Revista Facultad de Ingeniería, Universidad de Antioquia, No.110, pp. 9-22, Jan-Mar 2024
Integration of LFP-second life batteries as a
storage in a smart microgrid
Integración de baterías LFP-segunda vida como almacenamiento en una microrred inteligente
Oscar Izquierdo-Monge 1*, Nicolás Alonso González 1, Paula Peña-Carro 1
1Centro de Investigaciones Energéticas Medioambientales y Tecnológicas CIEMAT, Departamento de Energía. Autovía de
Navarra A15 Salida 56 42290 Lubia. C. P. 28040. Soria, España.
CITE THIS ARTICLE AS:
O. Izquierdo-Monge, N. Alonso
González and P. Peña-Carro.
Numerical analysis of soil
desaturation by an air injection
method, Revista Facultad de
Ingeniería Universidad de
Antioquia, no. 110, pp. 9-22,
Jan-Mar 2024. [Online].
Available: https:
//www.doi.org/10.17533/
udea.redin.20230211
ARTICLE INFO:
Received: March 08, 2022
Accepted: February 06, 2023
Available online: February 06,
2023
KEYWORDS:
Smart microgrid; LFP
batteries; Second life; Energy
storage system; energy
consumption
Microrred inteligente; baterías
LFP; segunda vida; sistema de
almacenamiento de energía;
consumo de energía
ABSTRACT: In recent years, there has been an increasing commitment to give batteries a
second life, as they are being consumed for different uses and the recycling methods are
not defined. This work aims to show how a storage system based on disused Lithium
Iron Phosphate (LFP) batteries has been recovered and integrated into the CE.D.E.R-
CIEMAT smart microgrid over a period of ten years during which the operation of the
system has been affected. During the recovery process, the cells have been classified
according to their voltage, and a series of charge-discharge processes have been carried
out on them at different voltages to determine their state of health and capacity. Once
characterised, the system was assembled and commissioned with the appropriate cells.
In addition, for the storage system, a Supervisory Control And Data Acquisition (SCADA)
has been developed in Home Assistant for its integration into the CE.D.E.R.’s microgrid
management system. This allows the microgrid to be managed more efficiently, storing
surplus energy from distributed generation sources and discharging the stored energy
during peak consumption periods to reduce peaks, reduce discharges to the distribution
grid and reduce the cost of electricity bills.
RESUMEN: Ante el aumento del consumo y producción de baterías para diferentes usos
en los últimos años y la problemática actual en la definición de los métodos de reciclaje,
se apuesta cada vez más por dar una segunda vida a las baterías. El propósito de este
trabajo es mostrar cómo se ha realizado la recuperación e integración en la microrred
inteligente del CE.D.E.R.-CIEMAT de un sistema de almacenamiento basado en baterías
Litio Ferro-Fosfato (LFP) en desuso durante diez años en los que se ha visto afectada la
operación del sistema. Durante el proceso de recuperación, se han clasificado las celdas
en base a su tensión y se les ha realizado una serie de procesos de carga-descarga
a diferentes tensiones para determinar el estado de salud y capacidad. Una vez
caracterizadas, se ha procedido al montaje y puesta en marcha del sistema con las
celdas adecuadas. Además, se ha desarrollado un Control Supervisor y Adquisición de
Datos (SCADA) del sistema de almacenamiento en Home Assistant para su integración
en el sistema de gestión de la microrred del CE.D.E.R.. Esto permite gestionar la
microrred de forma más eficiente, almacenando los excedentes energéticos de las
fuentes de generación distribuida y vertiendo la energía almacenada en periodos de
máximo consumo con el objetivo de reducir los picos, reducir los vertidos a la red de
distribución y disminuir el coste en la factura de la luz.
1. Introduction
In the last 30 years, society has faced many problems
related to the development of energy systems, the
exhaustion of sources all around the world, economic
recessions, climate change and CO2 emissions [1]. To
make society capable of confronting these problems,
experts around the world have been working on
investigations and developing new alternatives related to
renewable energy sources and innovative energy storage
9
* Corresponding author: Oscar Izquierdo-Monge
E-mail: oscar.izquierdo@ciemat.es
ISSN 0120-6230
e-ISSN 2422-2844
DOI: 10.17533/udea.redin.20230211 9
Integration of LFP-second life batteries as a
storage in a smart microgrid
Integración de baterías LFP-segunda vida como almacenamiento en una microrred inteligente
Oscar Izquierdo-Monge 1*, Nicolás Alonso González 1, Paula Peña-Carro 1
1Centro de Investigaciones Energéticas Medioambientales y Tecnológicas CIEMAT, Departamento de Energía. Autovía de
Navarra A15 Salida 56 42290 Lubia. C. P. 28040. Soria, España.
CITE THIS ARTICLE AS:
O. Izquierdo-Monge, N. Alonso
González and P. Peña-Carro.
Numerical analysis of soil
desaturation by an air injection
method, Revista Facultad de
Ingeniería Universidad de
Antioquia, no. 110, pp. 9-22,
Jan-Mar 2024. [Online].
Available: https:
//www.doi.org/10.17533/
udea.redin.20230211
ARTICLE INFO:
Received: March 08, 2022
Accepted: February 06, 2023
Available online: February 06,
2023
KEYWORDS:
Smart microgrid; LFP
batteries; Second life; Energy
storage system; energy
consumption
Microrred inteligente; baterías
LFP; segunda vida; sistema de
almacenamiento de energía;
consumo de energía
ABSTRACT: In recent years, there has been an increasing commitment to give batteries a
second life, as they are being consumed for different uses and the recycling methods are
not defined. This work aims to show how a storage system based on disused Lithium
Iron Phosphate (LFP) batteries has been recovered and integrated into the CE.D.E.R-
CIEMAT smart microgrid over a period of ten years during which the operation of the
system has been affected. During the recovery process, the cells have been classified
according to their voltage, and a series of charge-discharge processes have been carried
out on them at different voltages to determine their state of health and capacity. Once
characterised, the system was assembled and commissioned with the appropriate cells.
In addition, for the storage system, a Supervisory Control And Data Acquisition (SCADA)
has been developed in Home Assistant for its integration into the CE.D.E.R.’s microgrid
management system. This allows the microgrid to be managed more efficiently, storing
surplus energy from distributed generation sources and discharging the stored energy
during peak consumption periods to reduce peaks, reduce discharges to the distribution
grid and reduce the cost of electricity bills.
RESUMEN: Ante el aumento del consumo y producción de baterías para diferentes usos
en los últimos años y la problemática actual en la definición de los métodos de reciclaje,
se apuesta cada vez más por dar una segunda vida a las baterías. El propósito de este
trabajo es mostrar cómo se ha realizado la recuperación e integración en la microrred
inteligente del CE.D.E.R.-CIEMAT de un sistema de almacenamiento basado en baterías
Litio Ferro-Fosfato (LFP) en desuso durante diez años en los que se ha visto afectada la
operación del sistema. Durante el proceso de recuperación, se han clasificado las celdas
en base a su tensión y se les ha realizado una serie de procesos de carga-descarga
a diferentes tensiones para determinar el estado de salud y capacidad. Una vez
caracterizadas, se ha procedido al montaje y puesta en marcha del sistema con las
celdas adecuadas. Además, se ha desarrollado un Control Supervisor y Adquisición de
Datos (SCADA) del sistema de almacenamiento en Home Assistant para su integración
en el sistema de gestión de la microrred del CE.D.E.R.. Esto permite gestionar la
microrred de forma más eficiente, almacenando los excedentes energéticos de las
fuentes de generación distribuida y vertiendo la energía almacenada en periodos de
máximo consumo con el objetivo de reducir los picos, reducir los vertidos a la red de
distribución y disminuir el coste en la factura de la luz.
1. Introduction
In the last 30 years, society has faced many problems
related to the development of energy systems, the
exhaustion of sources all around the world, economic
recessions, climate change and CO2 emissions [1]. To
make society capable of confronting these problems,
experts around the world have been working on
investigations and developing new alternatives related to
renewable energy sources and innovative energy storage
9
* Corresponding author: Oscar Izquierdo-Monge
E-mail: oscar.izquierdo@ciemat.es
ISSN 0120-6230
e-ISSN 2422-2844
DOI: 10.17533/udea.redin.20230211 9
O. Izquierdo-Monge et al., Revista Facultad de Ingeniería, Universidad de Antioquia, No. 110, pp. 9-22, 2024
systems.
Thus, alternatives to conventional energy generation
and storage methods have become one of the priorities of
governments in most countries around the world. One of
the essential aims of these alternatives is to break up the
harmful effect of fossil fuels throughout their life, since
they are produced until their final uses [2–4].
One of the most important developing fields is the
automotive sector, in particular Electric Vehicles (EV).
The main reason why this sector has increased its offer
in different car brands and their sales is the reduction of
fossil fuel consumption and gas emissions, both at the
European and global levels. The International Energy
Agency (IEA) estimates that there are currently around 16
million EVs in the world, showing an increase that triples
the registered sales data concerning the year 2021 in the
People’s Republic of China, and 70% in Europe [5].
This huge increase in sales of hybrid, plug-in hybrid,
or pure EVs poses a challenge in battery management,
both in production, and, even more so, in the management
of batteries when they no longer meet the minimum
characteristics required for automotive use and need to
be replaced. This situation raises the question: what
will we do with such a large number of batteries in a few
years? To answer this question, studies on the useful
life of batteries in vehicles have been initiated. The great
majority of these studies, some of them such as [6–8],
agree that after a period of approximately eight years of
continued vehicle use, it would be appropriate to replace
the vehicle’s battery system because it no longer provides
the energy requirements necessary to supply the vehicle’s
needs.
