Energies

energies
Article
Hybrid System Assessment in On-Grid and Off-Grid
Conditions: A Technical and Economical Approach
Jonathan Muñoz Tabora 1, *, Ulisses Carvalho Paixão Júnior 1 , Carlos Eduardo Moreira Rodrigues 1 ,
Ubiratan Holanda Bezerra 1 , Maria Emília de Lima Tostes 1 , Bruno S. de Albuquerque 1 , Edson Ortiz de Matos 1
and Andréia Antloga do Nascimento 2
1 Post-Graduate Program in Electrical Engineering, Federal University of Para, Belém 66075-110, Brazil;
upaixaojr.eng@gmail.com (U.C.P.J.); carloseduardo.pa@hotmail.com (C.E.M.R.);
ubiratan.fapespa@gmail.com (U.H.B.); tostes@ufpa.br (M.E.d.L.T.); bsalbuquerque@live.com (B.S.d.A.);
ortiz@ufpa.br (E.O.d.M.)
2 Norte Energia S.A., Brasilia 70390-025, Brazil; andreianascimento@norteenergiasa.com.br
* Correspondence: jonathan_mt24@hotmail.com; Tel.: +55-91-98498-2482
Abstract: Photovoltaic systems can strengthen the energy matrix aiming at energy sustainability,
however, their intermittence, the availability time frame and seasonality effects are complicating
aspects for the operation of these systems. To mitigate such problems, a possible measure is the
operation of hybrid systems (HS) associated with storage systems to meet the energy demand. This
study presents real operational scenarios for the management of a hybrid mini-grid installed in the
Amazon region in Brazil, capable of managing energy supply and demand, changing the generation

curve, reducing fossil fuel consumption and energy costs, in addition to providing flexibility and
operational safety to the electrical operating under on-grid or off-grid conditions. Through technical
Citation: Tabora, J.M.; Paixão Júnior,
and economic analyses, the objective of this study is to show that hybrid systems can be implemented
U.C.; Rodrigues, C.E.M.; Bezerra,
U.H.; Tostes, M.E.d.L.; Albuquerque,
in small and medium consumers, in addition to generating a new market option for utilities. This
B.S.d.; Matos, E.O.d.; Nascimento, research evaluates the operation of HS concurrently with the electrical grid and the results show
A.A.d. Hybrid System Assessment in that the system meets the technical requirements, in addition to bringing financial and ecological
On-Grid and Off-Grid Conditions: A benefits. This information can support the decisions of consumers, utilities and energy operators for
Technical and Economical Approach. the development of the national market for on-grid and off-grid HS.
Energies 2021, 14, 5284. https://
doi.org/10.3390/en14175284 Keywords: hybrid system; storage system; photovoltaic generation; diesel generation; hybrid mini
grid; economic analysis
Academic Editor: Adel Merabet
Received: 28 June 2021

Accepted: 31 July 2021
1. Introduction
Published: 26 August 2021
Over the past few years, the Brazilian energy matrix has been expanding its capacity to
Publisher’s Note: MDPI stays neutral
generate electricity from renewable sources. Figure 1 shows the percentage distribution of
with regard to jurisdictional claims in
generation sources in 2020, based on 2019, which adds up to a total generation of 651.3 TWh,
published maps and institutional affil- according to the National Energy Balance Synthesis Report [1].
iations. Given the vast territorial extension of Brazil, it is possible to identify regions of low
and medium latitudes, with different patterns of precipitation and temperature. The
Amazon region has very high average rainfall, especially between the months of November
and March, when compared to regions in the northeastern semiarid region. In Brazil, the
Copyright: © 2021 by the authors.
average annual temperature ranges from 18 ◦ C to 29 ◦ C in a large part of the territory,
Licensee MDPI, Basel, Switzerland.
with higher average values observed in the North and Northeast regions, so that, in these
This article is an open access article
regions, the maximum temperature averages reach values as high 33 ◦ C in December [2].
distributed under the terms and According to [3,4], it is possible to visualize the monthly variation of irradiation in different
conditions of the Creative Commons regions in Brazil, showing its large solar photovoltaic generation capacity at any time of
Attribution (CC BY) license (https:// the year, despite the variation in its intensity.
creativecommons.org/licenses/by/ The ANEEL (National Electric Energy Agency) normative resolution 687/2015, has
4.0/). contributed to the growth of distributed generation sources in Brazil. In addition, a set of
Energies 2021, 14, 5284. https://doi.org/10.3390/en14175284 https://www.mdpi.com/journal/energies

Energies 2021, 14, x FOR PEER REVIEW 2 of 21
Energies 2021, 14, 5284 2 of 21
government incentives, high electricity costs, reduction in the cost of photovoltaic panels,
Energies 2021, 14, x FOR PEER REVIEW
as well as the availability and potential of the solar resource throughout the country,2 of
have
21
allowed photovoltaic generation source to grow exponentially in the country, as can be
seen in Figure 2.
Figure 1. Electricity generation sources in Brazil, base year 2019. Reproduced from [1], Energy
research company: 2020.
The ANEEL (National Electric Energy Agency) normative resolution 687/2015, has
contributed to the growth of distributed generation sources in Brazil. In addition, a set of
as well as the availability and potential of the solar resource throughout the country, have
Figure
Figure Electricity
1.1.Electricity
allowed photovoltaicgeneration
generation sources
generation
sources in Brazil,
source tobase
in Brazil, base
grow year
2019.2019.
exponentially
year Reproduced
in the
Reproduced from
country,
from [1], [1],
asEnergy
Energycan be
research
seen incompany:
research Figure 2.2020.
company: 2020.
The ANEEL (National Electric Energy Agency) normative resolution 687/2015, has
contributed to the growth of distributed generation sources in Brazil. In addition, a set of
as well as the availability and potential of the solar resource throughout the country, have
allowed photovoltaic generation source to grow exponentially in the country, as can be
seen in Figure 2.
Figure2.2.Solar
Figure Solargeneration
generationin
inBrazil.
Brazil.Reproduced
Reproducedfrom
from[5],
[5],National
NationalElectric
ElectricSystem
SystemOperator:
Operator:2021.
2021.
Despitethe
Despite thebenefits
benefitsthat thatdistributed
distributed solar
solar generation
generation brings
brings in economic
in economic andand ecolog-
ecological
ical terms,
terms, due due to non-dispatched
to its its non-dispatched nature
nature andand thethe limited
limited availabilitywindow
availability windowofofsolar solar
irradiation, its generation
irradiation, generation isislimited limitedtotomeeting
meetingdemand demand at at
thethesame time
same it isitbeing
time gen-
is being
erated. Until
generated. now,
Until the the
now, existing
existingrules in Brazil
rules in Brazildo not
do notallow the the
allow use useof energy
of energy storage sys-
storage
systems connected to the power grid, making some generation
tems connected to the power grid, making some generation systems less efficient. How- systems less efficient.
Figure
ever,2.this
However, Solar generation
this scenario
scenario in is
Brazil.
is expected Reproduced
expected to change
to change infrom
the[5],
in National
the
coming coming Electric
years, years,
since,System Operator:
since,
in 2016, inthe
2016, 2021.
ANEELthe
ANEEL
created created
a strategica strategic publicfor
public notice notice for research
research and development
and development (R&D) (R&D)
on battery on battery
energy
Despite
energy
storage the technologies
storage benefits
technologies that distributed
through through callsolar
public public N°call N◦ 21/2016.
generation
21/2016. Thebrings The
project inaims
economic
project
to aims and
develop to ecolog-
develop
and test
ical
and terms,
energy due tosystems,
teststorage
energy its non-dispatched
storage systems, nature
operatingoperating
in on-grid and the limited
insystems.
on-grid availability
systems.
Among itsAmong window of solar
its advantages,
advantages, the imple-
irradiation,
the its generation
implementation and is limited
approval to
of meeting
on-grid demand
storage at
systems,the same
will
mentation and approval of on-grid storage systems, will result in the possibility of altering time
result it
in is
the being gen-
possibility
erated.
the energy demand and generation curves, greater stability of the electricity grid, sys-
of Until
altering now,
the the
energy existing
demand rules
and in Brazil
generation do not allow
curves, the
greater use of energy
stability of storage
the electricity
better
tems
use connected
grid, better
of use to
intermittentof the power grid,
intermittent
renewable making
renewable
energy sourcessome
energy and, generation
sources and,
consequently, systems less efficient.
consequently,
reductions reductions
in energyHow- in
con-
ever,
energythis
sumption scenario
consumption is
and carbonand expected to change
carbon emissions.
emissions. in the coming years, since, in 2016, the ANEEL
created a strategic
Countries
Countries suchpublic
such as
asthenotice
the United for States,
United research
States,Japan,and development
Japan, Korea,
Korea,China
Chinaand (R&D)
andGermanyon battery
Germany have
have energy
imple-
imple-
storage
mented technologies through public call N° 21/2016. The project
mented storage systems through policies and incentives for their use, and by 2018,there
storage systems through policies and incentives for aims
their to
use, develop
and by and
2018, test
there
were
energy approximately
storage systems, 9 GW/17
operating GWhin of energy
on-grid storage
systems. power
Among
were approximately 9 GW/17 GWh of energy storage power capacity in operation globally capacity
its in
advantages, operation
the glob-
imple-
ally [6]. Although
mentation and approval storage systems
of on-grid continue
storage to be will
systems, developed,
result inathe study by [7] reports
possibility of alteringthat
the implementation of new materials resulted in a 87% reduction
the energy demand and generation curves, greater stability of the electricity grid, better in the cost of lithium-ion
use of intermittent renewable energy sources and, consequently, reductions in energy con-
sumption and carbon emissions.
Countries such as the United States, Japan, Korea, China and Germany have imple-
mented storage systems through policies and incentives for their use, and by 2018, there
were approximately 9 GW/17 GWh of energy storage power capacity in operation globally
Energies 2021, 14, 5284 3 of 21
batteries from 2010 (1183 US/kWh) to 2019 (156 US/kWh). Currently technologies such as
lithium iron phosphate (LiFePO4 ), lithium-nickel cobalt aluminum (NCA), lithium oxide
cobalt (LCO), lithium-nickel-manganese-cobalt (NMC), lithium-manganese spinel (LMO)
or lithium titanate (LTO) are used in storage systems with which longer life cycles, high
power capability, high discharge rates, and constant voltage profiles can be obtained [8–10].
There are still hundreds of locations in the world that have difficulty or no connection
to their respective countries main electricity grid. In Brazil, ONS, the National Electric
System Operator, reports that there are 212 isolated locations [11]. Specifically in the
Amazon region, public service utilities have difficulties in providing electricity services
in forested regions, mainly due to the difficulties or total impossibility of reaching them
due to the characteristics of the region. In these regions, when supplied, the electricity is
mainly provided by community generators powered by fossil fuels or photovoltaic energy,
normally installed with stationary lead acid batteries, which have a reduced useful lifespan.
Experiments and studies show that hybrid systems are more efficient than diesel systems
both in economic and environmental terms, especially in regions without a connection to
the power grid, revealing themselves as a possible solution to these problems.
To improve the efficiency of distributed generation, whether operating with a hybrid
system (HS) and storage system, it is also recommended to use technologies to implement
smart grid features, to manage the variability of generation sources and the demand of users
and the electric grid itself. The smart grid aims to make each process efficient (generation,
storage and consumption), making the service continuous, reliable and economical.
Studies considering hybrid systems including photovoltaic-diesel-wind and storage
systems have been presented in [12–17]. In this HS model, both batteries and genera-
tors perform backup functions for continuity of service. However, given the operational
characteristics of batteries, such as better control of voltage and frequency, as well as less
environmental impact, these are activated first than diesel generators. With this, the HS
has more flexibility to manage the electric grid energy balance and the useful life of the
equipment, increasing the reliability, continuity, and stability to the loads.
In some European communities, there has been discussion about the reduction of
fossil fuels through the use of renewable generation sources and energy management with
the help of storage systems, even if connected to the utility’s electricity grid, transforming
themselves into self-sustainable communities, as shown in the works [18–20]. In Brazil, due
to local legislation, only large companies have implemented self-generation, in addition to
managing energy generation and demand, using by-products from their production line as
fuel for steam heating in boilers coupled to electric turbines (thermoelectric) [21,22].
Previous studies have separately discussed the use of renewable sources as a way
to increase energy generation, as shown [23–25]. The work elaborated by [26,27] shows
the benefits of using the energy storage and management system to demand control as a
financial attraction.
Reference [28] highlights the importance of smart meters in electrical grids. Refer-
ences [29–31] describe the computational development and the web module of an energy
management system implemented in a university, as a way to mitigate possible problems
in electrical energy quality, measure the efficiency of systems and reduce energy costs
electric. To identify abnormalities in electricity consumption, references [32,33] used heuris-
tic methods to detect unexpected consumption behaviors, aiming to reduce waste and,
consequently, unnecessary energy costs. In [34], a stochastic method was sought to explain
the uncertainties of both consumption and energy generation.
Using different simulation methods, the integration of renewable sources, storage and
diesel generation was present in other studies such as [34–37]. In [38], the energy quality
of a photovoltaic-diesel hybrid mini-grid (HMG) was analyzed using the simulation of
input and output loads in the building. Thus, the main research and results found in the
management of HMG are based on computer simulations, where few validate the results
obtained experimentally.
Energies 2021, 14, 5284 4 of 21
A summary of the literature review related to HMGs and energy management, classi-
fied according to the approaches analyzed in this work, is presented in Table 1.
Table 1. Review of literature regarding hybrid mini-grid and energy management.
Papers’ Main Subject Relevant Literature

