalization allows the operation of distributed energy storage systems as an
aggregated one by unifying the SoC profiles. This fact allows simultaneous
transitions between charging stages by facilitating the definition of coordina-
tion actions. Additionally, the Soc equalization avoids uneven degradation of
the distributed storage units, by exposing them to similar levels of stress, or
proportional to their capacity, during their operation.
Both. distributed and centralized strategies have been explored. The dis-
tributed approaches take advantages of their inherent modularity and ex-
pandability making them more suitable for applications which require a pro-
gressive expansion or where the use of dedicated communications channels
are not possible. Meanwhile, centralized approaches take advantages of the
global vision provided by the central controller in order to perform more ac-
curate and multi-objective tasks. Therefore, centralized strategies add more
flexibility to the operation of the distributed units, since it makes possible the
definition of additional collaborative behaviours.
The coordination problem of multi-units systems has been widely ex-
plored in different fields. For instance, the coordination of multi-robot in
automation systems. In this sense, similar principles and approaches can be
adopted for the coordination of islanded microgrids as was proposed in this
work by defining a structured architecture based on tactics, roles behaviours
previously used for teams of robot soccer.
2 Contributions
This sections summarizes the main contributions from the point of view of
the Author:
• The definition knowledge-based fuzzy functions for enabling an adap-
tive operation of droop control loops, which allows the equalization of
64
3. Future Work
the state of charge of distributed energy storage systems. The fuzzy

systems rely only on qualitative knowledge, because of that, the can be
easily adapted for ac and dc microgrids. Apart from that, the fuzzy
system can perform multi-objective functions such as the reduction of
voltage deviation in dc microgrids.
• The definition of complete coordination architectures which ensure ap-

propriate stages and profiles for charging battery-based energy storage
systems. The proposed strategies consider the batteries as voltage lim-
iting devices, then the control modes in the energy storage units and in
the whole microgrid are adjusted for ensuring proper stages of charge.
• The development of distributed coordination strategies based on bus

signalling methods without the use of additional communication in-
frastructure. The proposed strategies are completely modular and scal-
able and allow an easy integration of additional distributed generators
and storage units.
• The development of small signal models, which considers different

topological circuit model for an islanded microgrid. The proposed
small-signal models allow the stability analysis of a microgrid com-
posed by parallel connected units in voltage control mode with droop
control loops and their interaction with constant power sources and
constant power loads.
• The design of a novel SoC equalization method based on a centralized

function, which has proved to be effective for equalizing the state of
charge of distributed energy storage systems and for ensuring uniform
charge/discharge ratios regardless of differences in the capacity of the
storage units.
• The adaptation of structured coordination architectures, which have

been previously used in other fields, in the coordination of distributed
energy resources in islanded microgrid.
3 Future Work
The strategies and contributions developed in this thesis show the following
promising aspects to be investigated in the future:
• This work has considered the coordination of microgrids with battery-

based energy storage systems. However, the interaction with energy
storage systems with high power density, such as Supercapacitors, should
be addressed in order to enhance the transient behaviour under sudden
65
variation in generation and consumption and also providing support
for enhancing the dynamic response and smoothing the operation of
the microgrid under the transitions in operation modes.
• The earliest form of artificial intelligence, such as the knowledge-based

system based on logical assertions has been considered within this the-
sis for defining the coordination architecture for islanded microgrids.
In this sense more sophisticated reasoning tools such as optimization
methods based on probability and economics goals, neural networks,
probabilistic methods for uncertain reasoning, etc. which should be
considered in order to have more learning, planning, self problem-
solving and reasoning capability, within the coordination architecture.
• A centralized program has been proposed, which has proved to be

effective for enabling uniform charging/discharging profiles for dis-
tributed ESSs. Therefore, this program should be approached from a
distributed perspective in order to reduce the dependence of a central
controller. In this sense, multi-agent approaches should be considered.
In this way, the advantages of distributed and centralized approaches
can be put together.
• Additional cooperative approaches can be included within the coor-

dination strategy which take into account the reactive power sharing,
the harmonic sharing and compensation, unbalance compensation, and
other power quality issues.
• Coordination architectures for distributed energy storage systems in

grid-connected microgrids should be also considered by taking into ac-
count the uniform operation of distributed units.
66
Literature List
References
[1] N. L. Díaz, D. Wu, T. Dragičević, J. C. Vasquez, and J. M. Guerrero,
“Stored energy balance for distributed pv-based active generators in an
ac microgrid,” in 2015 IEEE Power Energy Society General Meeting, July
2015, pp. 1–5.
[2] N. L. Díaz, A. C. Luna, J. C. Vasquez, and J. M. Guerrero, “Centralized