With these results, we see that batteries have suffered
degradation in their primary use energy capacities and
are not suitable for operation in this field, but are they
suitable for use in other applications, or do they need to
be discarded and taken to a recycling point? Different
authors, such as [9–13], agree that batteries still have
suitable capacities for use in other types of applications
with a categorisation that they call ”second-life battery”.
The most prominent applications, which would be
carried out with optimum performance, for this are area
regulation and energy storage systems to support grid
services, microgrids, or renewable energies (RES). All
these systems are characterised by being stationary and
needing smaller amounts of energy over longer periods of
time than in their first life uses, where the need for energy
implies more powerful systems with higher volumes of
energy over shorter periods of time [5].
An important factor in the battery industry is the cost
of batteries, and this is also a major factor in the field
of second-life batteries, as the cost of a user device
with certain diminished characteristics is also positively
affected by the fact that the cost is discounted from
the initial price. The authors of a comprehensive study
on second-life batteries [8] suggest, after a critique of
numerous papers, a cost of second-life batteries ranging
from $44 to $180/kWh, which is more affordable than
purchasing new batteries for these proposed secondary
applications.
So far, this document has focused on the second life
of batteries from EVs, as this is the largest market niche in
which we can enter batteries with sizes and percentages
of residual capacity suitable for these second-life
applications mentioned above. We must also consider
other types of the origin or first use of these batteries, like
a storage battery system already installed and that, due
to its use or disuse, the system no longer performs as it
originally did, and its working capacity has been affected.
This is our case, the Renewable Energy Development
Centre (CE.D.E.R.) - Centre for Energy, Environmental
and Technological Research (CIEMAT) located in Lubia
(Soria, Spain). It is a public research organisation. This
centre is part of the CIEMAT centre and is attached to the
Department of Energy of the Spanish Government. The
CE.D.E.R. specialises in the development and promotion
of renewable energies associated with research projects
in this field. For the development of all this work, it has
different types of energy generation and storage systems.
One of the storage systems affected by this problem
available at the centre is an LFP battery system. These
LFP batteries consist of two racks of 627.3 V in total. For
various reasons, the system has been in disuse for several
years, which has had a notably negative influence on the
system, greatly compromising its operation and, therefore,
its performance.
Whether we are dealing with a battery system for
primary vehicle or storage use, to assess its validity for a
second use, it is necessary to examine the cells that make
up the system and their capacities [8, 14].
Those batteries are known as “retired” and still maintain
nearly 80% of their initial capacity after their first life uses
[15]. Despite this, they are considered “out of service” as
they cannot fully satisfy the demand established for the
systems they were initially designed for. This aspect leads
to the application of the concept of “second life”, “reuse”
or “recycle” into other different fields with demands of
fewer current values involved or uninterruptable energy
storage systems [16].
Concretely, LFP batteries have a useful period between
10
systems.
Thus, alternatives to conventional energy generation
and storage methods have become one of the priorities of
governments in most countries around the world. One of
the essential aims of these alternatives is to break up the
harmful effect of fossil fuels throughout their life, since
they are produced until their final uses [2–4].
One of the most important developing fields is the
automotive sector, in particular Electric Vehicles (EV).
The main reason why this sector has increased its offer
in different car brands and their sales is the reduction of
fossil fuel consumption and gas emissions, both at the
European and global levels. The International Energy
Agency (IEA) estimates that there are currently around 16
million EVs in the world, showing an increase that triples
the registered sales data concerning the year 2021 in the
People’s Republic of China, and 70% in Europe [5].
This huge increase in sales of hybrid, plug-in hybrid,
or pure EVs poses a challenge in battery management,
both in production, and, even more so, in the management
of batteries when they no longer meet the minimum
characteristics required for automotive use and need to
be replaced. This situation raises the question: what
will we do with such a large number of batteries in a few
years? To answer this question, studies on the useful
life of batteries in vehicles have been initiated. The great
majority of these studies, some of them such as [6–8],
agree that after a period of approximately eight years of
continued vehicle use, it would be appropriate to replace
the vehicle’s battery system because it no longer provides
the energy requirements necessary to supply the vehicle’s
needs.
With these results, we see that batteries have suffered
degradation in their primary use energy capacities and
are not suitable for operation in this field, but are they
suitable for use in other applications, or do they need to
be discarded and taken to a recycling point? Different
authors, such as [9–13], agree that batteries still have
suitable capacities for use in other types of applications
with a categorisation that they call ”second-life battery”.
The most prominent applications, which would be
carried out with optimum performance, for this are area
regulation and energy storage systems to support grid
services, microgrids, or renewable energies (RES). All
these systems are characterised by being stationary and
needing smaller amounts of energy over longer periods of
time than in their first life uses, where the need for energy
implies more powerful systems with higher volumes of
energy over shorter periods of time [5].
An important factor in the battery industry is the cost
of batteries, and this is also a major factor in the field
of second-life batteries, as the cost of a user device
with certain diminished characteristics is also positively
affected by the fact that the cost is discounted from
the initial price. The authors of a comprehensive study
on second-life batteries [8] suggest, after a critique of
numerous papers, a cost of second-life batteries ranging
from $44 to $180/kWh, which is more affordable than
purchasing new batteries for these proposed secondary
applications.
So far, this document has focused on the second life
of batteries from EVs, as this is the largest market niche in
which we can enter batteries with sizes and percentages
of residual capacity suitable for these second-life
applications mentioned above. We must also consider
other types of the origin or first use of these batteries, like
a storage battery system already installed and that, due
to its use or disuse, the system no longer performs as it
originally did, and its working capacity has been affected.
This is our case, the Renewable Energy Development
Centre (CE.D.E.R.) - Centre for Energy, Environmental
and Technological Research (CIEMAT) located in Lubia
(Soria, Spain). It is a public research organisation. This
centre is part of the CIEMAT centre and is attached to the
Department of Energy of the Spanish Government. The
CE.D.E.R. specialises in the development and promotion
of renewable energies associated with research projects
in this field. For the development of all this work, it has
different types of energy generation and storage systems.
One of the storage systems affected by this problem
available at the centre is an LFP battery system. These
LFP batteries consist of two racks of 627.3 V in total. For
various reasons, the system has been in disuse for several
years, which has had a notably negative influence on the
system, greatly compromising its operation and, therefore,
its performance.
Whether we are dealing with a battery system for
primary vehicle or storage use, to assess its validity for a
second use, it is necessary to examine the cells that make
up the system and their capacities [8, 14].
Those batteries are known as “retired” and still maintain
nearly 80% of their initial capacity after their first life uses
[15]. Despite this, they are considered “out of service” as
they cannot fully satisfy the demand established for the
systems they were initially designed for. This aspect leads
to the application of the concept of “second life”, “reuse”
or “recycle” into other different fields with demands of
fewer current values involved or uninterruptable energy
storage systems [16].
Concretely, LFP batteries have a useful period between
10
O. Izquierdo-Monge et al., Revista Facultad de Ingeniería, Universidad de Antioquia, No. 110, pp. 9-22, 2024
500 and 3500 cycles. When cells reach 80% of charge from
their initial capacity, it is considered out of the range of the
storage systems for which they were designed (currently
EV) [17]. However, after the initial 2000 cycles of these
batteries, their residual capacity can be 4 or 5 times higher
than in the case of any acid battery (lead, sulphuric acid).
It is estimated that LFP batteries still can have around
2000 more cycles in second-life applications until their
capacity is reduced to 60% and the loss of voltage is
enough to consider the end of this second period of life
[17].
From an economic aspect, these applications are
attractive, given the fact that their cost is reduced to
minimum values if we consider the price of a new battery,
as well as a useful extended period of life. The first step
in the investigation of LFP batteries and their life cycles is
aging them and knowing their capacity and their behaviour
against changes in charging and discharging processes,
which bring the answer to whether they are valid or not for
different energy storage systems and why. This represents
the principal aim of this document.
These applications are directly related to the energy
sources they are connected to, which has a strong
influence on the charge and discharge conditions. Deep
charging and discharging processes deplete battery life
very quickly, while those partial charging and discharging
processes extend this period [18]. This document is based
on the paper “Second life for LiFePo4 batteries as an
energy storage system in a smart microgrid” presented at
IV Iberoamerican Congress on Smart Cities (ISCS-CITIES
2021). Throughout this document, we present a study of
the selection and integration of a bank of LFP batteries
into the CE.D.E.R. microgrid to give a second life to those
cells in optimal working conditions after a period of
disuse. The rest of the paper is divided into different
sections: Section 2 represents the description of the
storage system, Section 3 involves the jobs carried out for
the recovery of the system, and Section 4 explains in detail
the integration of the battery system on a smart microgrid
for a second application. Finally, the conclusions obtained
are presented and the bibliography is cited.
2. Material and methods
CE.D.E.R. has an electrochemical storage system based
on LFP batteries, composed of 2 racks with 14 modules
and 14 cells in each module, making a total of 392
cells. The operation of the system is completed with a
grid three-phase inverter Ingecon Sun 30 (30 kW at 400
V AC and 50Hz) that can work with each of the racks
independently, or with both at the same time.
This storage system was installed in 2012 and was
involved in a research project in CE.D.E.R. with the aim of
making it able to supply high-power energy during short
periods of time (less than 1 hour).