Energetic balance [1,18–22,27,35,36,39]
Renewable energy [2–5,23–25,40]
Storage system [6–10,26,41]
Off-grid system [11,12]
Hybrid systems [13–17,37,38]
Smart meters [28–34]
Zero energy buildings [42]
Statistics for data analysis [43]
Electric Power Quality [44]
The main contributions of this study is to provide real data on the operation and
management of a HMG integrated with a utility’s electricity grid, as a strategy to reduce
electricity costs, improve operational safety, increase the efficiency of using intermittent
sources, in addition to discussing compensation at smaller scales, through the application
of the HMG to small and medium-sized consumers in any region of the country, such
as universities, residential and commercial condominiums, hospitals, shopping malls,
among others. In addition, there is also the possibility of using hybrid systems in isolated
communities that have difficulty in electrification via cables. As a way of exemplifying
the contributions of this article for academic application and for future installations, the
following are discussed:
• Technical and operational analysis of real data from the HMG in off-grid and on-
grid modes;
• Statistical analysis of the management system behavior, according to stochastic vari-
ables such as load, photovoltaic generation, storage system and diesel generator;
• The economic analysis of the financial benefits arising from energy management in
small- and large-scale applications.
This work analyzes the functioning of a true hybrid system installed at the Center
for Energy Efficiency in the Amazon (CEAMAZON) of the Federal University of Pará
(UFPA), and composed of photovoltaic systems, electricity grid, diesel generator (DG)
and a LiFePO4 battery storage system. As a way to organize the article, it was divided as
follows: Section 2 provides a description of the HMG, composed of photovoltaic systems,
storage system, diesel generator, CEAMAZON building loads and the energy management
system. Section 3 presents the management system methodology used to manage the
energy generation and demand from the HMG operating in on-grid and off-grid modes.
Section 4 shows the technical analysis of three real scenarios of HMG operation, including
results, technical solutions and possible applications in society through new business
models for energy consumers and utilities. Section 5 presents a financial analysis, given
the implementation of the HMG with energy management. Finally, Section 6 summarizes
the main conclusions of the work.
2. Case Study
The hybrid system analyzed in this study includes three mini-plants of photovoltaic
solar energy, a diesel generation plant, a battery storage system, using lithium-ion batteries,
forming a HMG connected to the electric utility grid, located in the CEAMAZON at UFPA,
in the city of Belém, Pará, Brazil. The system is an integral part of the research and devel-
opment project (R&D) “Intelligent System for the Efficient Management of Multimodal
Electric Mobility” with financial support from Norte Energia S.A. a hydrogeneration utility
in the State of Pará. Figure 3 shows some images of the system and its location.
solar energy,
The hybrid system aanalyzed
diesel generation plant,
in this study a battery
includes storage
three system,ofusing
mini-plants lithium-ion batter-
photovoltaic
ies, forming a HMG connected to the electric utility grid, located in
solar energy, a diesel generation plant, a battery storage system, using lithium-ion batter- the CEAMAZON at
ies, forming a HMG connected to the electric utility grid, located in the CEAMAZON research
UFPA, in the city of Belém, Pará, Brazil. The system is an integral part of the at and
development project (R&D) “Intelligent System for the Efficient Management
UFPA, in the city of Belém, Pará, Brazil. The system is an integral part of the research and of Multi-
Energies 2021, 14, 5284 developmentmodal Electric
project (R&D)Mobility” with System
“Intelligent financialfor
support from Norte
the Efficient Energia S.A.
Management a hydrogenera-
of Multi- 5 of 21
tion utility
modal Electric in the
Mobility” State
with of Pará.support
financial Figure 3from
shows some
Norte imagesS.A.
Energia of the system and its location.
a hydrogenera-
tion utility in the State of Pará. Figure 3 shows some images of the system and its location.
Figure 3. Installed hybrid system at the Federal University of Pará, in Belém, Pará, Brazil.
Figure 3. Installed hybrid system at the Federal University of Pará, in Belém, Pará, Brazil.
Figure 3. Installed
The hybrid system
hybrid system at the Federal
components University
and their of Pará,
respective in Belém, Pará,
descriptions Brazil. below
are detailed
and presented
The hybrid system in Figure 4. and their respective descriptions are detailed below
components
The hybrid system components and their respective descriptions are detailed below
and presented
and in Figure 4.in Figure 4.
presented
Figure 4. Hybrid system components at CEAMAZON.
Figure 4. Hybrid
2.1. Photovoltaic system components at CEAMAZON.
Generation
Figure 4. Hybrid system components
The photovoltaic at CEAMAZON.
(PV) panels were installed on ground supports, positioned to the
north, with a 12◦ azimuth inclination, to make the panels self-cleaning efficient. The PV
renewable energy generation is composed of three PV systems totaling 80 kWp as presented
in the following.
• PV-CEAMAZON 01 (58 kWp): Represents the classic grid-connected photovoltaic
system, that is, for its operation, it is necessary that the University’s distribution grid
be operating normally, otherwise the system will be automatically disconnected. This
system has a 50 kW inverter.
• PV-CEAMAZON 02 (16 kWp): It is connected to the CEAMAZON mini-grid. There-
fore, in the event of an electrical grid fault, this generation system will operate normally
to supply energy to the mini-grid in off-grid mode, meeting the buildings’ loads. This
system, that is, for its operation, it is necessary that the University’s distribution grid
be operating normally, otherwise the system will be automatically disconnected. This
 PV-CEAMAZON 02 (16 kWp): It is connected to the CEAMAZON mini-grid. There-
Energies 2021, 14, 5284 fore, in the event of an electrical grid fault, this generation system will operate nor- 6 of 21
mally to supply energy to the mini-grid in off-grid mode, meeting the buildings’
loads. This system has a 15 kW inverter.
 PV-CEAMAZON 03 (6 kWp): This system will be exclusively for supplying the stor-
• PV-CEAMAZON 03 (6 kWp): This system will be exclusively for supplying the storage
age system, so that its entire generation will meet the battery demand for stored en-
system, so that its entire generation will meet the battery demand for stored energy
ergy that in turn, will serve as a backup for the hybrid system. This system is con-
that in turn, will serve as a backup for the hybrid system. This system is connected to
nected to the direct current (DC) bus through a charge controller.
the direct current (DC) bus through a charge controller.
2.2. Management System
2.2. Management System
To manage the Togeneration sources,
manage the the storage
generation system
sources, operation,
the storage as well
system as protection
operation, as well as protection
devices and loads,
devices and loads, a centralized system controller was used whichby
a centralized system controller was used which is composed is six
composed by six
inverters/charger controllers with a total power of 60 kW. The management system allows
inverters/charger controllers with a total power of 60 kW. The management system allows
remote parameter configuration
remote parameter as well as real-time
configuration monitoring
as well directly
as real-time from the
monitoring system
directly from the system
webpage. Figure 5 presents
webpage. an example
Figure 5 presentsofan
the available
example information
of the available that can be accessed
information that can be accessed in
in the management system webpage.
the management system webpage.
Figure 5. Hybrid system