control architecture for coordination of distributed renewable genera-
tion and energy storage in islanded ac microgrids,” IEEE Transactions on
Power Electronics, Early Access, DOI. 10.1109/TPEL.2016.2606653.
[3] N. L. Díaz, J. C. Vasquez, and J. M. Guerrero, “A communication-less

distributed control architecture for islanded microgrids with renewable
generation and storage,” IEEE Transactions on Power Electronics, Under
review.
[4] U. E. I. Administration, “International energy outlook 2016 with

projections to 2040,” Office of Energy Analysis U.S. Department of
Energy, Tech. Rep., May 2016. [Online]. Available: www.eia.gov/
forecasts/ieo/pdf/0484(2016).pdf
[5] A. Keyhani, Design of Smart Power Grid Renewable Energy Systems. Wiley,
2011.
[6] Siemens, “Microgrids white paper,” Siemens AG, Tech. Rep.,

2011. [Online]. Available: http://www.climateactionprogramme.org/
images/uploads/documents/Microgrids_as_a_solution_to_integrate_
renewable_generation_Siemens_White_Paper.pdf
[7] T. Ackermann, E. M. Carlini, B. Ernst, F. Groome, A. Orths, J. O’Sullivan,

M. de la Torre Rodriguez, and V. Silva, “Integrating variable renewables
in europe : Current status and recent extreme events,” IEEE Power and
Energy Magazine, vol. 13, no. 6, pp. 67–77, Nov 2015.
67
References
[8] M. Milligan, B. Frew, B. Kirby, M. Schuerger, K. Clark, D. Lew, P. Den-

holm, B. Zavadil, M. O’Malley, and B. Tsuchida, “Alternatives no more:
Wind and solar power are mainstays of a clean, reliable, affordable grid,”
IEEE Power and Energy Magazine, vol. 13, no. 6, pp. 78–87, Nov 2015.
[9] M. Castillo-Cagigal, E. Caamano-Martinn, E. Matallanas, D. Masa-
Bote, A. Gutierrez, F. Monasterio-Huelin, and J. Jimenez-Leube, “Pv
self-consumption optimization with storage and active dsm for the
residential sector,” Solar Energy, vol. 85, no. 9, pp. 2338 – 2348, 2011.
[Online]. Available: http://www.sciencedirect.com/science/article/
pii/S0038092X11002404
[10] S. E. Company, “The role of energy storage in smart microgrids white
paper,” S&C Electric Company, Tech. Rep., 2013.
[11] M. Beaudin, H. Zareipour, A. Schellenberg, and W. Rosehart,
“Chapter 1 - energy storage for mitigating the variability of renewable
electricity sources,” in Energy Storage for Smart Grids, P. D. Lu, Ed.
Boston: Academic Press, 2015, pp. 1 – 33. [Online]. Available: http:
//www.sciencedirect.com/science/article/pii/B9780124104914000014
[12] R. H. Lasseter, “Smart distribution: Coupled microgrids,” Proceedings of
the IEEE, vol. 99, no. 6, pp. 1074–1082, June 2011.
[13] J. M. Guerrero, J. C. Vasquez, J. Matas, L. G. de Vicuna, and M. Castilla,
“Hierarchical control of droop-controlled ac and dc microgrids - a gen-
eral approach toward standardization,” IEEE Transactions on Industrial
Electronics, vol. 58, no. 1, pp. 158–172, Jan 2011.
[14] X. Liu, P. Wang, and P. C. Loh, “A hybrid ac/dc microgrid and its co-
ordination control,” IEEE Transactions on Smart Grid, vol. 2, no. 2, pp.
278–286, June 2011.
[15] F. Marra and G. Yang, “Chapter 10 - decentralized energy storage
in residential feeders with photovoltaics,” in Energy Storage for Smart
Grids, P. D. Lu, Ed. Boston: Academic Press, 2015, pp. 277
– 294. [Online]. Available: http://www.sciencedirect.com/science/
article/pii/B9780124104914000105
[16] P. Asmus, “Microgrids, virtual power plants and our distributed
energy future,” The Electricity Journal, vol. 23, no. 10, pp. 72 –
82, 2010. [Online]. Available: http://www.sciencedirect.com/science/
article/pii/S1040619010002873
[17] B. K. Paul Denholm, Erik Ela and M. Milligan, “The role of energy stor-
age with renewable electricity generation,” National Renewable Energy
Laboratory, Tech. Rep., January 2010.
68
References
[18] J. P. Chaves-Avila, K. Wurzburg, T. Gomez, and P. Linares, “The green

impact: How renewable sources are changing eu electricity prices,” IEEE
Power and Energy Magazine, vol. 13, no. 4, pp. 29–40, July 2015.
[19] A. Olson, A. Mahone, E. Hart, J. Hargreaves, R. Jones, N. Schlag,