Once the project was completed, and after several
years of non-use, an objective was proposed, which was to
put the system back into operation and integrate it into the
CE.D.E.R. smart microgrid as an energy storage system.
In this case, the purpose of the system is completely
different from the initial one, as instead of working with
high-power energy supplies for short periods of time, (30
kW for a maximum of 1 hour), the current main objective
is to work with lower power energy supplies for longer
periods of time (4-8 kW for a few hours).
When starting with the process we call battery recycling
(same components but for different end use), we face
a process where we only have the basic cell and rack
specifications that you can see in Table 1 and Table 2
without their charge and discharge curves and efficiency.
Table 1 Cell specifications
Cell technology Lithium Iron Phosphate (LFP)
Nominal Voltage 3.2 V
Min. Voltage 2.0 V
Max. Voltage 3.8 V
Capacity 50 Ah
Nominal Energy 160 Wh
Nominal Energy (discharging C2) 167 Wh
Nominal Energy (discharging C3) 176 Wh
Nominal Power 160 W
Max. Power 640 W
Table 2 Rack specifications
Cell technology Lithium Iron Phosphate (LFP)
Cells Number 196
Modules Number 14
Nominal Voltage 627.2 V
Min. Voltage 529.2 V
Max. Voltage 729.1 V
Capacity 50 Ah
Nominal Energy 31360 Wh
Nominal Power 31360 W
Max. Power 125400 W
To use those batteries as an energy storage system for the
microgrid, it is essential to work on two different fields.
On the one hand, the recovery of the damaged cells or
those that have suffered a decrease in their efficiency as
a consequence of the passage of time and their first use
application. On the other hand, it is necessary to work
on the communication system and its integration on the
11
500 and 3500 cycles. When cells reach 80% of charge from
their initial capacity, it is considered out of the range of the
storage systems for which they were designed (currently
EV) [17]. However, after the initial 2000 cycles of these
batteries, their residual capacity can be 4 or 5 times higher
than in the case of any acid battery (lead, sulphuric acid).
It is estimated that LFP batteries still can have around
2000 more cycles in second-life applications until their
capacity is reduced to 60% and the loss of voltage is
enough to consider the end of this second period of life
[17].
From an economic aspect, these applications are
attractive, given the fact that their cost is reduced to
minimum values if we consider the price of a new battery,
as well as a useful extended period of life. The first step
in the investigation of LFP batteries and their life cycles is
aging them and knowing their capacity and their behaviour
against changes in charging and discharging processes,
which bring the answer to whether they are valid or not for
different energy storage systems and why. This represents
the principal aim of this document.
These applications are directly related to the energy
sources they are connected to, which has a strong
influence on the charge and discharge conditions. Deep
charging and discharging processes deplete battery life
very quickly, while those partial charging and discharging
processes extend this period [18]. This document is based
on the paper “Second life for LiFePo4 batteries as an
energy storage system in a smart microgrid” presented at
IV Iberoamerican Congress on Smart Cities (ISCS-CITIES
2021). Throughout this document, we present a study of
the selection and integration of a bank of LFP batteries
into the CE.D.E.R. microgrid to give a second life to those
cells in optimal working conditions after a period of
disuse. The rest of the paper is divided into different
sections: Section 2 represents the description of the
storage system, Section 3 involves the jobs carried out for
the recovery of the system, and Section 4 explains in detail
the integration of the battery system on a smart microgrid
for a second application. Finally, the conclusions obtained
are presented and the bibliography is cited.
2. Material and methods
CE.D.E.R. has an electrochemical storage system based
on LFP batteries, composed of 2 racks with 14 modules
and 14 cells in each module, making a total of 392
cells. The operation of the system is completed with a
grid three-phase inverter Ingecon Sun 30 (30 kW at 400
V AC and 50Hz) that can work with each of the racks
independently, or with both at the same time.
This storage system was installed in 2012 and was
involved in a research project in CE.D.E.R. with the aim of
making it able to supply high-power energy during short
periods of time (less than 1 hour).
Once the project was completed, and after several
years of non-use, an objective was proposed, which was to
put the system back into operation and integrate it into the
CE.D.E.R. smart microgrid as an energy storage system.
In this case, the purpose of the system is completely
different from the initial one, as instead of working with
high-power energy supplies for short periods of time, (30
kW for a maximum of 1 hour), the current main objective
is to work with lower power energy supplies for longer
periods of time (4-8 kW for a few hours).
When starting with the process we call battery recycling
(same components but for different end use), we face
a process where we only have the basic cell and rack
specifications that you can see in Table 1 and Table 2
without their charge and discharge curves and efficiency.
Table 1 Cell specifications
Cell technology Lithium Iron Phosphate (LFP)
Nominal Voltage 3.2 V
Min. Voltage 2.0 V
Max. Voltage 3.8 V
Capacity 50 Ah
Nominal Energy 160 Wh
Nominal Energy (discharging C2) 167 Wh
Nominal Energy (discharging C3) 176 Wh
Nominal Power 160 W
Max. Power 640 W
Table 2 Rack specifications
Cell technology Lithium Iron Phosphate (LFP)
Cells Number 196
Modules Number 14
Nominal Voltage 627.2 V
Min. Voltage 529.2 V
Max. Voltage 729.1 V
Capacity 50 Ah
Nominal Energy 31360 Wh
Nominal Power 31360 W
Max. Power 125400 W
To use those batteries as an energy storage system for the
microgrid, it is essential to work on two different fields.
On the one hand, the recovery of the damaged cells or
those that have suffered a decrease in their efficiency as
a consequence of the passage of time and their first use
application. On the other hand, it is necessary to work
on the communication system and its integration on the
11
O. Izquierdo-Monge et al., Revista Facultad de Ingeniería, Universidad de Antioquia, No. 110, pp. 9-22, 2024
CE.D.E.R. microgrid.
3. Cells recovery
For the recovery process of the storage system, we always
refer and work at the cell level, as the average voltage
values of the modules can falsify the individual voltages
and, thus, compromise certain cells to the system as a
whole.
In addition to knowing their voltage and, therefore, their
state of charge, first of all, we must know their physical
state, so that we can establish an initial classification of
cells (suitable – unsuitable) to continue with the process
depending on whether they have swollen, suffered a
rupture or are structurally damaged.
To do this, all cells are removed from the two racks
and visually assessed, leaving only those cells that have
not suffered visible damage in the process. These will go
on to the next stages of recovery.
The next step in the storage system reuse process is
to check the capacity of each cell in the storage system.
It is necessary to take note of the voltage level of each
cell to estimate its State Of Health (SOH) with respect to
its nominal voltage. These measurements are used for
grouping similar cells in voltage and applying the same
criteria for cells belonging to the same group.
This group division is carried out with its own criteria,
taking into account significant values in the cells and in the
system such as: minimum voltage above which the BMS
does not recognise the cell (2.0 V), nominal cell voltage
(3.2 V) and maximum cell voltage (3.8 V) defined by the
manufacturer.
Another aspect that influences the definition of the
groups is the number of cells with similar voltage values.
When noting down the individual voltage values, it can
be seen that quite a few cells have the same voltage
value. This leads to the definition of groups with different
amplitude ranges in order to be able to group cells that
are as similar in voltage as possible and therefore behave
similarly. The highest voltage found was 3.3 V, and the
lowest, 0 V.
Taking into consideration the above, the 392 cells
that compound the system are classified into 5 groups,
depending on their voltage levels:
• 31 Cells with a voltage higher than 3.3 V.
• 86 Cells with a voltage between 3 and 3.3 V.
• 107 Cells with a voltage between 2 and 3 V.
• 121 Cells with a voltage between 0.5 and 2 V.
• 47 Cells with a voltage of 0.0 V.
The Battery Management System (BMS) of the system
does not detect cells with a voltage lower than 2 V, as that
corresponds to the minimum level established, and it is
considered that cells with a lower level are damaged and
not useful for this purpose. For this reason, it is necessary
to raise the voltage of those cells with other methods to
make the BMS able to recognise them.
In the process of charging the cells and increasing
their voltage, a charger is necessary. In our case, we
used a Revolectrix GT-500. It allowed us to charge and
discharge up to 6 cells of several types (different kinds
of lithium batteries, NiCd, NiMH and Lead Acid batteries)
simultaneously.
The charge/discharge process is user-defined. It sets the
current at which the process is carried out (with a limit of
20 A for charging and 8 A for discharging process) and the
voltage limit at which the process ends. The voltage and
current of the battery can be seen at any given time via
the display. However, generally, chargers (as in this case)
are not able to recognise voltage levels lower than 2 V in
LFP batteries. That is why it is necessary to connect these
cells to a power supply, so they can achieve a recognising
voltage level and, after that, charge them until a voltage
close to the nominal level. By way of this method, we
make those cells that initially had a voltage between 0.6
and 2 V convert this level until the nominal voltage (3.2 V).
On the other hand, the 47 cells with a level of 0.0 V could
not be recovered and were thrown out from the experiment.
Cells with a voltage higher than 2.0 V can be detected by
the BMS. Those cells between 2.0 V and 3.0 V are charged
with GT-500 until their nominal voltage, while those with a
charge higher than 3.0 V are not manipulated.