Figure components
5. Hybrid systematcomponents
control panel at CEAMAZON.
at control panel at CEAMAZON.
2.3. Diesel Generation

2.3. Diesel Generation
An existing DGAn
setexisting
was usedDGinset
this study,
was usedwith a 50
in this kW nominal
study, with a 50power in stand-by
kW nominal power in stand-by
mode and 45 kW nominal
mode and 45power in primepower
kW nominal mode.in prime mode.
2.4. Storage System

The storage system consists of a set of eight lithium iron phosphate (LiFePO4 ) batteries
with a usable energy of 13.8 kWh each one, totaling 110.4 kWh, and an individual capacity
of 270 Ah, totaling a 2160 Ah capacity. For the charging mode, a 0.035 C rate was defined
with 75 A. The batteries are equipped with battery management system (BMS) and battery
management unit (BMU) that allow correct operation and communication with external
inverters/charger for a safe and reliable operation that extends their useful life.
3. Methodology for Hybrid Mini-Grid Operation Parameters Configuration

Despite the large solar irradiation in the region where the HMG is located, the season-
ality of the rainy season, between November and March, affects PV generation. For the
HMG operation, different configurations and parameters were defined for both on-grid and
off-grid operation. Therefore, the proposed methodology for the management system aims
to prioritize the use of renewable sources, predicting critical low solar availability scenarios,
Energies 2021, 14, 5284 7 of 21
unavailability of energy supply by the utility, meeting the energy demand required by the
building load, the cost of energy at times of largest consumption and also at times when
energy is more expensive.
The on-grid and off-grid operation configurations are implemented as described in
the flowchart of Figure 6. These configurations generally aim to reduce electricity costs
Energies 2021, 14, x FOR PEER REVIEW 8o
and increase the energy availability to the CEAMAZON building.
Figure 6. Hybrid system operation configurations flowchart.

Figure 6. Hybrid system operation configurations flowchart.
4. Technical Evaluation of the Hybrid Mini-Grid
The off-grid management mode aims to maximize the use of photovoltaic generation
and the batteries storedFor the on-grid
energy and off-grid
before using theoperating
dieselconditions
generatortechnical and economic
set. During off-grid evaluati
three scenarios will be evaluated in this work and
operation the PV-CEAMAZON 01 system will be out of service, according to ABNT NBR are presented in Table 2.
IEC 62116: 2012 [39], operational anti-island protection regulation in Brazil. However, the
Table 2. Operation scenarios for the hybrid system sources.
PV-CEAMAZON 02 and 03 systems will continue to generate energy to directly supply the
loads and batteries, respectively.
Source Scenario 1 Scenario 2 Scenario 3
At time a grid power failure
Grid occurs, the off-grid
ON operation mode is
OFF activated, so the OFF
batteries assumePV-CEAMAZON—01
the loads, maintaining the service ON continuity and reliability.
OFF The system OFF
controller will carry out the energy balance between
PV-CEAMAZON—02 ON the use of the PV ON generation systems, ON
supplementing with the battery
PV-CEAMAZON—03 stored energy, if necessary.
ON Even so, if the
ON energy supply via ON
the photovoltaic andStorage
batterysystem
sources is not enough, the
OFF (Off Peak hours) system controllerONwill start the OFFDG (* SoC < 20
set to service the loads in the CEAMAZON building. The mini grid management follows
OFF (Until Batteries *
Diesel
the instructions of the generator
battery manufacturer, who OFFrecommends a maximum discharge of ON
SoC = 20%)
80%, that is, its lower charging limit must not be less than 20%. In the case of an off-grid
* SoC—State of Charge.
condition and a battery SoC less than 20%, the diesel generator will assume both the loads
and the storage system Thecharge until the
first scenario utility
looks grid
at the energy
hybrid is available.
system operation in on-grid conditions, aim
The on-grid at operation
the operationaims in to carry
urban out connected
areas the mini-gridto theenergy
utility’sbalance,
electric in order
grid and todisconnect
reduce the expenses with
from the electricity,
grid at scheduledat peak timesThis
periods. andscenario
in the hours
showsof thehighest
benefitsdemand
for consumers c
from the grid, innected
addition
to thetoutility
guaranteeing
grid throughthe energy
HMG. The quality duringoftemporary
evaluation the demand off-grid
curve, in combi
operation. While the external grid is available (on-grid operation), the three photovoltaic
tion with renewable energy generation and the storage system can be useful to red
systems (PV-CEAMAZON
peak demand01, and02energy
and 03) will becost
at higher operating
from thenormally,
utility grid.according to the
This scenario can be used
availability of sunlight, and the batteries will start operating in the programmed
residential, commercial, public and industrial consumers, located in cities and who s periods,
that is, peak times and times
flexibility of highest
in demand demand from the grid, and temporarily operating
management.
Then, the second scenario aims to simulate on-grid operation with the need for s
den off-grid operation, due to an interruption in the supply of the local utility’s electric
grid. This scenario aims to present the performance, system security and power quality
the energy supply, only with the use of photovoltaic and storage systems during the l
of the utility grid. Given the reliability, this scenario shows that HMG with batteries
Energies 2021, 14, 5284 8 of 21
in off-grid mode in case of grid fault conditions. In this condition, the DG will not start
operating, even if the batteries reach the limit of equal or less than 20% of their capacities.
4. Technical Evaluation of the Hybrid Mini-Grid

For the on-grid and off-grid operating conditions technical and economic evaluation,
three scenarios will be evaluated in this work and are presented in Table 2.
Table 2. Operation scenarios for the hybrid system sources.
Source Scenario 1 Scenario 2 Scenario 3

Grid ON OFF OFF
PV-CEAMAZON—01 ON OFF OFF
PV-CEAMAZON—02 ON ON ON
PV-CEAMAZON—03 ON ON ON
Storage system OFF (Off Peak hours) ON OFF (* SoC < 20%)
Diesel generator OFF OFF (Until Batteries * SoC = 20%) ON
* SoC—State of Charge.
The first scenario looks at the hybrid system operation in on-grid conditions, aiming at
the operation in urban areas connected to the utility’s electric grid and disconnecting from
the grid at scheduled periods. This scenario shows the benefits for consumers connected to
the utility grid through HMG. The evaluation of the demand curve, in combination with
renewable energy generation and the storage system can be useful to reduce peak demand
and energy at higher cost from the utility grid. This scenario can be used by residential,
commercial, public and industrial consumers, located in cities and who seek flexibility in
demand management.
Then, the second scenario aims to simulate on-grid operation with the need for sudden
off-grid operation, due to an interruption in the supply of the local utility’s electricity grid.
This scenario aims to present the performance, system security and power quality in the
energy supply, only with the use of photovoltaic and storage systems during the loss of the
utility grid. Given the reliability, this scenario shows that HMG with batteries can be used
in places that have an extreme need for continuity of operation, such as hospitals, airports,
railways, ports, etc.
Finally, the third scenario also analyzes the off-grid operation after an energy inter-
ruption from the grid, however, the DG set operation was included, due to the batteries
reaching their minimum limit (SoC < 20%) and the system being with little or no solar
availability. This scenario can be replicated both in locations connected to the electrical
grid, as in the second scenario, having even more security due to the redundancy of power
supply; and in isolated locations, without the assistance of utility grid and which, for
operational safety, can also activate the DG set until it has re-established the battery charge
and/or energy through renewable sources, such as solar photovoltaic.
4.1. First Scenario: Normal Operation

4.1.1. Generation and Demand of Energy from the Hybrid Mini-Grid
The first scenario involves the production of electric energy from the three photovoltaic
systems (PV-CEAMAZON 01, 02 and 03), to meet the loads of the mini-grid (CEAMAZON
building and batteries) and the surplus is made available to UFPA’s internal electrical grid.
However, during the off-grid operation, the loads in the CEAMAZON building will be
served only by the PV-CEAMAZON 02 and 03 systems, inserted in the mini-grid, while
the PV-CEAMAZON 01 system will be shut down. Figure 7 shows the energy production
curve of the three photovoltaic systems, over a week. In this period, the three photovoltaic
systems achieve a production of up to 64 kWp.
taic systems (PV-CEAMAZON 01, 02 and 03), to meet the loads of the mini-grid (CEAMA-
ZON building and batteries) and the surplus is made available to UFPA’s internal electri-
cal grid. However, during the off-grid operation, the loads in the CEAMAZON building
will be served only by the PV-CEAMAZON 02 and 03 systems, inserted in the mini-grid,
Energies 2021, 14, 5284 while the PV-CEAMAZON 01 system will be shut down. Figure 7 shows the energy pro-
9 of 21
duction curve of the three photovoltaic systems, over a week. In this period, the three
photovoltaic systems achieve a production of up to 64 kWp.
Figure7.7.Total
Figure Totalphotovoltaic
photovoltaicgeneration
generationininthe
theCEAMAZON
CEAMAZONmini
minipower
powerplant,
plant,within
withina aweek.
week.
Inthe
In theState
StateofofPará,
Pará,Brazil,
Brazil,for
forgroup
groupAAconsumers,
consumers,there
thereare
aredifferent
differenttariff
tariffoptions
options
that allow the customer to choose the contracted demand value (kW), according tototheir
that allow the customer to choose the contracted demand value (kW), according their
loadprofile.
load profile.UFPA
UFPA opted
opted forfor
thethe green
green tariff
tariff modality,
modality, which which
has ahas a contracted
contracted demand demand
(kW)
(kW)
for for all weekdays
all weekdays and a difference
and a difference in tariffinfor
tariff for consumption
consumption (kWh) during
(kWh) during off-peakoff-peak
hours
hoursa.m.
(00:00 (00:00toa.m.
18:29top.m.
18:29and
p.m.21:31
and p.m.
21:31 top.m. to 23:59
23:59 p.m.)p.m.) and
and at at peak
peak hourhour
(18:30(18:30
p.m.p.m.
to
to 21:30
21:30 p.m.)
p.m.) [40].[40].
For For the green
the green hour-seasonal
hour-seasonal tarifftariff modality,
modality, the for
the tariff tariff for energy
energy con-
consump-
tion
Energies 2021, 14, x FOR PEER REVIEW (kWh)(kWh)
sumption duringduring
peak hours is almost
peak hours 10 times
is almost 10 higher than the
times higher thanvalue duringduring
the value off-peak
10 ofoff-
21
hours. UnlikeUnlike
peak hours. residential sectorssectors
residential and some commercial
and some sectors,
commercial UFPAUFPA
sectors, and CEAMAZON
and CEAMA-
have
ZONpeakhavedemand duringduring
peak demand the daythe
(off-peak hours),hours),
day (off-peak however, even if even
however, reduced, they main-
if reduced, they
tain considerable
maintain consumption
considerable duringduring
consumption peak hours.
peak hours.
Figure 8 shows the load curve characteristic at UFPA and the CEAMAZON building,
Figure 8 shows the load curve characteristic at UFPA and the CEAMAZON building,
considering the period of teaching classes from Monday to Friday and highlighting peak
considering the period of teaching classes from Monday to Friday and highlighting peak
hours. Although the demand for the CEAMAZON building represents 0.7% of UFPA’s
hours. Although the demand for the CEAMAZON building represents 0.7% of UFPA’s
maximum demand, it is possible to verify that both have the same behavior.
maximum demand, it is possible to verify that both have the same behavior.
Figure 8.
Figure Energy demand
8. Energy demand at
at UFPA
UFPA and
and CEAMAZON.
CEAMAZON.
Analyzing the HMG behavior, it was found that it is possible to carry out the energy
balance between generation, storage, and demand of energy, providing better perfor-
mance of generation sources and less use of the external electrical grid. In addition, the
Energies 2021, 14, 5284 10 of 21
Analyzing the HMG behavior, it was found that it is possible to carry out the energy
balance between generation, storage, and demand of energy, providing better performance
of generation sources and less use of the external electrical grid. In addition, the energy
surplus generated by photovoltaic systems that was delivered to the utility, can be con-
sumed in times of higher energy demand or less photovoltaic generation at other times
of the year. Considering the building’s energy management, the energy balance allows
to outline the CEAMAZON be defined as a zero-energy building (ZEB), since the energy
supplied by the utility grid will be less than the generated renewable energy [42].
4.1.2. Scheduled Operation of the Hybrid Mini-Grid to Demand Control