G. Kwok, N. Ryan, R. Orans, and R. Frowd, “Halfway there: Can cal-
ifornia achieve a 50% renewable grid?” IEEE Power and Energy Magazine,
vol. 13, no. 4, pp. 41–52, July 2015.
[20] W. Marańda and M. Piotrowicz, “Short-time energy buffering for pho-

tovoltaic system,” in Mixed Design of Integrated Circuits and Systems
(MIXDES), 2010 Proceedings of the 17th International Conference, June 2010,
pp. 525–528.
[21] C. L. Moreira and J. A. P. Lopes, MicroGrids Operation and Control

under Emergency Conditions. Berlin, Heidelberg: Springer Berlin
Heidelberg, 2012, ch. Chapter 12, pp. 351–399. [Online]. Available:
http://dx.doi.org/10.1007/978-3-642-21578-0_12
[22] Q. Jiang, M. Xue, and G. Geng, “Energy management of microgrid in

grid-connected and stand-alone modes,” IEEE Transactions on Power Sys-
tems, vol. 28, no. 3, pp. 3380–3389, Aug 2013.
[23] P. A. Østergaard, “Reviewing optimisation criteria for energy systems

analyses of renewable energy integration,” Energy, vol. 34, no. 9, pp.
1236 – 1245, 2009. [Online]. Available: http://www.sciencedirect.com/
science/article/pii/S0360544209001777
[24] A. C. Luna, N. L. D. Aldana, M. Graells, J. C. Vasquez, and J. M.

Guerrero, “Mixed-integer-linear-programming based energy manage-
ment system for hybrid pv-wind-battery microgrids: Modeling, design
and experimental verification,” IEEE Transactions on Power Electronics,
vol. PP, no. 99, pp. 1–1, 2016.
[25] F. Katiraei, R. Iravani, N. Hatziargyriou, and A. Dimeas, “Microgrids

management,” IEEE Power and Energy Magazine, vol. 6, no. 3, pp. 54–65,
May 2008.
[26] J. de Matos, F. e Silva, and L. Ribeiro, “Power control in ac isolated

microgrids with renewable energy sources and energy storage systems,”
IEEE Transactions on Industrial Electronics,, vol. 62, no. 6, pp. 3490–3498,
2015.
[27] Y. Zhang, H. J. Jia, and L. Guo, “Energy management strategy of is-

landed microgrid based on power flow control,” in 2012 IEEE PES Inno-
vative Smart Grid Technologies (ISGT), Jan 2012, pp. 1–8.
69
References
[28] M. Alam, S. D. Ramchurn, and A. Rogers, “Cooperative energy

exchange for the efficient use of energy and resources in remote
communities,” in Proceedings of the 2013 International Conference
on Autonomous Agents and Multi-agent Systems, ser. AAMAS ’13.
Richland, SC: International Foundation for Autonomous Agents
and Multiagent Systems, 2013, pp. 731–738. [Online]. Available:
http://dl.acm.org/citation.cfm?id=2484920.2485035
[29] B. Heap, “Smart villages: New thinking for off-grid communities world-
wide,” Banson/Smart Villages Initiative., Tech. Rep., 2015.
[30] N. L. Díaz, E. A. Coelho, J. C. Vasquez, and J. M. Guerrero, “Stability

analysis for isolated ac microgrids based on pv-active generators,” in
2015 IEEE Energy Conversion Congress and Exposition (ECCE), Sept 2015,
pp. 4214–4221.
[31] J. A. P. Lopes, C. L. Moreira, and A. G. Madureira, “Defining control

strategies for microgrids islanded operation,” IEEE Transactions on Power
Systems, vol. 21, no. 2, pp. 916–924, May 2006.
[32] S. McCluer and J.-F. Christin, “Comparing data center batteries, fly-
wheels, and ultracapacitors (white paper),” Schneider Electric’s, Tech.
Rep., 2011.
[33] N. Gueissaz, “Distributed vs. centralized energy storage for

power system applications,” Master’s thesis, Electrical Engi-
neering, University College Dublin, 2014. [Online]. Available:
http://energyexemplar.com/wp-content/uploads/publications/2014_
Hanze/Distributed%20vs.%20Centralized%20Energy%20Storage%
20for%20Power%20Sytsem%20Applications.pdf
[34] B. Sun, T.Dragičević, F. D. Freijedo, J. C. Vasquez, and J. M. Guerrero,