After a few days of rest, it is observed that in many
of those cells with voltages between 0.6 and 2.0 V, their
voltage level falls again until levels under 2.0 V. This
means that they were not going to be recognised by the
BMS and they are thrown out as well. The rest of the cells
that initially were between 0.6-2.0 V and 2.0-3.0 V also lost
some of their voltage levels but without reaching a level
lower than 2.0 V.
To know the functioning of cells in each group, they
are subjected to processes of charge and discharge with
the GT-500 charger to study their behaviour.
3.1 Cells charging process
To study the current state of charge on the different cells,
it is selected one of each group, a new cell (cell 1), a cell
12
CE.D.E.R. microgrid.
3. Cells recovery
For the recovery process of the storage system, we always
refer and work at the cell level, as the average voltage
values of the modules can falsify the individual voltages
and, thus, compromise certain cells to the system as a
whole.
In addition to knowing their voltage and, therefore, their
state of charge, first of all, we must know their physical
state, so that we can establish an initial classification of
cells (suitable – unsuitable) to continue with the process
depending on whether they have swollen, suffered a
rupture or are structurally damaged.
To do this, all cells are removed from the two racks
and visually assessed, leaving only those cells that have
not suffered visible damage in the process. These will go
on to the next stages of recovery.
The next step in the storage system reuse process is
to check the capacity of each cell in the storage system.
It is necessary to take note of the voltage level of each
cell to estimate its State Of Health (SOH) with respect to
its nominal voltage. These measurements are used for
grouping similar cells in voltage and applying the same
criteria for cells belonging to the same group.
This group division is carried out with its own criteria,
taking into account significant values in the cells and in the
system such as: minimum voltage above which the BMS
does not recognise the cell (2.0 V), nominal cell voltage
(3.2 V) and maximum cell voltage (3.8 V) defined by the
manufacturer.
Another aspect that influences the definition of the
groups is the number of cells with similar voltage values.
When noting down the individual voltage values, it can
be seen that quite a few cells have the same voltage
value. This leads to the definition of groups with different
amplitude ranges in order to be able to group cells that
are as similar in voltage as possible and therefore behave
similarly. The highest voltage found was 3.3 V, and the
lowest, 0 V.
Taking into consideration the above, the 392 cells
that compound the system are classified into 5 groups,
depending on their voltage levels:
• 31 Cells with a voltage higher than 3.3 V.
• 86 Cells with a voltage between 3 and 3.3 V.
• 107 Cells with a voltage between 2 and 3 V.
• 121 Cells with a voltage between 0.5 and 2 V.
• 47 Cells with a voltage of 0.0 V.
The Battery Management System (BMS) of the system
does not detect cells with a voltage lower than 2 V, as that
corresponds to the minimum level established, and it is
considered that cells with a lower level are damaged and
not useful for this purpose. For this reason, it is necessary
to raise the voltage of those cells with other methods to
make the BMS able to recognise them.
In the process of charging the cells and increasing
their voltage, a charger is necessary. In our case, we
used a Revolectrix GT-500. It allowed us to charge and
discharge up to 6 cells of several types (different kinds
of lithium batteries, NiCd, NiMH and Lead Acid batteries)
simultaneously.
The charge/discharge process is user-defined. It sets the
current at which the process is carried out (with a limit of
20 A for charging and 8 A for discharging process) and the
voltage limit at which the process ends. The voltage and
current of the battery can be seen at any given time via
the display. However, generally, chargers (as in this case)
are not able to recognise voltage levels lower than 2 V in
LFP batteries. That is why it is necessary to connect these
cells to a power supply, so they can achieve a recognising
voltage level and, after that, charge them until a voltage
close to the nominal level. By way of this method, we
make those cells that initially had a voltage between 0.6
and 2 V convert this level until the nominal voltage (3.2 V).
On the other hand, the 47 cells with a level of 0.0 V could
not be recovered and were thrown out from the experiment.
Cells with a voltage higher than 2.0 V can be detected by
the BMS. Those cells between 2.0 V and 3.0 V are charged
with GT-500 until their nominal voltage, while those with a
charge higher than 3.0 V are not manipulated.
After a few days of rest, it is observed that in many
of those cells with voltages between 0.6 and 2.0 V, their
voltage level falls again until levels under 2.0 V. This
means that they were not going to be recognised by the
BMS and they are thrown out as well. The rest of the cells
that initially were between 0.6-2.0 V and 2.0-3.0 V also lost
some of their voltage levels but without reaching a level
lower than 2.0 V.
To know the functioning of cells in each group, they
are subjected to processes of charge and discharge with
the GT-500 charger to study their behaviour.
3.1 Cells charging process
To study the current state of charge on the different cells,
it is selected one of each group, a new cell (cell 1), a cell
12
O. Izquierdo-Monge et al., Revista Facultad de Ingeniería, Universidad de Antioquia, No. 110, pp. 9-22, 2024
with a charge between 3.0-3.3 V (cell 2), a cell between
2-3 V (cell 3) and another with a voltage less than 2 V (cell 4).
The first step is fully discharging the four cells, to
make them be in a similar initial state of charge and, after
that, subject them to a process of charge at 8A (with a
limited voltage of 3.6 V for the four cells). The results are
the following:
Cell 1: This is the case of a new cell that represents
the correct operational behaviour of a LFP battery and
works as a reference to compare the state with the rest of
the cells. Its voltage starting point is 3.05 V, and at least
380 minutes to achieve the 3.6 V scheduled in the charger
as the maximum voltage.
In approximately 15 minutes, the cell has converted
the initial 3.05 V until 3.26 V. In the following 280 minutes,
close to 5 hours later, it produced a constant increase in
its voltage until achieving 3.4 V. The next 30 minutes, the
cell suffers again a quick increase of the voltage until 3.6
V. Finally, the cell maintains its level (3.6 V) for a while.
After stopping the process of charge, the voltage starts to
decrease its level until it stabilizes itself at 3.34 V.
Cell 2: At the beginning of the study, this cell had a
voltage level of between 3-3.3 V. This level is considered
right for the charge. During the process, it is observed 3
phases, as in cell 1, but with some differences.
The starting point is similar to cell 1 (3.06 V). In this
case, the first phase of charge is shorter than in case 1
(13 minutes) and rapidly achieves a voltage slightly higher
than the case before (3.27 V). In the second phase, where
the increase is slow and progressive, the time involved was
250 minutes approximately, until achieving 3.52 V. Finally,
the last phase lasts 20 minutes and the voltage is rapidly
increased again, until achieving the 3.6 V scheduled in the
charger.
The time involved in achieving 3.6 V is 280 minutes, a
notable difference if we compare it with the 380 minutes
with cell 1.
When the charge is finished, as in cell 1, the voltage
level starts decreasing until it stabilizes after a couple of
hours in a value close to the nominal (3.3 V).
Cell 3: At the beginning of this study, cell 3 had a
voltage level of between 2 and 3 V. This range of voltage
is a bit low. However, it is within the limits defined by the
fabricant and understandable, considering that the system
has not been used for a long time. Its behaviour is similar
to cell 2, as can be observed in Figure 1.
Cell 4: This cell had a starting point at the beginning
of the study lower than 2 V, at the limit of the voltage
established by the manufacturer. This is a sign that the
cell may not be suitable for recovery for a second life
application.
As the starting voltage point is lower than 2 V (concretely
0.6 V) the charger cannot recognise it, so it is necessary to
raise the voltage using the power supply until more than
2 V. By doing this, the charger can recognize the cell and
start charging it.
During the first phase of charge, the voltage quickly
rises to 3.24 V in 30 minutes. From this moment, its
behaviour is similar to cell 1. Phase 2 lasts 240 minutes
and reaches 3.4 V. Finally, in phase 3, the voltage level
increases again quickly until 3.6 V. The total time involved
in the charging process is slightly lower than in cell 1 (330
minutes).
This could lead us to the conclusion that, in contrast
to what initially seemed, these cells could be reused in
second-life applications. However, when the charge is
stopped, the voltage levels decrease until values are lower
than 2 V. This makes the systems unable to recognise
them.
The summary of the results obtained is shown in Figure
1. The charging process is repeated with the same cells,
but at different current levels to check if there are any
differences. The results obtained lead to the conclusion
that the behaviour is similar in all cases. Figure 2 shows
the results obtained in a charging process with 16 A.
3.2 Cells discharging process
After charging the cells, they were subjected to discharging
processes, limiting the voltage in the charger to 3 V, to
compare the results. The initial current value of the
discharging processes is 8 A.
The results obtained for the four cells are shown in
Figure 3.
Cell 1: Discharging process starts with 3.34 V. In about 10
minutes, this level is decreased until 3.26 V, maintaining
this value until almost 3 hours later. After this moment,
the voltage decreases again, but much more quickly than
before, until it achieves the 3 V scheduled in the charger.
The total time involved in the discharging process of
this cell from the initial 3.34 V to 3 V was 360 minutes.
When the discharge was stopped, and after hours, the cell
slightly increased the voltage to 3.2 V.
13
with a charge between 3.0-3.3 V (cell 2), a cell between
2-3 V (cell 3) and another with a voltage less than 2 V (cell 4).
The first step is fully discharging the four cells, to
make them be in a similar initial state of charge and, after
that, subject them to a process of charge at 8A (with a
limited voltage of 3.6 V for the four cells). The results are
the following:
Cell 1: This is the case of a new cell that represents
the correct operational behaviour of a LFP battery and
works as a reference to compare the state with the rest of
the cells. Its voltage starting point is 3.05 V, and at least
380 minutes to achieve the 3.6 V scheduled in the charger
as the maximum voltage.