Given the batteries energy charge/discharge functionality and grid connectivity, an
operation cycle was programmed in the CEAMAZON management system, in such a way
that the HS was programmed to operate in off-grid condition between 10:00 a.m. and
14:00 p.m. (highest demand period) and from 18:30 p.m. to 21:30 p.m. (peak hours). The
battery is recharged at the remaining periods, when exists solar potential and/or when
the utility energy cost is cheaper. Thus, at these times, the power supply is interrupted by
the external electrical grid, the HMG operates in off-grid mode and the load is fed by the
storage system. Figure 9 shows the HMG behavior on a typical weekend day and during
the week (for the example, Sunday and Monday were used, respectively) considering the
energy curves supplied by the external electrical grid, batteries and CEAMAZON building
loads consumption. On weekends the storage system is in standby condition, due to the
lack of peak hours on those days, in addition, due to low consumption and the incidence
of photovoltaic generation, HMG supplies energy to the external grid. During the week, it
is possible to check the times when the battery is supplying power to the building’s loads
and when it is being charged. In reliability terms, according to the consumption curve (red
color of Figure 9), which represents the building loads, it is possible to verify that HMG is
able to keep the loads supply continuity, without any loss in the supplied energy power,
even in situations of sudden changes in the loads or solar power along the day.
Figure 9.
Figure 9. Total
Total power
power balance
balancefor
forhybrid
hybridmini-grid
mini-gridoperation.
operation.
As
As seen
seen in
in Figure
Figure 9,
9, the batteries are in a discharging cycle period between 10:00 a.m.
and
and 14:00
14:00 p.m.
p.m. (highest
(highest demand
demand period)
period) and
and from
from 18:30
18:30 p.m.
p.m. to
to 21:30
21:30 p.m.
p.m. (peak
(peak hours)
hours)
and recharging at other times. In the 06:00 a.m. to 18:00 p.m. time period, the storage
and recharging at other times. In the 06:00 a.m. to 18:00 p.m. time period, the storage sys-
system relies
tem relies onon thethe powersupply
power supplyofofthe
thePV-CEAMAZON
PV-CEAMAZON02 02 and
and 03
03 systems.
systems. Figure
Figure 10
10
shows the mini-grid energy flow, over a week. During this period, the energy balance was
carried out, so that the operation model of the batteries and photovoltaic systems would
meet the building load. The external grid was also used as storage and energy supply
source, according to the energy supply or need of the HMG. The operation allowed the
mini-grid to end with a positive balance of 113.33 kWh at the end of a week, which repre-
Figure 9. Total power balance for hybrid mini-grid operation.
As seen in Figure 9, the batteries are in a discharging cycle period between 10:00 a.m.
Energies 2021, 14, 5284 and 14:00 p.m. (highest demand period) and from 18:30 p.m. to 21:30 p.m. (peak hours) 11 of 21
and recharging at other times. In the 06:00 a.m. to 18:00 p.m. time period, the storage sys-
tem relies on the power supply of the PV-CEAMAZON 02 and 03 systems. Figure 10
shows the mini-grid energy flow, over a week. During this period, the energy balance was
shows the mini-grid energy flow, over a week. During this period, the energy balance
carried out, so that the operation model of the batteries and photovoltaic systems would
was carried out, so that the operation model of the batteries and photovoltaic systems
meet the building load. The external grid was also used as storage and energy supply
would meet the building load. The external grid was also used as storage and energy
source, according to the energy supply or need of the HMG. The operation allowed the
supply source, according to the energy supply or need of the HMG. The operation allowed
mini-grid to end with a positive balance of 113.33 kWh at the end of a week, which repre-
the mini-grid to end with a positive balance of 113.33 kWh at the end of a week, which
sents 5.22% of the consumption of a week in the CEAMAZON building. In some countries,
represents 5.22% of the consumption of a week in the CEAMAZON building. In some
this amount of energy surplus could be sold on the market, however, in Brazil, this energy
countries, this amount of energy surplus could be sold on the market, however, in Brazil,
surplus can surplus
this energy be offsetcan
at other times,
be offset at but it cannot
other be sold,
times, but according
it cannot to the
be sold, actual regula-
according to the
tion.
actual regulation.
discharge until reaching 21%. Under these conditions, the system operation as well as the
Figure
batteries
Figure 10.
10. Total energy balance
charging/discharging
Total energy balancecycles
in the
in the hybrid
hybrid mini-grid.
will bemini-grid.
analyzed. Figure 11 presents the power
curves in off-grid conditions.
4.2. Second
Second Scenario:
Initially, Off-Grid Operation
Off-Gridhow
it is observed Operation
the electrical grid is feeding both the loads and charging
the storage
For thesystem, when at 6:00
second scenario, a.m., anoperating
the system
system interruption
operating in in the utility
in off-grid
off-grid grid will
conditions
conditions service
will be was pro-
be analyzed.
analyzed.
grammed
For in athe
this case, management
failure system,
in the utility simulating
grid was an external
simulated, fault. Aswere
and the batteries initially com-
placed to
mented,
dischargeinuntil
grid reaching
absence conditions,
21%. Under thethese
photovoltaic
conditions,system 01 (59 kWp)
the system is disconnected
operation as well as
due to the anti-islanding
the batteries protection,
charging/discharging and in
cycles this
will be way the system
analyzed. Figureis11operating
presents in
theoff-grid
power
mode
curveswith the PV-CEAMAZON
in off-grid conditions. 02 and 03 photovoltaic systems only.
Figure 11. Storage system operation cycle in off-grid conditions.
Initially,
The it is observed
off-grid operationhow the was
period electrical grid is feeding
determined both the loadsdemanded
by the consumption and charging
ac-
the storage
cording system, loads
to building when and
at 6:00 a.m.,
lasted anThe
9 h. interruption
transitionin the the
from utility grid in
on-grid service
to thewas pro-
off-grid
grammed in
condition andthevice
management system,
versa, occurs simulating anwithout
instantaneously externalany
fault. As initially
impact on thecommented,
connected
loads, so that when the photovoltaic systems are switched on, the load delivered by the
storage system turns out to be the difference between the demanded power and the pho-
tovoltaic solar generation. Finally, when the batteries SoC reaches 21%, the electrical grid
is restored automatically, at approximately 15 h:00 p.m. As soon as the system is reestab-
Energies 2021, 14, 5284 12 of 21
in grid absence conditions, the photovoltaic system 01 (59 kWp) is disconnected due to the
anti-islanding protection, and in this way the system is operating in off-grid mode with the
PV-CEAMAZON 02 and 03 photovoltaic systems only.
The off-grid operation period was determined by the consumption demanded accord-
ing to building loads and lasted 9 h. The transition from the on-grid in to the off-grid
condition and vice versa, occurs instantaneously without any impact on the connected
loads, so that when the photovoltaic systems are switched on, the load delivered by the
storage system turns out to be the difference between the demanded power and the photo-
voltaic solar generation. Finally, when the batteries SoC reaches 21%, the electrical grid
is restored automatically, at approximately 15 h:00 p.m. As soon as the system is reestab-
lished, the external grid starts to recharge the battery, however, at 18:30 p.m. (peak time)
the battery will be used again, until reaching its minimum limit (20%) again. The battery is
not recharged by the external grid during peak hours, being recharged after 21:30 p.m.
The off-grid condition does not represent any impact for the system voltages and
frequency,
Energies 2021, 14, x FOR PEER REVIEW and in Figure 12a,b it can be seen how there is no difference between the on-grid
13 of 21
and off-grid conditions when the load is powered by the batteries. A detail of the storage
system operation in its charge/discharge cycle will be presented in the next section.
(a) (b)
Figure 12. (a)
Figure 12. (a) voltage
voltage and
and (b)
(b) frequency variation during
frequency variation during discharging
discharging period
period in
in storage
storage system.
system.
Storage System Assessment in Charge/Discharge