“A control algorithm for electric vehicle fast charging stations equipped
with flywheel energy storage systems,” IEEE Transactions on Power Elec-
tronics, vol. 31, no. 9, pp. 6674–6685, Sept 2016.
[35] D. Linden and T. Reddy, Handbook of Batteries, third edition ed.,

ser. McGraw-Hill handbooks. McGraw-Hill Education, 2001. [Online].
Available: https://books.google.dk/books?id=XquySsZp5jsC
[36] I. Hadjipaschalis, A. Poullikkas, and V. Efthimiou, “Overview

of current and future energy storage technologies for electric
power applications,” Renewable and Sustainable Energy Reviews,
vol. 13, no. 6–7, pp. 1513 – 1522, 2009. [Online]. Available:
http://www.sciencedirect.com/science/article/pii/S1364032108001664
70
References
[37] H. Ibrahim, A. Ilinca, and J. Perron, “Energy storage systems-

characteristics and comparisons,” Renewable and Sustainable Energy
Reviews, vol. 12, no. 5, pp. 1221 – 1250, 2008. [Online]. Available:
[38] “IEEE Guide for Optimizing the Performance and Life of Lead-Acid
Batteries in Remote Hybrid Power Systems,” IEEE Std 1561-2007, pp.
C1–25, May 2008.
[39] R. Perez, “Lead-acid battery state of charge vs. voltage .” Home Power,
no. 36, pp. 66–69, August / September 1993.
[40] F. Marra, G. Yang, C. Træholt, J. Østergaard, and E. Larsen, “A decen-

tralized storage strategy for residential feeders with photovoltaics,” IEEE
Transactions on Smart Grid, vol. 5, no. 2, pp. 974–981, March 2014.
[41] I. Buchmann and C. E. Inc, Batteries in a Portable World: A Handbook on

Rechargeable Batteries for Non-engineers, 4th ed. Cadex Electronics, 2016.
[42] U S E RŚ G U I D E Trijan Battery., Trojan Battery Company, 12380 Clark

Street, Santa Fe Springs, CA 90670 USA, 2013.
[43] R. B. Europe, BATTERY USER MANUAL, Rolls Battery Eginnering, 5

Little Park Farm Segensworth West Fareham Hampshire England UK
PO15 5SJ.
[44] C. A. Hill, M. C. Such, D. Chen, J. Gonzalez, and W. M. Grady, “Battery

energy storage for enabling integration of distributed solar power gen-
eration,” IEEE Transactions on Smart Grid, vol. 3, no. 2, pp. 850–857, June
2012.
[45] J. Cui, K. Li, Y. Sun, Z. Zou, and Y. Ma, “Distributed energy storage
system in wind power generation,” in Electric Utility Deregulation and
Restructuring and Power Technologies (DRPT), 2011 4th International Con-
ference on, July 2011, pp. 1535–1540.
[46] W. Li and G. Joos, “Performance comparison of aggregated and dis-

tributed energy storage systems in a wind farm for wind power fluc-
tuation suppression,” in Power Engineering Society General Meeting, 2007.
IEEE, June 2007, pp. 1–6.
[47] E. Liegmann and R. Majumder, “An efficient method of multiple storage

control in microgrids,” IEEE Transactions on Power Systems, vol. 30, no. 6,
pp. 3437–3444, Nov 2015.
[48] H. Kanchev, D. Lu, F. Colas, V. Lazarov, and B. Francois, “Energy man-

agement and operational planning of a microgrid with a pv-based active
71
References
generator for smart grid applications,” IEEE Transactions on Industrial