In approximately 15 minutes, the cell has converted
the initial 3.05 V until 3.26 V. In the following 280 minutes,
close to 5 hours later, it produced a constant increase in
its voltage until achieving 3.4 V. The next 30 minutes, the
cell suffers again a quick increase of the voltage until 3.6
V. Finally, the cell maintains its level (3.6 V) for a while.
After stopping the process of charge, the voltage starts to
decrease its level until it stabilizes itself at 3.34 V.
Cell 2: At the beginning of the study, this cell had a
voltage level of between 3-3.3 V. This level is considered
right for the charge. During the process, it is observed 3
phases, as in cell 1, but with some differences.
The starting point is similar to cell 1 (3.06 V). In this
case, the first phase of charge is shorter than in case 1
(13 minutes) and rapidly achieves a voltage slightly higher
than the case before (3.27 V). In the second phase, where
the increase is slow and progressive, the time involved was
250 minutes approximately, until achieving 3.52 V. Finally,
the last phase lasts 20 minutes and the voltage is rapidly
increased again, until achieving the 3.6 V scheduled in the
charger.
The time involved in achieving 3.6 V is 280 minutes, a
notable difference if we compare it with the 380 minutes
with cell 1.
When the charge is finished, as in cell 1, the voltage
level starts decreasing until it stabilizes after a couple of
hours in a value close to the nominal (3.3 V).
Cell 3: At the beginning of this study, cell 3 had a
voltage level of between 2 and 3 V. This range of voltage
is a bit low. However, it is within the limits defined by the
fabricant and understandable, considering that the system
has not been used for a long time. Its behaviour is similar
to cell 2, as can be observed in Figure 1.
Cell 4: This cell had a starting point at the beginning
of the study lower than 2 V, at the limit of the voltage
established by the manufacturer. This is a sign that the
cell may not be suitable for recovery for a second life
application.
As the starting voltage point is lower than 2 V (concretely
0.6 V) the charger cannot recognise it, so it is necessary to
raise the voltage using the power supply until more than
2 V. By doing this, the charger can recognize the cell and
start charging it.
During the first phase of charge, the voltage quickly
rises to 3.24 V in 30 minutes. From this moment, its
behaviour is similar to cell 1. Phase 2 lasts 240 minutes
and reaches 3.4 V. Finally, in phase 3, the voltage level
increases again quickly until 3.6 V. The total time involved
in the charging process is slightly lower than in cell 1 (330
minutes).
This could lead us to the conclusion that, in contrast
to what initially seemed, these cells could be reused in
second-life applications. However, when the charge is
stopped, the voltage levels decrease until values are lower
than 2 V. This makes the systems unable to recognise
them.
The summary of the results obtained is shown in Figure
1. The charging process is repeated with the same cells,
but at different current levels to check if there are any
differences. The results obtained lead to the conclusion
that the behaviour is similar in all cases. Figure 2 shows
the results obtained in a charging process with 16 A.
3.2 Cells discharging process
After charging the cells, they were subjected to discharging
processes, limiting the voltage in the charger to 3 V, to
compare the results. The initial current value of the
discharging processes is 8 A.
The results obtained for the four cells are shown in
Figure 3.
Cell 1: Discharging process starts with 3.34 V. In about 10
minutes, this level is decreased until 3.26 V, maintaining
this value until almost 3 hours later. After this moment,
the voltage decreases again, but much more quickly than
before, until it achieves the 3 V scheduled in the charger.
The total time involved in the discharging process of
this cell from the initial 3.34 V to 3 V was 360 minutes.
When the discharge was stopped, and after hours, the cell
slightly increased the voltage to 3.2 V.
13
O. Izquierdo-Monge et al., Revista Facultad de Ingeniería, Universidad de Antioquia, No. 110, pp. 9-22, 2024
Figure 1 Cells charging process with GT-500 at 8 A
Figure 2 Cells charging process with GT-500 at 16 A
Figure 3 Cells discharging process with GT-500 at 8 A
14
Figure 1 Cells charging process with GT-500 at 8 A
Figure 2 Cells charging process with GT-500 at 16 A
Figure 3 Cells discharging process with GT-500 at 8 A
14
O. Izquierdo-Monge et al., Revista Facultad de Ingeniería, Universidad de Antioquia, No. 110, pp. 9-22, 2024
Figure 4 Cells discharging process with GT-500 at 4 A
Cell 2: Initially, this cell had a voltage of 3.32 V. This
level is a bit higher than cell 1. The voltage was decreased
in a few minutes until 3.21 V and maintained in that value
for almost 1 hour. After that, it goes down to 3.16 V, staying
at that value for more than 1 hour. From that moment, the
voltage starts decreasing quickly until 3 V.
In this case, the total time involved was 280 minutes
approximately. When the discharge finished, as occurred
in cell 1, the voltage value of the cell achieved its nominal
value.
Cell 3: As in the charging process, the discharge of
cell 3 is like cell 2. However, there are some differences
related to its bad conditions. A clear example of this is the
time involved in the process, 240 minutes.
Cell 4: The total time involved in this cell was 300
minutes, less than cell 1 but higher than cells 2 and 3.
In phase 1, its voltage changes from the initial 3.32 V
to 3.24 V in less than 5 minutes. This value is maintained
for almost 2 hours and after that, it starts decreasing more
quickly until 3 V.
In this case, the nominal value it achieves once the
cell is not connected to the charger for some time, is
enough evidence to demonstrate that the cell is not
working properly. As it happened during the charging
process, once disconnected, the nominal value increases
(up to 3.1 V), but instead of stabilizing, it starts falling as
time passes. The discharging process is repeated with
the same cells at different current levels to check if there
are any differences. The behaviour is similar in all cases.
Figure 4 shows discharging process at 4 A.
3.3 Cells selection
Considering the results obtained in the sections of this
document, it is demonstrated that the best options for
second-life applications are those cells whose behaviour
is like 2 and 3; otherwise, 4 is not.
Apparently, cells like 4 have a better capacity and
could be used during long and continuous processes of
charge, given the fact that if there are not long periods
where the voltage falls, the cells work correctly. However,
when they are in long periods of time without the charger
connected to them, the voltage falls significantly. This
can lead to a voltage lower than 2 V, which means that
they are not going to be recognised by the system anymore.
On the one hand, cells 2 and 3 have worse charging
capacity, but on the other hand, they can keep constant
voltage levels without being connected to the system;
meaning that, from a functional point of view, they work
better for our purpose. That is the reason why they are
selected to take part in the system. Once the best cells
were chosen to take part in the energy storage system
formed by the 196 batteries necessary in each rack,
the next step is integrating them into the system of the
microgrid.
4. Second life. Integration of the
energy storage system in the
microgrid
In order to integrate the energy storage system with LFP
in the microgrid of CE.D.E.R., it is necessary to run out the
next tasks:
• Establish the communication between the
15
Figure 4 Cells discharging process with GT-500 at 4 A
Cell 2: Initially, this cell had a voltage of 3.32 V. This
level is a bit higher than cell 1. The voltage was decreased
in a few minutes until 3.21 V and maintained in that value
for almost 1 hour. After that, it goes down to 3.16 V, staying
at that value for more than 1 hour. From that moment, the
voltage starts decreasing quickly until 3 V.
In this case, the total time involved was 280 minutes
approximately. When the discharge finished, as occurred
in cell 1, the voltage value of the cell achieved its nominal
value.
Cell 3: As in the charging process, the discharge of
cell 3 is like cell 2. However, there are some differences
related to its bad conditions. A clear example of this is the
time involved in the process, 240 minutes.
Cell 4: The total time involved in this cell was 300
minutes, less than cell 1 but higher than cells 2 and 3.
In phase 1, its voltage changes from the initial 3.32 V
to 3.24 V in less than 5 minutes. This value is maintained
for almost 2 hours and after that, it starts decreasing more
quickly until 3 V.
In this case, the nominal value it achieves once the
cell is not connected to the charger for some time, is
enough evidence to demonstrate that the cell is not
working properly. As it happened during the charging
process, once disconnected, the nominal value increases
(up to 3.1 V), but instead of stabilizing, it starts falling as
time passes. The discharging process is repeated with
the same cells at different current levels to check if there
are any differences. The behaviour is similar in all cases.
Figure 4 shows discharging process at 4 A.
3.3 Cells selection
Considering the results obtained in the sections of this
document, it is demonstrated that the best options for
second-life applications are those cells whose behaviour
is like 2 and 3; otherwise, 4 is not.
Apparently, cells like 4 have a better capacity and
could be used during long and continuous processes of
charge, given the fact that if there are not long periods
where the voltage falls, the cells work correctly. However,
when they are in long periods of time without the charger
connected to them, the voltage falls significantly. This
can lead to a voltage lower than 2 V, which means that
they are not going to be recognised by the system anymore.
On the one hand, cells 2 and 3 have worse charging
capacity, but on the other hand, they can keep constant
voltage levels without being connected to the system;
meaning that, from a functional point of view, they work
better for our purpose. That is the reason why they are
selected to take part in the system. Once the best cells
were chosen to take part in the energy storage system
formed by the 196 batteries necessary in each rack,
the next step is integrating them into the system of the
microgrid.
4. Second life. Integration of the
energy storage system in the
microgrid
In order to integrate the energy storage system with LFP
in the microgrid of CE.D.E.R., it is necessary to run out the
next tasks:
• Establish the communication between the
15
O. Izquierdo-Monge et al., Revista Facultad de Ingeniería, Universidad de Antioquia, No. 110, pp. 9-22, 2024
management app of the smart grid at CE.D.E.R.