Charge/Discharge Cycle
Cycle
The storage system is composed by eight lithium iron phosphate (LiFePO44) batteries,
characterized
characterized by byaatwo-phase
two-phasetransition
transitionduring
duringcharging
chargingandand
discharging
discharging cycles withwith
cycles dif-
different
ferent open
open circuit
circuit voltages
voltages (OCV)
(OCV) [41].Each
[41]. Eachbattery
batterypackage
packageconsists
consistsofoftwo
two25.6
25.6 Volt,
Volt,
270 Ah series batteries. In turn, each battery package is composed of eight cells in series
with a nominal voltage of 3.2 Volts and and an an open
open circuit
circuit voltage
voltage of
of 3.8
3.8 Volts.
Volts. Figure 13
the hysteresis
presents the hysteresisvoltage
voltagecurve
curveduring
duringthe the storage
storage system
system charging/discharging
charging/discharging cy-
cycle,
cle, a C-rate
a C-rate ofof 0.035
0.035 was
was configuredfor
configured forthe
thecharging
chargingcycle,
cycle,with
withan
anapproximate
approximate recharge
time of 28.8 h at 75 Amperes.
As can be seen in Figure 13, the discharge and charge curves show some differences in
voltage levels for the same battery SoC, within both charge and discharge curves, and the
lowest voltage gap occurs at SoC = 27%, when the difference between charge and discharge
voltage is 2.34 Volts. It also can be seen how the discharge cycle presents a random voltage
variation, mainly due to the instability of photovoltaic generation and building loads
demand variation. When a SoC = 21% is reached, the storage systems stop feeding the
loads and the on-grid condition is restored. At this point the charging cycle starts with a
75 Amps charging current from the grid and PV systems until batteries are fully charged
again, then the voltage decreases to the floating voltage value defined in 55 Volts (pink
points), aiming to maintain the battery in fully charged state (trickle charging).
Figure 13. Storage system voltage discharging/charging cycle.

with a nominal voltage of 3.2 Volts and an open circuit vol
presents the hysteresis voltage curve during the storage syste
cle, a C-rate of 0.035 was configured for the charging cycle, wi
Energies 2021, 14, 5284 13 of 21
time of 28.8 h at 75 Amperes.

By making a correlation between the voltage, the charge status and the power deliv-
ered by the battery, through Figure 14a,b, it can be verified that when SoC < 55% and the
power
power >>12.5 As can be seen in Figure 13, the discharge and charge cu
12.5kW,
kW,lower
lower voltage
voltage levels
levels in
inthe
thedischarge
discharge process
process are
are observed,
observed, asas presented
presented
in
in Figure
Figure 14a.
14a. However,
However, during
during the
the charging
charging process,
process, aa constant
constant voltage
voltage is
is observed
observed in in
in voltage levels for the same battery SoC, within both charge
Figure
Figure 14b.
14b. The main difference
differencebetween
betweenthe thecharge
chargeand
anddischarge
dischargediagrams
diagrams forfor
thethe stor-
storage
the lowest voltage gap occurs at SoC = 27%, when the differen
age system,
system, observed
observed in Figures
in Figures 14 and
14 and 15,that
15, is is that
the the charge
charge process
process is constant,
is constant, at aatC-rate
a C-
of 0.035
rate (75 A),
of 0.035 while
(75 A), the the
while second depends
second depends on on
thethe
demand
demand of the building
of the loads,
building loads,varying
vary-
charge voltage is 2.34 Volts. It also can be seen how the discha
throughout
ing throughoutthe the
day.day.
Dicharging Voltage Charging Cycle
0
Battery
voltage variation, mainly due to the instability of photovolta
5,000 Battery
Voltage
-5,000
< 52
52 – 53
loads demand variation. When a SoC = 21% is reached, the st Voltage
< 52
52 – 53
53 – 54 53 – 54
54 – 55 the loads and the on-grid condition is restored. At this point th
4,000
54 – 55
Battery Power (W)
Battery Power (W)
55 – 56 55 – 56
-10,000
> 56 > 56
a 75 Amps charging current from the grid and PV systems unt
3,000
-15,000
again, then the voltage decreases to the floating voltage valu
2,000
-20,000 points), aiming to maintain the battery in fully charged state (

1,000
-25,000 By making a correlation between the voltage, the charge

0
30 40 50
ered by the battery, through Figure
60 70 80
Battery State of Charge (%)
90
14a,b, it can be verified th
20
Battery State of Charge (%)
30 40 50 60 70 80 90 100
(a) (b)
Figure
Figure14.
14.Voltage
Voltage variation
variation in
in function
function of
of batteries
batteries power
power and
and SoC
SoC for:
for: (a)
(a) discharging
discharging cycle
cycle and
and (b)
(b) charging
charging cycle.
cycle.
Temperature
Temperature is is another
anotherimportant
importantparameter
parameterininthe
the batteries’
batteries’ operation,
operation, and and to ana-
to analyze
lyze it, correlation plots of its variation in function of the SoC and the power
it, correlation plots of its variation in function of the SoC and the power delivered in the delivered in
the charge/discharge
charge/discharge cyclecycle
were were created
created andpresented
and are are presented in Figure15.
in Figure In discharging
15. In discharging mode
mode
(Figure(Figure
15a), it15a), it is observed
is observed how thehow the temperature
temperature increasesincreases as of
as the state thecharge
state of charge
decreases,
decreases,
mainly formainly
SoC < 45% for SoC
and <discharge
45% andpowers
discharge
> 15powers
kW. For > the
15 kW. For the
charging charging
cycle, on thecycle,
other
on the other hand, the temperature does not show considerable increases, with higher
temperatures at the beginning of the recharge process and in the final phase, as shown in
Figure 15b.
Dicharging Temperature Charging Cycle
0
Battery 5,000 Battery
Figure 14. Voltage variation in function of batteries power and SoC for: (a) discharging cycle and (b) charging cycle.
2
6 ∑𝑛𝑖=1 𝐷𝑖
𝑟𝑠 = 1 − (1)
Temperature is another important parameter 𝑛(𝑛2in−the 1) batteries’ operation, and to ana-
lyze it, correlation plots of its variation in function of the SoC and the power delivered in
where
the n is the numbercycle
charge/discharge of pairswere of created
values and and 𝐷are 𝑖 = presented
𝑋𝑖 − 𝑌𝑖 is in theFigure15.
differenceInbetween
dischargingeach
Energies 2021, 14, 5284 corresponding
mode (Figure 15a), 𝑋 and 𝑌 value rank.
𝑖 it is 𝑖observed how the temperature increases as the state of charge 14 of 21
The correlation
decreases, mainly foranalysis
SoC < 45% results
andindischarge
a numberpowers between> −1 15 and
kW.+1, Forcalled the correlation
the charging cycle,
on the other hand, the temperature does not show considerable increases,the
coefficient. The higher the coefficient, the closer the relationship between withvariables.
higher
The analysis was
temperatures at theperformed
beginning for the
thestorage system charging
and in and discharging cycle.
hand, the temperature does notofshow recharge
considerable process increases, the
withfinal
higherphase, as shown in
temperatures at
Figure In the
15b. correlation matrix, the upper cell shows the Spearman coefficient while the
the beginning of the recharge process and in the final phase, as shown in Figure 15b.
lower cellTemperature
Dicharging shows the p-value, useful for rejecting the null hypothesis when compared to
Charging Cycle
0 the significance level (0.05 assumed).
Battery 5,000 negative Battery
temperature In general, greater positive and temperature correlations are observed for the batteries
< 21 < 21
-5,000
21 – 22 discharge cycle, in relation to the charge cycle. 21 – 22This is mainly due to the fact that the charg-
22 – 23 4,000 22 – 23
23 – 24 ing process occurs at a constant electric current 23 – 24 independent of the load demand or avail-
Battery Power (W)
Battery Power (W)

24 – 25 24 – 25
-10,000
> 25 able solar generation during the entire process, > 25 that is different from discharging cycle, in
3,000
which solar availability, as well as load demand, condition the power delivered by the
-15,000
system. For this reason, the PV-CEAMAZON 02 and 03 as well as the load demanded by
2,000
the building were also considered as they are variables that can influence the power de-
-20,000
livered by the batteries. Regular correlations 1,000
can be observed in relation to both photovol-
-25,000
taic systems in both charging and discharging cycle, as they are dependent on solar avail-
30 40
ability.
50 60 70 80 90
0
20 30 40 50 60 70 80 90 100
Regarding
Battery State of Charge (%)the discharge cycle (Figure 16), Battery it is State
observed that voltage not only de-
of Charge (%)
pends on delivered power and the SoC, but also on the delivered current, while it varies
(a)
inversely with the demanded load. Temperature also (b)increases as the loads power de-
Figure
Figure 15.
15. Temperature
Temperature variationmandin
variation increases,
in function
function of inbatteries
of this case
batteries with
soc
soc and negative
and power
power for: for:values,
(a) however,
(a) discharging
discharging lower
cycle
cycle and correlation
and (b)
(b) charging values are
charging cycle.
cycle.
observed in relation to the power delivered by the batteries, mainly due to photovoltaic
generation.
To
To better It can
better alsothe
assess
assess be relationship
observed how
relationship the delivered
between
between the currentassociated
the variables
variables and powerwith
associated fromthe
with thestorage
the batter-
storage
ies haveas
system,
system, lower
as wellcorrelation
well as
as their with photovoltaic
their influences,
influences, aa Spearman’s
Spearman’s generation, as it depends
correlation
correlation analysis on
analysis was
was loads demand.
developed
developed in
in
Minitab
Minitab In this way,
18 [43], the batteries
[43], between
between the operation
thestorage
storage systemis not
system only
voltage, influenced
voltage, current,
current, bytemperature,
the demanded
temperature, power load,state
power
and but
and
also
state by
of charge, theasavailable
of charge, wellasaswellsolar
theas generation,
the
demanded demandedloadwhich load contributes
power power
and theand to greater
the PV-CEAMAZON
PV-CEAMAZON autonomy
02 and 03 in off-grid
02systems.
and 03
operating
systems.
Figures 16 conditions.
Figures
and 1716 and 17the
present present the correlation
correlation matrices between matricesthe between
analyzed the variables.
analyzed varia-
bles.
Discharging Cycle Battery Voltage Battery Current Battery Temperature Battery SoC PV03 - (kW) PV02 - (kW) Loads Demand (kW)
Battery Current 0.802 Cell Content
1
0 Spearman Coefficient
p-value
Battery Temperature -0.832 -0.423 0.8
0 0 0.6
0.4
Battery SoC 0.882 0.499 -0.941 0.2
0 0 0
0
PV03 - (kW) -0.541 -0.136 0.623 -0.692
0 0.025 0 0
-0.2
PV02 - (kW) -0.528 -0.117 0.614 -0.681 0.981 -0.4
0 0.055 0 0 0 -0.6
-0.8
Loads Demand (kW) -0.884 -0.766 0.712 -0.774 0.636 0.628
0 0 0 0 0 0
-1
Battery Power (kW) 0.874 0.97 -0.541 0.624 -0.329 -0.311 -0.88
0 0 0 0 0 0 0
Figure
Figure16.
16.Spearman
Spearmancorrelation
correlationmatrix
atrix inindischarging
dischargingcycle
cycleofofstorage
storagesystem.
system.
The Spearman correlation coefficient assesses the monotonic relationship between