Electronics, vol. 58, no. 10, pp. 4583–4592, Oct 2011.
[49] H. Kanchev, F. Colas, V. Lazarov, and B. Francois, “Emission reduction
and economical optimization of an urban microgrid operation including
dispatched pv-based active generators,” IEEE Transactions on Sustainable
Energy, vol. 5, no. 4, pp. 1397–1405, Oct 2014.
[50] A. N. Azmi, M. L. Kolhe, and A. G. Imenes, “Review on photovoltaic
based active generator,” in 2015 9th International Symposium on Advanced
Topics in Electrical Engineering (ATEE), May 2015, pp. 812–815.
[51] D. Wu, “Decentralized coordinated control strategy of islanded micro-
grids,” Ph.D. dissertation, 11 2015.
[52] R. Newnham and W. Baldsing, “Advanced management strategies
for remote-area power-supply systems,” Journal of Power Sources, vol.
133, no. 1, pp. 141 – 146, 2004, proceedings of the Tenth Asian
Battery Conference. [Online]. Available: http://www.sciencedirect.
com/science/article/pii/S0378775303011224
[53] V. Salas, E. Olías, A. Barrado, and A. Lázaro, “Review of the
maximum power point tracking algorithms for stand-alone photovoltaic
systems,” Solar Energy Materials and Solar Cells, vol. 90, no. 11, pp.
1555 – 1578, 2006. [Online]. Available: http://www.sciencedirect.com/
science/article/pii/S0927024805003582
[54] C. Patsios, A. Chaniotis, M. Rotas, and A. Kladas, “A comparison
of maximum-power-point tracking control techniques for low-power
variable-speed wind generators,” in 8th International Symposium on Ad-
vanced Electromechanical Motion Systems Electric Drives Joint Symposium,
2009. ELECTROMOTION 2009, July 2009, pp. 1–6.
[55] C. L. T. Rodríguez, N. L. D. Aldana, and J. A. H. Mora, “Control of an in-
verter for photovoltaic arrays, using an estimator of short circuit current
for maximum power point tracking,” IEEE Latin America Transactions,
vol. 13, no. 10, pp. 3294–3303, Oct 2015.
[56] S. X. Chen and H. B. Gooi, “Sizing of energy storage system for micro-
grids,” in Probabilistic Methods Applied to Power Systems (PMAPS), 2010
IEEE 11th International Conference on, June 2010, pp. 6–11.
[57] E. Hittinger, T. Wiley, J. Kluza, and J. Whitacre, “Evaluating the value
of batteries in microgrid electricity systems using an improved energy
systems model,” Energy Conversion and Management, vol. 89, pp. 458 –
472, 2015. [Online]. Available: http://www.sciencedirect.com/science/
article/pii/S0196890414008929
72
References
[58] H. Hassanzadehfard, S. Moghaddas-Tafreshi, and S. Hakimi, “Optimal

sizing of an islanded micro-grid for an area in north-west iran using
particle swarm optimization based on reliability concept,” in World Re-
newable Energy Congress-Sweden; 8-13 May; 2011; Linköping; Sweden, no.
057. Linköping University Electronic Press, 2011, pp. 2969–2976.
[59] C. Li, H. Shi, Y. Cao, J. Wang, Y. Kuang, Y. Tan, and

J. Wei, “Comprehensive review of renewable energy curtailment and
avoidance: A specific example in china,” Renewable and Sustainable
Energy Reviews, vol. 41, pp. 1067 – 1079, 2015. [Online]. Available:
[60] J. Y. Kim, J. H. Jeon, S. K. Kim, C. Cho, J. H. Park, H. M. Kim, and

K. Y. Nam, “Cooperative control strategy of energy storage system and
microsources for stabilizing the microgrid during islanded operation,”
IEEE Transactions on Power Electronics, vol. 25, no. 12, pp. 3037–3048, Dec
2010.
[61] H. Mahmood, D. Michaelson, and J. Jiang, “Strategies for independent

deployment and autonomous control of pv and battery units in islanded
microgrids,” IEEE Journal of Emerging and Selected Topics in Power Electron-
ics, vol. 3, no. 3, pp. 742–755, Sept 2015.
[62] T. Dragičević, J. M. Guerrero, J. C. Vasquez, and D. Škrlec, “Supervisory

control of an adaptive-droop regulated dc microgrid with battery man-
agement capability,” IEEE Transactions on Power Electronics, vol. 29, no. 2,
pp. 695–706, Feb 2014.
[63] T. Dragičević, J. M. Guerrero, and J. C. Vasquez, “A distributed control

strategy for coordination of an autonomous lvdc microgrid based on
power-line signaling,” IEEE Transactions on Industrial Electronics, vol. 61,
no. 7, pp. 3313–3326, July 2014.
[64] M. A. C. S. Michalis Foukarakis, Asterios Leonidis, “Combining finite

state machine and decision-making tools for adaptable robot behavior,”
in Universal Access in Human-Computer Interaction. Aging and Assistive En-
vironments. Springer International Publishing, 2014, vol. 8515.
[65] I. Millington and J. Funge, “5 - decision making,” in Artificial

Intelligence for Games (Second Edition), second edition ed., I. Millington
and J. Funge, Eds. Boston: Morgan Kaufmann, 2009, pp. 293
– 491. [Online]. Available: http://www.sciencedirect.com/science/
article/pii/B9780123747310000050
[66] D. M. Harris and S. L. Harris, “1 - from zero to one,” in