(developed with Home Assistant) and the control
system of batteries (BMS-Battery Management
System) to know the information about cells and
regulate the power supply of charges and discharges
with the converter.
• Once the communication is established, it
is necessary to develop a SCADA inside the
management software to control the operation
of the batteries, so they could be managed manually
with the microgrid operator.
• Finally, it is necessary to establish the automation,
so the strategies of energy management, defined by
the grid administrator, can be applied automatically
without the need for an operator.
4.1 Communication with the management
system and batteries control system
The control system is formed by many different elements:
• BMS of each module: Each of the 14 modules of a rack
has a BMS that allows keeping under observation the
voltage and temperature of each o of the 14 cells of
each module. This BMS carries out a cell calibration
during its charge and controls the fan to keep the
system cool. The communication between the BMS
of the rack is carried through by means of a CAN bus.
• BMS of each rack: keeps the voltage and temperature
of the 14 modules that take part in each rack under
observation. It also calculates the state of charge and
the health conditions (SOH-State of Health) of each
module. It establishes the communication with BMS
of the system by means of the CAN bus.
• BMS of the system: keeps under observation the
voltage and temperature of the two racks that
make up the system. It leads the Modbus-RTU to
communicate with a computer, where the software for
the control of the storage system is installed.
The management app for the microgrid can communicate
with BMS of the system using the protocol of Modbus
communication in two different ways:
• Connecting a computer/Raspberry/Arduino etc., to
the cable RS485 of the BMS of the storage system
with a SCADA. This allows to monitor the information
by way of a Modbus-RTU communication protocol,
and after that, the management app of the microgrid
must communicate with the SCADA to receive the
information.
• Using a Modbus RTU-TCP/IP converter. A Modbus
RTU (RS485) connects the exit of the BMS to the
entrance of the converter. At the same time, the
RTU-TCI/IP converter is connected to the microgrid’s
data network, thanks to a RJ45-Ethernet cable. In this
way, the software can communicate directly with it.
The SCADA allows to monitor the information and it
has been developed with the same managing app as
the microgrid.
The first option presents some disadvantages that make
the second one much more interesting. On the one hand,
by using Modbus RTU, only one device at a time is allowed
to communicate; whereas by using Modbus TCP/IP, it is
possible to communicate with more than one device at the
same time. On the other hand, connecting the RS485 of
the BMS to a device (pc, Raspberry or Arduino) is more
expensive and complex than an RTU-TCP/IP converter, as
it would be necessary to have two different SCADAs. One
of them for the monitoring of the BMS and the other, for
the development of the control system of the microgrid, to
receive that information.
Also, apart from the communication with BMS of the
energy storage system, the control system of the
microgrid has to communicate with the converter, so
the transmission of the data of charging and discharging
processes is possible, as the BMS only produces the
information about the batteries state, but does not allow
its charge and discharge.
The converter allows communication by means of
Modbus TCP/IP, so it can be connected to the data grid
Ethernet of the microgrid. The control software can also
directly communicate with it once the parameters are
defined.
Figure 5 shows a diagram of communications of the
control system in the microgrid with the energy storage
system and the converter [19].
Once Modbus is established as a way of communication
with the BMS of the system as well as with the converter
of the grid, it is necessary to define the equipment in the
configuration of the software used for the management of
the system. In this case, the managing software is carried
out with Home Assistant, software developed in Python.
Despite it being used generally for home automation, it is
a good solution for the monitoring and controlling of the
microgrid in real time.
To establish communication, it is necessary to define
both elements. On one side, the BMS of the system and,
on the other side, the converter to grid in the configuration
file. To do this, it is necessary to know the communication
protocol (Modbus TCP), IP direction of each element and
the connection link (which generally is 502 for Modbus
TCP).
16
management app of the smart grid at CE.D.E.R.
(developed with Home Assistant) and the control
system of batteries (BMS-Battery Management
System) to know the information about cells and
regulate the power supply of charges and discharges
with the converter.
• Once the communication is established, it
is necessary to develop a SCADA inside the
management software to control the operation
of the batteries, so they could be managed manually
with the microgrid operator.
• Finally, it is necessary to establish the automation,
so the strategies of energy management, defined by
the grid administrator, can be applied automatically
without the need for an operator.
4.1 Communication with the management
system and batteries control system
The control system is formed by many different elements:
• BMS of each module: Each of the 14 modules of a rack
has a BMS that allows keeping under observation the
voltage and temperature of each o of the 14 cells of
each module. This BMS carries out a cell calibration
during its charge and controls the fan to keep the
system cool. The communication between the BMS
of the rack is carried through by means of a CAN bus.
• BMS of each rack: keeps the voltage and temperature
of the 14 modules that take part in each rack under
observation. It also calculates the state of charge and
the health conditions (SOH-State of Health) of each
module. It establishes the communication with BMS
of the system by means of the CAN bus.
• BMS of the system: keeps under observation the
voltage and temperature of the two racks that
make up the system. It leads the Modbus-RTU to
communicate with a computer, where the software for
the control of the storage system is installed.
The management app for the microgrid can communicate
with BMS of the system using the protocol of Modbus
communication in two different ways:
• Connecting a computer/Raspberry/Arduino etc., to
the cable RS485 of the BMS of the storage system
with a SCADA. This allows to monitor the information
by way of a Modbus-RTU communication protocol,
and after that, the management app of the microgrid
must communicate with the SCADA to receive the
information.
• Using a Modbus RTU-TCP/IP converter. A Modbus
RTU (RS485) connects the exit of the BMS to the
entrance of the converter. At the same time, the
RTU-TCI/IP converter is connected to the microgrid’s
data network, thanks to a RJ45-Ethernet cable. In this
way, the software can communicate directly with it.
The SCADA allows to monitor the information and it
has been developed with the same managing app as
the microgrid.
The first option presents some disadvantages that make
the second one much more interesting. On the one hand,
by using Modbus RTU, only one device at a time is allowed
to communicate; whereas by using Modbus TCP/IP, it is
possible to communicate with more than one device at the
same time. On the other hand, connecting the RS485 of
the BMS to a device (pc, Raspberry or Arduino) is more
expensive and complex than an RTU-TCP/IP converter, as
it would be necessary to have two different SCADAs. One
of them for the monitoring of the BMS and the other, for
the development of the control system of the microgrid, to
receive that information.
Also, apart from the communication with BMS of the
energy storage system, the control system of the
microgrid has to communicate with the converter, so
the transmission of the data of charging and discharging
processes is possible, as the BMS only produces the
information about the batteries state, but does not allow
its charge and discharge.
The converter allows communication by means of
Modbus TCP/IP, so it can be connected to the data grid
Ethernet of the microgrid. The control software can also
directly communicate with it once the parameters are
defined.
Figure 5 shows a diagram of communications of the
control system in the microgrid with the energy storage
system and the converter [19].
Once Modbus is established as a way of communication
with the BMS of the system as well as with the converter
of the grid, it is necessary to define the equipment in the
configuration of the software used for the management of
the system. In this case, the managing software is carried
out with Home Assistant, software developed in Python.
Despite it being used generally for home automation, it is
a good solution for the monitoring and controlling of the
microgrid in real time.
To establish communication, it is necessary to define
both elements. On one side, the BMS of the system and,
on the other side, the converter to grid in the configuration
file. To do this, it is necessary to know the communication
protocol (Modbus TCP), IP direction of each element and
the connection link (which generally is 502 for Modbus
TCP).
16
O. Izquierdo-Monge et al., Revista Facultad de Ingeniería, Universidad de Antioquia, No. 110, pp. 9-22, 2024
Figure 5 Communications diagram
-name: BMS_sistema_LFP # name (BMS)
type: tcp # type of Modbus (RTU o TCP)
host: 192.168.15.102 # BMS IP address
port: 502 # communication port
-name: inversor_red_LFP # name (Inverter)
type: tcp # type of Modbus (RTU o TCP)
host: 192.168.15.90 # Inverter IP address
port: 502 # communication port
Once the elements of the configuration file are defined, it
is necessary to read the desired Modbus directions and to
know the plot of each Modbus equipment. In the case of the
BMS, it is not possible to send codes to it, as it only provides
information about the state of the system; therefore, in
order to monitor, only reading the direction is necessary.
Examples are the temperatures and the voltage of cells. All
this is defined with the configuration file of Home Assistant
in the following way:
- platform: Modbus # Modbus
scan_interval: 1 # monitoring interval
registers:
- name: Tensión_maxima_célula # variable to monitoring
hub: BMS_sistema_LFP # hub previously defined
register_type: input # type of register (input,
holding, etc.)
unit_of_measurement: W # unit_of_measurement
slave: 1 # identification number
register: 44 # Modbus address
count: 1 # number of address
scale: 1 # scale (x1, x10, etc.)
Regarding the inverter, apart from reading the parameters
like voltage, current values, and power supply for its
monitoring (defining them in the configuration file), it is
also necessary to send codes to carry out the charging and
discharging processes in the specific conditions (previously
defined). This is done by means of a script:
potencia_inversor_litio: # script name
mode: single
sequence: # sequence of commands
- data:
address: ’1000’ # Modbus address
hub: inversor_red_LFP # hub previously defined
unit: ’1’ # identification number
value:
- 8 # Defined value to battery charge
/discharge
- sensor.consigna_potencia # charge/discharge power
service: modbus.write_register
The last step in the integration of the storage system
is the creation of a control panel (interface) with Home
Assistant. This allows visualising the parameters in real
time, monitoring the BMS of the system and the converter,
and sending codes for charging and discharging the
batteries intuitively.