two variables using their rank values and not the values themselves. The development of
the Spearman’s rank correlation coefficient is presented in (1):
6 ∑in=1 Di2
rs = 1 − (1)
n ( n2 − 1)
where n is the number of pairs of values and Di = Xi − Yi is the difference between each
corresponding Xi and Yi value rank.
The correlation analysis results in a number between −1 and +1, called the correlation
coefficient. The higher the coefficient, the closer the relationship between the variables. The
analysis was performed for the storage system charging and discharging cycle.
Energies 2021, 14, 5284 A different scenario is observed for the batteries charging cycle (Figure 17), with15 low
of 21
correlation coefficients, as well as with high p-values, with which it is not possible to con-
clude on the relationship between some of the analyzed variables.
Charging Cycle Battery Voltage Battery Current Battery Temperature Battery SoC PV03 - (kW) PV02 - (kW) Loads Demand (kW)
Battery Current 0.269
1
0
Battery Temperature -0.028 -0.208 0.8

0.429 0 0.6
0.4
Battery SoC 0.996 0.245 -0.021 0.2
0 0 0.556
0
PV03 - (kW) 0.398 0.041 0.606 0.388
0 0.235 0 0
-0.2
PV02 - (kW) 0.381 0.038 0.605 0.373 0.992 -0.4
0 0.281 0 0 0 -0.6
-0.8
Loads Demand (kW) 0.22 -0.273 0.766 0.227 0.761 0.765
0 0 0 0 0 0
-1
Battery Power (kW) -0.002 0.658 -0.495 -0.01 -0.549 -0.551 -0.681
0.959 0 0 0.766 0 0 0
Figure 17. Spearman correlation matrix in charging cycle of storage system.

Figure 17. Spearman correlation matrix in charging cycle of storage system.
4.3. Third
In theScenario: Off-Grid
correlation Operation
matrix, with Diesel
the upper Generator
cell shows the Spearman coefficient while the
lowerFor
cellthe
shows the p-value,
last scenario, an useful forDG
existing rejecting the nulltohypothesis
was utilized when
analyze its compared
operation in theto hy-
the
significance
brid system to level (0.05 CEAMAZON
supply assumed). building energy demand in off-grid conditions. The
In general,
DG operation greater
will be onlypositive and conditions
in critical negative correlations
and as a last are observed
option, for the
aiming batteries
at the use of
discharge cycle, in relation to the charge cycle. This is mainly due to the
renewable energies, its operation will be in case of utility grid absence and a storage sys- fact that the
charging process occurs at a constant electric current independent of
tem SoC < 20%. For the test, a power grid interruption was conducted, and the DG was the load demand or
available solar
configured generation
to assume theduring the entire
load with a SoCprocess,
definedthat is different
in 86%. Underfrom discharging
these conditions, cycle,
the
in which load
building solarwas
availability,
increased asin
well as load
order demand,
to analyze thecondition
generatorthe power delivered
operation at full loadbycon-
the
system. (45
ditions ForkW).
this reason,
Figurethe18 PV-CEAMAZON
shows the events02sequence
and 03 asinwell
theasDG
theset
load demanded
operation, by the
divided
building
into four were also
events, considered
which will beas they are below.
described variables that can influence the power delivered
by the batteries. Regular correlations can be observed in relation to both photovoltaic
systems in both charging and discharging cycle, as they are dependent on solar availability.
Regarding the discharge cycle (Figure 16), it is observed that voltage not only depends
on delivered power and the SoC, but also on the delivered current, while it varies inversely
with the demanded load. Temperature also increases as the loads power demand increases,
in this case with negative values, however, lower correlation values are observed in relation
to the power delivered by the batteries, mainly due to photovoltaic generation. It can also
be observed how the delivered current and power from the batteries have lower correlation
with photovoltaic generation, as it depends on loads demand.
In this way, the batteries operation is not only influenced by the demanded load, but
also by the available solar generation, which contributes to greater autonomy in off-grid
operating conditions.
A different scenario is observed for the batteries charging cycle (Figure 17), with
low correlation coefficients, as well as with high p-values, with which it is not possible to
conclude on the relationship between some of the analyzed variables.
4.3. Third Scenario: Off-Grid Operation with Diesel Generator