Digital Design and Computer Architecture (Second Edition), second
73
References
edition ed., D. M. Harris, , and S. L. Harris, Eds. Boston:

Morgan Kaufmann, 2013, pp. 2 – 52. [Online]. Available: http:
//www.sciencedirect.com/science/article/pii/B978012394424500001X
[67] J. Rocabert, A. Luna, F. Blaabjerg, and P. Rodríguez, “Control of power
converters in ac microgrids,” IEEE Transactions on Power Electronics,
vol. 27, no. 11, pp. 4734–4749, Nov 2012.
[68] J. C. Vasquez, J. M. Guerrero, A. Luna, P. Rodriguez, and R. Teodorescu,
“Adaptive droop control applied to voltage-source inverters operating
in grid-connected and islanded modes,” IEEE Transactions on Industrial
Electronics, vol. 56, no. 10, pp. 4088–4096, Oct 2009.
[69] F. Guo, C. Wen, J. Mao, J. Chen, and Y. D. Song, “Distributed cooperative
secondary control for voltage unbalance compensation in an islanded
microgrid,” IEEE Transactions on Industrial Informatics, vol. 11, no. 5, pp.
1078–1088, Oct 2015.
[70] M. Savaghebi, A. Jalilian, J. C. Vasquez, and J. M. Guerrero, “Secon-
dary control scheme for voltage unbalance compensation in an islanded
droop-controlled microgrid,” IEEE Transactions on Smart Grid, vol. 3,
no. 2, pp. 797–807, June 2012.
[71] Q. Shafiee, J. M. Guerrero, and J. C. Vasquez, “Distributed secondary
control for islanded microgrids - a novel approach,” IEEE Transactions
on Power Electronics, vol. 29, no. 2, pp. 1018–1031, Feb 2014.
[72] L. Meng, “Hierarchical control for optimal and distributed operation of
microgrid systems,” Ph.D. dissertation, 10 2015.
[73] Y. U. Cao, A. S. Fukunaga, and A. Kahng, “Cooperative mobile robotics:
Antecedents and directions,” Autonomous Robots, vol. 4, no. 1, pp. 7–27,
1997. [Online]. Available: http://dx.doi.org/10.1023/A:1008855018923
[74] A. Kahrobaeian and Y. A. R. I. Mohamed, “Networked-based hybrid
distributed power sharing and control for islanded microgrid systems,”
IEEE Transactions on Power Electronics, vol. 30, no. 2, pp. 603–617, Feb
2015.
[75] R. Majumder, A. Ghosh, G. Ledwich, and F. Zare, “Power management
and power flow control with back-to-back converters in a utility con-
nected microgrid,” IEEE Transactions on Power Systems, vol. 25, no. 2, pp.
821–834, May 2010.
[76] N. L. Díaz, J. G. Guarnizo, M. Mellado, J. C. Vasquez, and J. M. Guerrero,
“A robot-soccer-coordination inspired control architecture applied to is-
landed microgrids,” IEEE Transactions on Power Electronics, Early Access,
DOI. 10.1109/TPEL.2016.2572262.
74
References
[77] X. Lu, K. Sun, J. Guerrero, J. Vasquez, and L. Huang, “State-of-charge

balance using adaptive droop control for distributed energy storage sys-
tems in dc microgrid applications,” IEEE Transactions on Industrial Elec-
tronics, vol. 61, no. 6, pp. 2804–2815, June 2014.
[78] W. Shi, X. Xie, C. C. Chu, and R. Gadh, “Distributed optimal energy

management in microgrids,” IEEE Transactions on Smart Grid, vol. 6,
no. 3, pp. 1137–1146, May 2015.
[79] B. Wang, M. Sechilariu, and F. Locment, “Intelligent dc microgrid with

smart grid communications: Control strategy consideration and de-
sign,” IEEE Transactions on Smart Grid, vol. 3, no. 4, pp. 2148–2156, Dec
2012.
[80] M. Mao, H. Huang, and L. Chang, “Real-time energy coordinated and

balance control strategies for microgrid with photovoltaic generators,” in
2013 4th IEEE International Symposium on Power Electronics for Distributed
Generation Systems (PEDG), July 2013, pp. 1–7.
[81] H. Mahmood, D. Michaelson, and J. Jiang, “Strategies for independent