Using Home Assistant is very easy to develop this
interface. The starting point is a frame that allows
inserting the cards with many different functions. It is
possible to add cards with all the values registered in the
configuration files.
17
Figure 5 Communications diagram
-name: BMS_sistema_LFP # name (BMS)
type: tcp # type of Modbus (RTU o TCP)
host: 192.168.15.102 # BMS IP address
port: 502 # communication port
-name: inversor_red_LFP # name (Inverter)
type: tcp # type of Modbus (RTU o TCP)
host: 192.168.15.90 # Inverter IP address
port: 502 # communication port
Once the elements of the configuration file are defined, it
is necessary to read the desired Modbus directions and to
know the plot of each Modbus equipment. In the case of the
BMS, it is not possible to send codes to it, as it only provides
information about the state of the system; therefore, in
order to monitor, only reading the direction is necessary.
Examples are the temperatures and the voltage of cells. All
this is defined with the configuration file of Home Assistant
in the following way:
- platform: Modbus # Modbus
scan_interval: 1 # monitoring interval
registers:
- name: Tensión_maxima_célula # variable to monitoring
hub: BMS_sistema_LFP # hub previously defined
register_type: input # type of register (input,
holding, etc.)
unit_of_measurement: W # unit_of_measurement
slave: 1 # identification number
register: 44 # Modbus address
count: 1 # number of address
scale: 1 # scale (x1, x10, etc.)
Regarding the inverter, apart from reading the parameters
like voltage, current values, and power supply for its
monitoring (defining them in the configuration file), it is
also necessary to send codes to carry out the charging and
discharging processes in the specific conditions (previously
defined). This is done by means of a script:
potencia_inversor_litio: # script name
mode: single
sequence: # sequence of commands
- data:
address: ’1000’ # Modbus address
hub: inversor_red_LFP # hub previously defined
unit: ’1’ # identification number
value:
- 8 # Defined value to battery charge
/discharge
- sensor.consigna_potencia # charge/discharge power
service: modbus.write_register
The last step in the integration of the storage system
is the creation of a control panel (interface) with Home
Assistant. This allows visualising the parameters in real
time, monitoring the BMS of the system and the converter,
and sending codes for charging and discharging the
batteries intuitively.
Using Home Assistant is very easy to develop this
interface. The starting point is a frame that allows
inserting the cards with many different functions. It is
possible to add cards with all the values registered in the
configuration files.
17
O. Izquierdo-Monge et al., Revista Facultad de Ingeniería, Universidad de Antioquia, No. 110, pp. 9-22, 2024
Figure 6 Control panel in Home Assistant
Figure 7 Battery errors
This can be a numerical or historic graphic. Figure 6
shows the control panel of the storage system during the
process of charge of the batteries. Figure 6 shows a 4-kW
discharge process. The system voltage is 630 V, and the
batteries are providing 4050 W to the grid (current value
6.43 A). The average voltage at the rack is 3.198 V; with
cell 3 of module 11 as the maximum (3.216 V) and cell
1 of module 8 as the minimum (3.139 V). The average
temperature of the rack is 20 ºC. Also, it can be graphically
observed that at the beginning of the discharge, the
voltage of cells (the maximum and the minimum) slightly
falls until achieving a value that is maintained for a long
time during the discharging process.
Figure 7 shows the interface with the errors that can
occur in batteries. In that moment, during the discharging
18
Figure 6 Control panel in Home Assistant
Figure 7 Battery errors
This can be a numerical or historic graphic. Figure 6
shows the control panel of the storage system during the
process of charge of the batteries. Figure 6 shows a 4-kW
discharge process. The system voltage is 630 V, and the
batteries are providing 4050 W to the grid (current value
6.43 A). The average voltage at the rack is 3.198 V; with
cell 3 of module 11 as the maximum (3.216 V) and cell
1 of module 8 as the minimum (3.139 V). The average
temperature of the rack is 20 ºC. Also, it can be graphically
observed that at the beginning of the discharge, the
voltage of cells (the maximum and the minimum) slightly
falls until achieving a value that is maintained for a long
time during the discharging process.
Figure 7 shows the interface with the errors that can
occur in batteries. In that moment, during the discharging
18
O. Izquierdo-Monge et al., Revista Facultad de Ingeniería, Universidad de Antioquia, No. 110, pp. 9-22, 2024
Figure 8 Charging and discharging process at 4 kW. Power
Figure 9 Charging and discharging process at 4 kW. Maximum and minimum voltage
Figure 10 Consumption/Discharging CE.D.E.R. microgrid – Network (with and without LFP batteries)
19
Figure 8 Charging and discharging process at 4 kW. Power
Figure 9 Charging and discharging process at 4 kW. Maximum and minimum voltage
Figure 10 Consumption/Discharging CE.D.E.R. microgrid – Network (with and without LFP batteries)
19
O. Izquierdo-Monge et al., Revista Facultad de Ingeniería, Universidad de Antioquia, No. 110, pp. 9-22, 2024
process, there are neither alarms (Warnings) nor errors
(Faults). The system notifies when the minimum voltage
of the cell falls under the minimum working parameters.
Once the system is integrated into the LFP batteries
system in the management app of the microgrid, it can
be used as a storage system for a second-life application.
This second life is under lower voltages than those that
the system was initially created for.
Figures 8 and 9 show graphically the entire charging
and discharging processes with a power value of 4 kW (the
charging power supply is -4000 W and the discharging
4000 W). Figure 8 shows the power supply of the system,
and Figure 9 shows the maximum and minimum voltage
of the cells.
4.2 Operation of the recovered LFP battery
system in CE.D.E.R.´s microgrid
This last section shows the operation and contribution of
the recovered LFP battery system within the microgrid,
taking into account the criteria for cell selection and
integration within the monitoring and control system
developed in the previous sections.
Thanks to the monitoring system that collects data
from the batteries, we have been able to carry out an
ex-post study of the system´s performance.
Figure 10 shows the operation of the microgrid versus
the distribution grid with particular reference to the LFP
batteries during a full day.
To better understand the graph and the influences of
the incorporation of these LFP batteries into the CE.D.E.R.
microgrid on the centre’s electricity bill, we will explain
how the current contract with the distribution network
breaks down.
The centre has an electricity contract with an electricity
distribution company on a commercial basis, which
disaggregates the 24 hours a day into six different billing
periods. The most expensive period is P1, and the
cheapest is P6, with costs decreasing progressively from
one to six.
The final costs for each period are broken down by
different concepts, including the energy consumed during
that period and the contracted power. If the consumption
is higher than the contracted power for that period, an
extra charge will be applied to the bill, so it is important to
ensure that the power never exceeds the power detailed in
the contract.
On the other hand, the only variable concepts in terms of
quantity are related to energy: the more kWh consumed,
the higher the electricity bill is. Conversely, the less
energy consumed, the lower the amount. With these
two aspects in mind, we shall look at the influence of
batteries on the microgrid as a whole. The graph shows
the operation of the microgrid in the centre versus the
distribution grid for a full day. In it, we see three periods
of battery operation. Two periods of discharge have
helped to flatten the peak consumption and, therefore,
reduce the energy consumed from the grid. And a battery
charging period has contributed to reducing the discharge
of surplus generation into the distribution grid.
Extrapolating these daily results to the rest of the
year would be similar. We see that the operation and
influence of the battery system in the microgrid of the
centre fulfils the following functions:
Flatten consumption peaks during the first and last
hours of the day. The periods coincide with the start of
the working day and late afternoon when the generation
decreases.
Reduce the discharge of surplus to the distribution
network during the peak generation period of the day.
Reduce the cost of the monthly electricity bill due to
lower consumption of active energy.
Consider reducing the contracted power to reduce
the cost of this item on the electricity bill.
5. Conclusions
This paper explains the steps followed in CE.D.E.R. to give
a second life to LFP batteries as an energy storage system
in a smart microgrid, after not being used for their initial
purpose for a long time.
After checking the initial state of health of each cell
and studying their charging and discharging curves (V-t),
those that performed the best were selected to make
operational one of the racks. There were not enough
cells to recover both racks because many of them were
damaged.
Once one of the racks has been completed with the
best-performance cells, the battery management system
(BMS) and the inverter to the grid were integrated into the
control system of CE.D.E.R.’s microgrid, allowing its use
(second life) as the energy storage system of the microgrid.
In its second life, this battery system cannot put up
with fast processes of charge and discharge with high
20
process, there are neither alarms (Warnings) nor errors
(Faults). The system notifies when the minimum voltage
of the cell falls under the minimum working parameters.
Once the system is integrated into the LFP batteries
system in the management app of the microgrid, it can
be used as a storage system for a second-life application.
This second life is under lower voltages than those that
the system was initially created for.
Figures 8 and 9 show graphically the entire charging
and discharging processes with a power value of 4 kW (the
charging power supply is -4000 W and the discharging
4000 W). Figure 8 shows the power supply of the system,
and Figure 9 shows the maximum and minimum voltage
of the cells.
4.2 Operation of the recovered LFP battery
system in CE.D.E.R.´s microgrid
This last section shows the operation and contribution of
the recovered LFP battery system within the microgrid,
taking into account the criteria for cell selection and
integration within the monitoring and control system
developed in the previous sections.