For the last scenario, an existing DG was utilized to analyze its operation in the
hybrid system to supply CEAMAZON building energy demand in off-grid conditions.
The DG operation will be only in critical conditions and as a last option, aiming at the use
Figure 18. Diesel
of renewable generatorits
energies, operation in off-grid
operation will becondition.
in case of utility grid absence and a storage
system SoC < 20%. For the test, a power grid interruption was conducted, and the DG
was configured to assume the load with a SoC defined in 86%. Under these conditions,
the building load was increased in order to analyze the generator operation at full load
conditions (45 kW). Figure 18 shows the events sequence in the DG set operation, divided
into four events, which will be described below.
DG operation will be only in critical conditions and as a last option, aiming at the use of
renewable energies, its operation will be in case of utility grid absence and a storage sys-
tem SoC < 20%. For the test, a power grid interruption was conducted, and the DG was
configured to assume the load with a SoC defined in 86%. Under these conditions, the
Energies 2021, 14, 5284 16 of 21
building load was increased in order to analyze the generator operation at full load con-
ditions (45 kW). Figure 18 shows the events sequence in the DG set operation, divided
into four events, which will be described below.
Figure18.18.Diesel
Figure Diesel generator
generator operation
operation in off-grid
in off-grid condition.
condition.
It can be seen that the batteries (green curve) assume the building’s loads after the
power interruption. Considering the first event, once the defined 86% SoC is reached, the
DG set goes into operation assuming the building loads, however, within a few seconds,
the management system stops the DG operation and the batteries assume the building
loads, aiming at the energy supply continuity.
In the second event, the management system makes a new attempt to put the DG set
into operation, however, again without success. The observed difficulties on the diesel
generator to stay on and its sudden exit, is mainly due to the operation of the loss of main
(LOM) detection function of the management system, used to detect a loss of the electric
utility grid and avoid power islanding, ensuring safety and quality of energy supplied. The
manufacturer of the management system records a history of incompatibility with certain
diesel generators, due to the update of its firmware. As a solution, LOM function in the
management system was disabled, so that the generator was able to enter into operation
and reach its nominal power, as recorded in the third event. In the fourth event, the LOM
function was again deactivated, thus, DG makes a new attempt, with the same result as the
first two attempts, thus confirming the impact of the LOM function on the operation of the
generator used in the HMG.
Analyzing the DG operation period with the LOM function disabled, it was found
that both frequency and voltage decrease in the hybrid system as the power demanded
increases. Figure 19 details the magnitudes of voltage and frequency falls to values below
those allowed in the national regulation, based on PRODIST module 8 [44], which can
result in damage and malfunction of the installed loads. The HMG frequency registers
values below PRODIST standards at 20 kW < power (kW) < 22.8 kW. The voltage, on the
other hand, presents unsatisfactory values, at 37.9 kW < power (kW) < 36.7 kW. The limits
transgression can result in installed equipment malfunctioning in the building, so that it is
important to maintain the LOM function enabled during the operation of the DG set.
increases. Figure 19 details the magnitudes of voltage and frequency falls to values below
those allowed in the national regulation, based on PRODIST module 8 [44], which can
result in damage and malfunction of the installed loads. The HMG frequency registers
values below PRODIST standards at 20 kW < power (kW) < 22.8 kW. The voltage, on the
Energies 2021, 14, 5284 other hand, presents unsatisfactory values, at 37.9 kW < power (kW) < 36.7 kW. The limits
17 of 21
transgression can result in installed equipment malfunctioning in the building, so that it
is important to maintain the LOM function enabled during the operation of the DG set.
Figure
Figure19.
19.Power,
Power,voltage
voltageand
andfrequency
frequencycomparation
comparationof
ofdiesel
dieselgenerator
generatorset
setin
inoff-grid
off-gridconditions.
conditions.
5.5.Economic
EconomicEvaluation
Evaluationof ofthe
theHybrid
HybridMini-Grid
Mini-Grid
Froma afinancial
From financialperspective,
perspective, overover a month,
a month, it appears
it appears that that the operation
the joint joint operationof pho- of
photovoltaic systems, storage and external grid, highlighted in Section
tovoltaic systems, storage and external grid, highlighted in Section 4.1 using the on-grid 4.1 using the on-
gridoff-grid
and and off-grid programmed
programmed operation,
operation, brings
brings financial
financial benefits
benefits to theto consumer.
the consumer. TheTheuseuse
of
of photovoltaic
photovoltaic generation
generation at the
at the timetime
of of highest
highest demandfor
demand forenergy
energyininthe theCEAMAZON
CEAMAZON
buildingand
building andthe
thestorage
storageofof energy
energy surplus
surplus in the
in the set set of batteries,
of batteries, make make it possible
it possible to re-
to reduce
duce
the the contracted
contracted demand demand
during during
times oftimes
highestof demand,
highest demand,
reducingreducing
the energy the energy con-
consumption
sumption
from from the
the external external
grid for thegrid for thepeak
building building
timespeak times demands,
demands, exactly when exactlythe when
utility´sthe
utility´is
energy s energy is more expensive.
more expensive. In addition,
In addition, the energy thesurplus
energygenerated
surplus generated by the pho-
by the photovoltaic
systems
tovoltaiccan be delivered
systems to the electric
can be delivered to theutility
electric forutility
consumption during times
for consumption during of times
higherof
consumption or lower photovoltaic generation at other times of the year,
higher consumption or lower photovoltaic generation at other times of the year, in accord- in accordance
with
anceBrazilian legislation.
with Brazilian legislation.
In
Inorder
ordertotofully
fullymonetize
monetizethe theHMC
HMCconsumption
consumptionand andgeneration,
generation,ititwas wasconsidered
considered
that
thatthe
theenergy
energysurplus
surpluswill
willbebesold
soldtotothe
thelocal
localutility
utilityfor
forthe
thesame
sametariff
tariffas
asititisispurchased
purchased
during
duringoff-peak
off-peak hours. Figure 20
hours. Figure 20shows
showsaacomparison
comparisonofofthe theCEAMAZON
CEAMAZON energy
energy billbill
for
for the period of one month, with and without
the period of one month, with and without using the HMG. using the HMG. In the month in question,
month in question,
CEAMAZON
CEAMAZONsaved savedUSDUSD655.50.
655.50.
Figure20.
Figure 20.Estimation
Estimationof
ofthe
theelectricity
electricitybill
billwith
withand
andwithout
withoutthe
themini-grid
mini-gridparticipation.
participation.
In Figure 21, an energy balance was made considering the complete system, using
the three photovoltaic systems, the storage system, the loads in the CEAMAZON building
and the utility grid energy. Together, the three photovoltaic systems indicate that the
CEAMAZON mini-plant (PV-CEAMAZON 01, 02 and 03) has a larger generation than
Energies 2021, 14, 5284 18 of 21
Figure 20. Estimation of the electricity bill with and without the mini-grid participation.
In Figure 21, an energy balance was made considering the complete system, using
In Figure
the three 21, an energy
photovoltaic systems,balance was made
the storage system,considering
the loads inthe
thecomplete
CEAMAZON system, using the
building
three
and thephotovoltaic
utility gridsystems,
energy.the storage the
Together, system,
threethe loads in thesystems
photovoltaic CEAMAZON
indicatebuilding
that the and
the utility gridmini-plant
CEAMAZON energy. Together, the three photovoltaic
(PV-CEAMAZON 01, 02 and 03)systems
has a indicate that the CEAMA-
larger generation than
ZON mini-plant
demanded. In this(PV-CEAMAZON 01, 02 and
case, the surplus energy 03) has a in
is distributed larger
UFPA’sgeneration
internal than demanded.
grid to com-
In this case,
pensate, the surplus
energetically energy is distributed
and financially, in UFPA’s internal
the energy consumption of other grid to compensate,
university build-
ings.
energetically and financially, the energy consumption of other university buildings.
Figure 21. Total power balance for mini power plant operation.
Using the CEAMAZON HMG, it is noted that, mainly at the weekend, a large amount
of energy is delivered to the university’s internal grid. According to the energy balance for
the analyzed period, a projection of the grid behavior over a typical month was developed,
the surplus of energy distributed in the campus would allow a reduction of 1.89% in
contracted demand and 4.06% in energy consumption for the entire university. Despite the
reduction appears to be small, given the magnitude of the UFPA loads, the results show
the economic and energy benefits that HMG can bring for large-scale uses.
Based on the results presented in this section, it is possible to conclude on the economic
benefits that the implementation of the hybrid systems brings to CEAMAZON and also to
UFPA. The implementation of renewable energy generation and storage systems allow the
reduction of contracted demand and energy consumption during peak hours, significantly
reducing energy bill costs. In addition, analyzing the electricity grid from the point of
view of distribution and generation utilities, when managing the energy mainly of large
consumers, it is possible to reduce immediate costs with the construction/expansion
of electricity grids, reduce energy theft, reduce energy demand at peak times, among
other benefits.
6. Conclusions
This paper presented an evaluation of a real hybrid system operation which is installed
in the Amazon region of Brazil, composed of a utility grid, photovoltaic generation systems,
energy storage system and DG, under both on-grid and off-grid operating conditions. To
assess its operation in typical conditions of urban areas as well as in remote areas of the
region, three operating scenarios were carried out. Some conclusions can be made from
the results:
• Under normal operating conditions (on-grid conditions) the batteries in the HMG
offer operational benefits in terms of service continuity and reliability in the face of
unexpected power failures, mainly in scenarios with little photovoltaic generation as
Energies 2021, 14, 5284 19 of 21
well as the variability associated with this type of source. Also, with the programmed
discharge operation in on-grid mode, the HMG is able to reduce the dependency on
the utility grid energy during the time period of highest consumer demand and, also,
during peak hours, when there is a higher energy shortage.
• In off-grid operating conditions, the storage system is responsible for maintaining
constant voltages, frequency and power through the centralized system controller that
performs the available load-generation balance, which allows higher autonomy for the
battery. It was observed how there is no difference between the on-grid conditions and
off-grid conditions in power quality terms for the storage system discharging cycle.
This makes its use promising for applications where high reliability and continuity of
service are required, such as hospitals.
• Finally, the decentralization of energy supply also generates a new business model for
service companies and energy utilities that will be able to sell energy, monitoring and
management services, as well as system maintenance, among others. With this, the
analyzed balance between power demand and generation in HMG can be replicated
in homes, hotels, condominiums, universities and even in cities that are connected to
the electricity grid. For isolated communities, hybrid systems offer a more sustainable
and good quality electrical system to meet the community’s energy demand.
Author Contributions: J.M.T. and U.C.P.J. designed the entire study. J.M.T., U.C.P.J. and C.E.M.R.
conducted the data acquisition, modeling and analysis of the results. J.M.T., U.C.P.J., B.S.d.A. and
C.E.M.R. wrote the original draft and edited the text. U.H.B., M.E.d.L.T., E.O.d.M. and A.A.d.N.
reviewed and supervised the study. A.A.d.N. guaranteed financial resources. All authors discussed
the results and the conclusions of the manuscript. All authors have read and agreed to the published
version of the manuscript.
Funding: This research was funded by Norte Energia S.A. and Pro-Rectory of Research and Post-
Graduate Studies-PROPESP/UFPA.
Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.
Data Availability Statement: Not applicable.
Acknowledgments: The article’s authors are grateful for the support of the Coordination for the
Improvement of Higher Education Personnel—Brazil (CAPES); Pro-Rectory of Research and Post-
Graduate Studies-PROPESP/UFPA and Norte Energia S.A. to carry out the present work, within
the scope of the ANEEL R&D project 07427-0319/2019 entitled “Intelligent system for efficient
management of multimodal electric mobility”, carried out within the scope of Call 22/2018—ANEEL.
Conflicts of Interest: The authors declare no conflict of interest.
References
1. Energy Research Company. National Energy Balance 2020—Base Year 2019; Energy Research Company: Rio de Janeiro, Brazil, 2020.
2. Pereira, E.B.; Martins, F.R.; Gonçalves, A.R.; Costa, R.S.; de Lima, F.J.L.; Rüther, R.; de Abreu, S.L.; Tiepolo, G.M.; Pereira, S.V.; de
Souza, J.G. Brazilian Solar Energy Atlas, 2nd ed.; INPE: São José dos Campos, Brazil, 2017. [CrossRef]
3. Irradiation Database in Brazil. Renewable Energy Resources Modeling and Studies Laboratory. Available online: http://labren.
ccst.inpe.br/atlas_2017.html (accessed on 16 July 2021).
4. Irradiation Database in Brazil. Reference Center for Solar and Wind Energy Sérgio Brito. Available online: http://www.cresesb.
cepel.br/ (accessed on 16 July 2021).
5. National Electric System Operator. History of Solar Generation in BRAZIL. Available online: http://www.ons.org.br/Paginas/
resultados-da-operacao/historico-da-operacao/geracao_energia.aspx (accessed on 20 April 2021).
6. Nitta, N.; Wu, F.; Lee, J.T.; Yushin, G. Li-ion battery materials: Present and future. Mater. Today 2015, 18, 252–264. [CrossRef]
7. Operational Energy Storage Capacity Globally by Country 2017. Statista. Available online: https://www.statista.com/statistics/
785692/global-energy-storage-capacity-by-country/ (accessed on 26 March 2021).
8. Miao, Y.; Hynan, P.; von Jouanne, A.; Yokochi, A. Current Li-Ion Battery Technologies in Electric Vehicles and Opportunities for
Advancements. Energies 2019, 12, 1074. [CrossRef]
9. Guzmán, G.; Vazquez-Arenas, J.; Ramos-Sánchez, G.; Bautista-Ramírez, M.; González, I. Improved performance of LiFePO4
cathode for Li-ion batteries through percolation studies. Electrochim. Acta 2017, 247, 451–459. [CrossRef]
Energies 2021, 14, 5284 20 of 21
10. Types of Lithium-Ion Batteries—Battery University. Available online: https://batteryuniversity.com/learn/article/types_of_