deployment and autonomous control of pv and battery units in islanded
microgrids,” IEEE Journal of Emerging and Selected Topics in Power Electron-
ics, vol. 3, no. 3, pp. 742–755, Sept 2015.
[82] D. Wu, F. Tang, T. Dragičević, J. C. Vasquez, and J. M. Guerrero, “A

control architecture to coordinate renewable energy sources and energy
storage systems in islanded microgrids,” IEEE Transactions on Smart Grid,
vol. 6, no. 3, pp. 1156–1166, May 2015.
[83] E. Serban and H. Serban, “A control strategy for a distributed power

generation microgrid application with voltage- and current-controlled
source converter,” IEEE Transactions on Power Electronics, vol. 25, no. 12,
pp. 2981–2992, Dec 2010.
[84] D. Wu, F. Tang, T. Dragičević, J. C. Vasquez, and J. M. Guerrero, “Au-

tonomous active power control for islanded ac microgrids with pho-
tovoltaic generation and energy storage system,” IEEE Transactions on
Energy Conversion, vol. 29, no. 4, pp. 882–892, Dec 2014.
[85] D. Wu, F. Tang, T. Dragičević, J. M. Guerrero, and J. C. Vasquez, “Coordi-

nated control based on bus-signaling and virtual inertia for islanded dc
microgrids,” IEEE Transactions on Smart Grid, vol. 6, no. 6, pp. 2627–2638,
Nov 2015.
[86] N. L. Díaz, T. Dragičević, J. C. Vasquez, and J. M. Guerrero, “Intelli-

gent distributed generation and storage units for dc microgrids - a new
75
References
concept on cooperative control without communications beyond droop

control,” IEEE Transactions on Smart Grid, no. 5, pp. 2476–2485, Sept 2014.
[87] W. Huang and J. A. A. Qahouq, “Energy sharing control scheme for

state-of-charge balancing of distributed battery energy storage system,”
IEEE Transactions on Industrial Electronics, vol. 62, no. 5, pp. 2764–2776,
May 2015.
[88] X. Lu, K. Sun, J. Guerrero, J. Vasquez, and L. Huang, “Double-quadrant

state-of-charge-based droop control method for distributed energy stor-
age systems in autonomous dc microgrids,” IEEE Transactions on Smart
Grid, vol. 6, no. 1, pp. 147–157, Jan 2015.
[89] Q. Shafiee, T. Dragičević, J. C. Vasquez, and J. M. Guerrero, “Hierarchical

control for multiple dc-microgrids clusters,” IEEE Transactions on Energy
Conversion, vol. 29, no. 4, pp. 922–933, Dec 2014.
[90] Y. Guan, J. Vasquez, and J. Guerrero, “Coordinated secondary control

for balanced discharge rate of energy storage system in islanded ac mi-
crogrids,” IEEE Transactions on Industry Applications, vol. PP, no. 99, pp.
1–1, 2016.
[91] T. R. Oliveira, W. W. A. G. Silva, and P. F. Donoso-Garcia, “Distributed

secondary level control for energy storage management in dc micro-
grids,” IEEE Transactions on Smart Grid, vol. PP, no. 99, pp. 1–11, 2016.
[92] C. Li, T. Dragičević, M. Garcia Plaza, F. Andrade, J. Vasquez, and J. Guer-

rero, “Multiagent based distributed control for state-of-charge balance of
distributed energy storage in dc microgrids,” in 40th Annual Conference of
the IEEE Industrial Electronics Society, (IECON), Oct 2014, pp. 2180–2184.
[93] C. Li, T. Dragičević, N. Diaz, J. Vasquez, and J. Guerrero, “Voltage

scheduling droop control for state-of-charge balance of distributed
energy storage in dc microgrids,” in Energy Conference (ENERGYCON),
2014 IEEE International, May 2014, pp. 1310–1314.
[94] O. Palizban and K. Kauhaniemi, “Distributed cooperative control of

battery energy storage system in {AC} microgrid applications,” Journal
of Energy Storage, vol. 3, pp. 43 – 51, 2015. [Online]. Available:
http://www.sciencedirect.com/science/article/pii/S2352152X15300074
[95] C. Li, T. Dragičević, J. C. Vasquez, J. M. Guerrero, and E. A. A. Coelho,

“Multi-agent-based distributed state of charge balancing control for dis-
tributed energy storage units in ac microgrids,” in 2015 IEEE Applied
Power Electronics Conference and Exposition (APEC), March 2015, pp. 2967–
2973.
76
References
[96] N. Díaz, D. Wu, T. Dragičević, J. Vasquez, and J. Guerrero, “Fuzzy droop

control loops adjustment for stored energy balance in distributed energy
storage system,” in 2015 9th International Conference on Power Electronics
and ECCE Asia (ICPE-ECCE Asia), June 2015, pp. 728–735.
[97] M. D. Mulligan, B. Broach, and T. H. Lee, “A constant-frequency method

for improving light-load efficiency in synchronous buck converters,”
IEEE Power Electronics Letters, vol. 3, no. 1, pp. 24–29, March 2005.
[98] L. Meng, T. Dragičević, J. M. Guerrero, and J. C. Vasquez, “Dynamic

consensus algorithm based distributed global efficiency optimization of
a droop controlled dc microgrid,” in Energy Conference (ENERGYCON),
2014 IEEE International, May 2014, pp. 1276–1283.
77
ISSN (online): 2246-1248
ISBN (online): 978-87-7112-846-8

PHD Nelson Leonardo Diaz Aldana E PDF

Uploaded by

Copyright:

Available Formats

PHD Nelson Leonardo Diaz Aldana E PDF

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

PHD Nelson Leonardo Diaz Aldana E PDF

Uploaded by

Copyright:

Available Formats

Aalborg Universitet

Coordinated Control and Management of Distributed battery-based energy Storage

Diaz Aldana, Nelson Leonardo

Link to publication from Aalborg University

Citation for published version (APA):

D ISS ERTAT ION S U B MITTE D 2016

Department of Energy Technology

PhD supervisor: Prof. Josep M. Guerrero

Assistant PhD supervisor: Assoc. Prof. Juan C. Vasquez

PhD committee: Associate Professor Erik Schaltz (chairman)

PhD Series: Faculty of Engineering and Science, Aalborg University

ISSN (online): 2246-1248

© Copyright: Nelson Leonardo Díaz Aldana

Printed in Denmark by Rosendahls, 2016

The integration of variable renewable energy sources such as solar or wind

Integrationen af variable vedvarende energikilder såsom sol eller vindmøller

tal driftstilstande, at sikre passende opladning/afladning profiler til energi-

Thesis Title: Coordinated Control and Management of Distributed

Supervisor: Prof. Josep M. Guerrero.

Co-supervisor: Prof. Juan C. Vasquez.

1st Authored Journal Papers:

• Díaz, N. L., Dragičević, T., Vasquez, J. C., & Guerrero, J. M. (2014).

• Díaz, N. L., Vasquez, J. C., & Guerrero, J. M. A Communication-less

Co-authored Journal Publications:

Selected Publications in Proceedings with Peer Review:

• Díaz, N. L., Dragičević, T., Vasquez, J. C., & Guerrero, J. M. (2014).

• Díaz, N. L., Coelho, E. A. A., Vasquez, J. C., & Guerrero, J. M. (2015).

AC Microgrids. In Proceedings of the 41th Annual Conference of IEEE

This document contains a summary of the Ph.D. project, entitled “Coordi-

possible all my process during my Ph.D. study.

Nelson Leonardo Díaz Aldana.

Curriculum Vitae iii

Thesis Details and Publications ix

List of Figures xvii

VIII Concluding remarks 61

16 Comparison between charging current behaviours in distributed

1 Energy Storage Systems in Microgrids

1.1 Grid-Connected and Islanded Operation

consumption [15]. Because of that, the participation of microgrids on the

YLine1 YLine 2 YLOAD YLine3 YLine 4

YLine 2 YLOAD YLine3 YLine 4

YLine1 Microgrid YLine 4

YLine 2 YLOAD YLine3

RESs or regulated process of charge in ESSs). Within the current trend in

However, the management of DERs in islanded microgrids involve addi-

1.2 Energy Storage Systems in Islanded Microgrids

• Supercapacitors: They are electrical energy storage systems character-

• Flywheels: They are mechanical ESSs based on a rotating mass, which

• Super Conducting Magnetic Energy Storage (SMES): It is a device that

able batteries technologies can be utilized in conjunction with variable

– Lead-Acid Batteries: They are characterized by their low cost, high

1.3 Charge/Discharge Characteristic of Batteries

Overcharge Safe Operation Window Over-discharge

Table 1: Summary Table of Typical Voltage Values per Cell

Charge Cut-off Cell Voltage Discharge Cut-off

Fig. 4: Charging/Discharging Characteristic of Lead-acid batteries.

ate charging is critically important to the life and performance of lead-acid

• In the first stage of charge it is ensured about an 80 % of the total charge,

• The topping charge (constant-voltage charging) is the most effective

• Since the Lead-acid batteries are voltage-limiting devices, the manu-