lithium_ion (accessed on 21 March 2021).
11. National Electric System Operator. Isolated System. Available online: http://www.ons.org.br/paginas/sobre-o-sin/sistemas-
isolados (accessed on 29 March 2021).
12. Pinho, J.T.; Barbosa, C.F.O.; Pereira, E.J.; Souza, H.M.S.; Blasques, L.C.M.; Galhardo, M.A.B.; Macêdo, W.N. Hybrid Systems—Energy
Solutions for the Amazon, 1st ed.; Ministry of Mines and Energy: Brasilia, Brazil, 2008.
13. Costa, T.S.; Villalva, M.G. Technical Evaluation of a PV-Diesel Hybrid System with Energy Storage: Case Study in the Tapajós-
Arapiuns Extractive Reserve, Amazon, Brazil. Energies 2020, 13, 2969. [CrossRef]
14. Madziga, M.; Rahil, A.; Mansoor, R. Comparison between Three Off-Grid Hybrid Systems (Solar Photovoltaic, Diesel Generator
and Battery Storage System) for Electrification for Gwakwani Village, South Africa. Environments 2018, 5, 57. [CrossRef]
15. de Freitas Moscardini Júnior, E.; Rüther, R. The influence of the solar radiation database and the photovoltaic simulator on the
sizing and economics of photovoltaic-diesel generators. Energy Convers. Manag. 2020, 210, 112737. [CrossRef]
16. Hauschild, L. Evaluation of Operation Strategies for Hybrid Photovoltaic-Wind-Diesel Systems. In Proceedings of the 2017 IEEE
International Conference on Environment and Electrical Engineering and 2017 IEEE Industrial and Commercial Power Systems
Europe, Milan, Italy, 6–9 June 2017; p. 117.
17. Sawle, Y.; Gupta, S.C.; Bohre, A.K. PV-wind hybrid system: A review with case study. Cogent Eng. 2016, 3, 1189305. [CrossRef]
18. Henrique, A.; Christian Heinrich, M. Energy Balance in a Renewable Energy Community. In Proceedings of the 10th International
Conference on Environment and Electrical Engineering, Rome, Italy, 8–11 May 2011. [CrossRef]
19. Henrique, A.; Christian Heinrich, M. Performance Comparison of the Energy Balance in a Renewable Energy Community. In
Proceedings of the 11th International Conference on Environment and Electrical Engineering, Venice, Italy, 18–25 May 2012.
[CrossRef]
20. Empowering Consumers through Energy Communities—2016. Available online: https://eusew.eu/empowering-consumers-
through-energy-communities (accessed on 16 July 2021).
21. Prospective Study of the Steel Industry (2025): Energy Efficiency in the Steel Industry; Technical Note; Center for Strategic Studies and
Management: Brasilia, Brazil, 2009.
22. de Araújo, E.Q. Energy Balance Model for Managing and Forecasting Demand for a Steel Plant; Postgraduate Program in Energy
Engineering at the Federal Center for Technological Education of Minas Gerais: São João Del Rei, Brazil, 2015.
23. Viviescas, C.; Lima, L.; Diuana, F.A.; Vasquez, E.; Ludovique, C.; Silva, G.N.; Huback, V.; Magalar, L.; Szklo, A.; Lucena, A.F.P.;
et al. Contribution of Variable Renewable Energy to increase energy security in Latin America: Complementarity and climate
change impacts on wind and solar resources. Renew. Sustain. Energy Rev. 2019, 113, 109232. [CrossRef]
24. de Lima, L.C.; de Ferreira, L.; de Morais, F.H.B. Performance analysis of a grid connected photovoltaic system in northeastern
Brazil. Renew. Sustain. Energy Rev. 2017, 37, 79–85. [CrossRef]
25. de Santos, J.A.F.; de Jong, P.; Costa, C.A.; Torres, E.A. Combining wind and solar energy sources: Potential for hybrid power
generation in Brazil. Renew. Sustain. Energy Rev. 2020, 67, 101084. [CrossRef]
26. Prüggler, N.; Prüggler, W.; Wirl, F. Storage and Demand Side Management as power generator’s strategic instruments to influence
demand and prices. Renew. Sustain. Energy Rev. 2011, 36, 6308–6317. [CrossRef]
27. Yao, A.W.L.; Chi, S.C. Analysis and design of a Taguchi–Grey based electricity demand predictor for energy management systems.
Energy Convers. Manag. 2004, 45, 1205–1217. [CrossRef]
28. Paixão, U.C., Jr.; Filho, C.L.P.; de Souza, A.C.D.B.; de Tostes, M.E. Insertion of smart meters in the Brazilian electrical system. In
Proceedings of the XXIII National Seminar on Electric Energy Distribution, Fortaleza, Brazil, 20–23 November 2018.
29. de Carvalho, I.S.; Paixão, U.C., Jr.; de Souza, A.C.D.B.; da Muniz, J.R.; Monteiro, F.P.; de Tostes, M.E. Modeling and validation of
a multiplatform system for Electric Energy Management. In Proceedings of the 13th Latin-American Congress on Electricity
Generation and Transmission (CLAGTEE 2019), Santiago, Chile, 20–23 October 2019.
30. de Carvalho, I.S.; Paixão, U.C., Jr.; de Souza, A.C.D.B.; da Muniz, J.R.; Monteiro, F.P.; de Tostes, M.E. Modeling and validation of
a multiplatform system for Electric Energy Management—SISGEE. In Proceedings of the Brazilian Symposium on Intelligent
Automation (SBAI), Ouro Preto, Brazil, 27 October 2019.
31. da Muniz, J.R. Computational Project of the Electric Energy Management System (SISGEE) in the University City José da Silveira Netto;
Federal University of Pará: Belém, Brazil, 2019.
32. Sial, A.; Singh, A.; Mahanti, A. Detecting Anomalous Energy Consumption Using Contextual Analysis of Smart Meter Data; Springer
Science+Business Media: Berlin, Germany, 2019. [CrossRef]
33. Sial, A.; Singh, A.; Mahanti, A.; Gong, M. Heuristics-Based Detection of Abnormal Energy Consumption. In International
Conference on Smart Grid Inspired Future Technologies; Springer: Cham, Switzerland, 2018. [CrossRef]
34. Jiang, Y.; Wan, C.; Chen, C.; Shahidehpour, M.; Song, Y. A Hybrid Stochastic-Interval Operation Strategy for Multi-Energy
Microgrids. IEEE Trans. Smart Grid 2020, 11, 440–456. [CrossRef]
35. Pilz, M.; Al-Fagih, L. Selfish Energy Sharing in Prosumer Communities: A Demand-Side Management Concept. In Proceedings of
the IEEE International Conference on Communications, Control, and Computing Technologies for Smart Grids (SmartGridComm),
Beijing, China, 21–23 October 2019.
36. Zhou, S.; Hu, Z.; Gu, W.; Jiang, M.; Zhang, X.P. Artificial Intelligence Based Smart Energy Community Management: A
Reinforcement Learning Approach. CSEE J. Power Energy Syst. 2019, 5, 1–10. [CrossRef]
Energies 2021, 14, 5284 21 of 21
37. Paixão, U.C., Jr.; Andrade, V.B.; Tabora, J.M.; do Nascimento, A.A.; de Tostes, M.E.; Bezerra, U.H.; de Matos, E.O.; de Albuquerque,
B.S.; Rodrigues, C.E.M.; Takeda, F.M.; et al. Modeling and simulation of operability scenarios of a hybrid mini-grid with
photovoltaic-diesel generation, energy storage connected to the grid. In Proceedings of the XII Brazilian Congress on Energy
Planning, Foz do Iguaçu, Brazil, 23 September 2020.
38. Paixão, U.C., Jr.; de Albuquerque, L.O.; de Barbosa, C.F.; Veríssimo, J.P.A.; Pinho, J.T.; Galhardo, M.A.B. Quality of electricity
supply in the GEDAE/UFPA hybrid systems and mini-grids laboratory. Braz. J. Sol. Energy 2019, 10, 38–48.
39. ABNT Catalog. Available online: https://www.abntcatalogo.com.br/norma.aspx?ID=090238 (accessed on 29 March 2021).
40. General Electricity Supply Conditions—Normative Resolution 414/2010—ANEEL. Available online: https://aneel.gov.br/ren-41
4 (accessed on 7 April 2021).
41. Roscher, M.A.; Bohlen, O.; Vetter, J. OCV Hysteresis in Li-Ion Batteries including Two-Phase Transition Materials. Int. J.
Electrochem. 2011, 2011, 984320. [CrossRef]
42. A Common Definition for Zero Energy Buildings. EnergyGov. Available online: https://www.energy.gov/eere/buildings/
downloads/common-definition-zero-energy-buildings (accessed on 25 March 2021).
43. Minitab 18 Statistical Software. Computer Software; Minitab, Inc.: State College, PA, USA, 2010; Available online: https://www.
minitab.com/es-mx/ (accessed on 15 August 2019).
44. PRODIST—Module 8—Electricity Power Quality—ANEEL. Available online: https://www.aneel.gov.br/modulo-8 (accessed on
29 March 2021).

Energies - Hybrid Systems

Uploaded by

Copyright:

Available Formats

You might also like

Energies - Hybrid Systems

Uploaded by

Document Information

Original Title

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

Energies - Hybrid Systems

Uploaded by

Copyright:

Available Formats

Received: 28 June 2021

Energies 2021, 14, 5284. https://doi.org/10.3390/en14175284 https://www.mdpi.com/journal/energies

Energies 2021, 14, 5284 2 of 21

Table 1. Review of literature regarding hybrid mini-grid and energy management.

Papers’ Main Subject Relevant Literature

Figure 4. Hybrid system components at CEAMAZON.

Figure 5. Hybrid system

2.3. Diesel Generation

2.4. Storage System

3. Methodology for Hybrid Mini-Grid Operation Parameters Configuration

Figure 6. Hybrid system operation configurations flowchart.

4. Technical Evaluation of the Hybrid Mini-Grid

Table 2. Operation scenarios for the hybrid system sources.

Source Scenario 1 Scenario 2 Scenario 3

4.1. First Scenario: Normal Operation

4.1.2. Scheduled Operation of the Hybrid Mini-Grid to Demand Control

Energies 2021, 14, x FOR PEER REVIEW 12 of 21

Figure 11. Storage system operation cycle in off-grid conditions.

Storage System Assessment in Charge/Discharge

Figure 13. Storage system voltage discharging/charging cycle.

Figure 13. Storage system voltage discharging/charging cycle.

Battery Power (W)

-20,000 points), aiming to maintain the battery in fully charged state (

-25,000 By making a correlation between the voltage, the charge

Battery Power (W)

The Spearman correlation coefficient assesses the monotonic relationship between

Battery Temperature -0.028 -0.208 0.8

Figure 17. Spearman correlation matrix in charging cycle of storage system.

4.3. Third Scenario: Off-Grid Operation with Diesel Generator

10. Types of Lithium-Ion Batteries—Battery University. Available online: https://batteryuniversity.com/learn/article/types_of_

You might also like