Thanks to the monitoring system that collects data
from the batteries, we have been able to carry out an
ex-post study of the system´s performance.
Figure 10 shows the operation of the microgrid versus
the distribution grid with particular reference to the LFP
batteries during a full day.
To better understand the graph and the influences of
the incorporation of these LFP batteries into the CE.D.E.R.
microgrid on the centre’s electricity bill, we will explain
how the current contract with the distribution network
breaks down.
The centre has an electricity contract with an electricity
distribution company on a commercial basis, which
disaggregates the 24 hours a day into six different billing
periods. The most expensive period is P1, and the
cheapest is P6, with costs decreasing progressively from
one to six.
The final costs for each period are broken down by
different concepts, including the energy consumed during
that period and the contracted power. If the consumption
is higher than the contracted power for that period, an
extra charge will be applied to the bill, so it is important to
ensure that the power never exceeds the power detailed in
the contract.
On the other hand, the only variable concepts in terms of
quantity are related to energy: the more kWh consumed,
the higher the electricity bill is. Conversely, the less
energy consumed, the lower the amount. With these
two aspects in mind, we shall look at the influence of
batteries on the microgrid as a whole. The graph shows
the operation of the microgrid in the centre versus the
distribution grid for a full day. In it, we see three periods
of battery operation. Two periods of discharge have
helped to flatten the peak consumption and, therefore,
reduce the energy consumed from the grid. And a battery
charging period has contributed to reducing the discharge
of surplus generation into the distribution grid.
Extrapolating these daily results to the rest of the
year would be similar. We see that the operation and
influence of the battery system in the microgrid of the
centre fulfils the following functions:
Flatten consumption peaks during the first and last
hours of the day. The periods coincide with the start of
the working day and late afternoon when the generation
decreases.
Reduce the discharge of surplus to the distribution
network during the peak generation period of the day.
Reduce the cost of the monthly electricity bill due to
lower consumption of active energy.
Consider reducing the contracted power to reduce
the cost of this item on the electricity bill.
5. Conclusions
This paper explains the steps followed in CE.D.E.R. to give
a second life to LFP batteries as an energy storage system
in a smart microgrid, after not being used for their initial
purpose for a long time.
After checking the initial state of health of each cell
and studying their charging and discharging curves (V-t),
those that performed the best were selected to make
operational one of the racks. There were not enough
cells to recover both racks because many of them were
damaged.
Once one of the racks has been completed with the
best-performance cells, the battery management system
(BMS) and the inverter to the grid were integrated into the
control system of CE.D.E.R.’s microgrid, allowing its use
(second life) as the energy storage system of the microgrid.
In its second life, this battery system cannot put up
with fast processes of charge and discharge with high
20
O. Izquierdo-Monge et al., Revista Facultad de Ingeniería, Universidad de Antioquia, No. 110, pp. 9-22, 2024
power supplies, close to 30 kW (just a few minutes).
However, it can work with lower power supplies (around
5 kW) for several hours, which is very useful in the
energy management of the microgrid. These allow the
storage of energy surplus from distributed generation
systems (wind turbines and photovoltaic systems) and use
that energy when needed, improving microgrid’s efficiency.
Future work would be to recover the second rack by
using the remaining cells and buying new ones to
complete the total of 196 cells. This would double the
storage capacity, in addition to improving the performance
of the entire system by introducing new cells.
6. Declaration of competing interest
We declare that we have no significant competing interests,
including financial or non-financial, professional, or
personal interests interfering with the full and objective
presentation of the work described in this manuscript.
7. Funding
This work was supported by CEDER-CIEMAT.
8. Author contributions
Conceptualization, O.I-M.; Methodology, O.I-M., P.P-C.
and N.A.G.; Software, O.I-M.; writing—original draft
preparation O.I-M. N.A.G. and P.P-C.; writing.
9. Data availability statement
Data were collected at CEDER-CIEMAT from May to
September 2021. We used the BMS of the LFP
battery systems and Ingeteam inverter, connected to
CEDER´s both microgrid energy management control
system developed with Home Assistant to measure power
and voltage.
References
[1] M. H. Lee and D. Shang-Chang, “Allocative efficiency of high-power
li-ion batteries from automotive mode (am) to storage mode (sm),”
Renewable and Sustainable Energy Reviews, vol. 64, Jun. 02, 2016.
[Online]. Available: https://doi.org/10.1016/j.rser.2016.06.002
[2] R. Tresa-Jacob and R. Liyanapathirana, “Technical feasibility in
reaching renewable energy targets; case study on australia,” in
4th International Conference on Electrical Energy Systems (ICEES),
Chennai, India, 2018, pp. 630–634.
[3] P. peña carro, O. Izquierdo-Monge, L. Hernández-Callejo,
and G. Martín-Jiménez, “Estudio e integración de pequeños
aerogeneradores en una microrred periurbana,” Revista Facultad
de Ingeniería, vol. 104, Jul. Sep. 2022. [Online]. Available:
https://doi.org/10.17533/udea.redin.20210845
[4] S. Nesmachnow, G. Colacurcio, D. G. Rossit, J. Toutouh, and
F. Luna, “Optimizing household energy planning in smart cities:
A multiobjective approach,” Revista de Ingeniería, vol. 101, Act.
Dic. 2021. [Online]. Available: https://doi.org/10.17533/udea.redin.
20200587
[5] L. Paoli and T. Gül. (2022, Nov. 30,) Electric cars fend off supply
challenges to more than double global sales, iea: International
energy agency. [Online]. Available: https://tinyurl.com/mr2c8393
[6] M. H. S. M. Haram, J. W. Lee, G. Ramasamy, E. E. Ngua,
S. P. Thiagarajaha, and et al., “Feasibility of utilising second life
ev batteries: Applications, lifespan, economics, environmental
impact, assessment, and challenges,” Alexandria Engineering
Journal, vol. 60, no. 5, Mar. 09, 2021. [Online]. Available: https:
//doi.org/10.1016/j.aej.2021.03.021
[7] E. Hossain, D. Murtaugh, J. Mody, H. M. R. Faruque,
M. S. Haque-Sunny, and et al., “A comprehensive review on
second-life batteries: Current state, manufacturing considerations,
applications, impacts, barriers & potential solutions, business
strategies, and policies,” IEEE Access, vol. 7, May. 20, 2019. [Online].
Available: http://doi.org/10.1109/ACCESS.2019.2917859.
[8] E. Martinez-Laserna, I. Gandiaga, E. Sarasketa-Zabala, J. Badeda,
D. I. Stroee, and et al., “Battery second life: Hype, hope or
reality? a critical review of the state of the art,” Renewable and
Sustainable Energy Reviews, vol. 93, Apr. 14, 2018. [Online]. Available:
https://doi.org/10.1016/j.rser.2018.04.035
[9] J. Neubauer and A. Pesaran, “The ability of battery second use
strategies to impact plug-in electric vehicle prices and serve
utility energy storage applications,” Journal of Power Sources,
vol. 196, no. 23, Jun. 15, 2011. [Online]. Available: https:
//doi.org/10.1016/j.jpowsour.2011.06.053
[10] Q. Liao, M. Mu, S. Zhao, L. Zhang, T. Jiang, and et al., “Performance
assessment and classification of retired lithium ion battery from
electric vehicles for energy storage,” International Journal of
Hydrogen Energy, vol. 42, no. 30, Jun. 06, 2017. [Online]. Available:
https://doi.org/10.1016/j.ijhydene.2017.06.043
[11] V. V. Viswanathan and M. Kintner-Meyer, “Second use of
transportation batteries: Maximizing the value of batteries for
transportation and grid service,” IEEE Transactions on Vehicular
Technology, vol. 60, no. 7, Jun. 23, 2011. [Online]. Available:
http://doi.org/10.1109/TVT.2011.2160378
[12] R. Reinhardt, I. Christodoulou, S. Gassó-Domingo, and
B. Amante-García, “Towards sustainable business models for
electric vehicle battery second use: A critical review,” Journal
of Environmental Management, vol. 245, May. 23, 2019. [Online].
Available: https://doi.org/10.1016/j.jenvman.2019.05.095l
[13] C. S. Ioakimidis, A. Murillo-Marrodán, A. Bagheri, D. Thomas,
and K. N. Genikomsakis, “Life cycle assessment of a lithium iron
phosphate (lfp) electric vehicle battery in second life application
scenarios,” Sustainability, vol. 11, no. 9, Apr. 28, 2019. [Online].
Available: https://doi.org/10.3390/su11092527
[14] Y. Jiang, J. Jiang, C. Zhang, W. Zhang, Y. Gao, and et al., “State of
health estimation of second-life lifepo4 batteries for energy storage
applications,” Journal of Cleaner Production, vol. 205, Sep. 17, 2008.
[Online]. Available: https://doi.org/10.1016/j.jclepro.2018.09.149
[15] F. Salek, S. Resalati, D. Morrey, P. Henshall, and A. Azizi, “Technical
energy assessment and sizing of a second life battery energy
storage system for a residential building equipped with ev charging
station,” Applied Sciences, vol. 12, no. 21, Oct. 31, 2022. [Online].
Available: https://doi.org/10.3390/app122111103
[16] H. C. Hesse, M. Schimpe, D. Kucevic, and A. Jossen, “Lithium-ion
battery storage for the grid—a review of stationary battery storage
system design tailored for applications in modern power grids,”
Energies, vol. 10, no. 12, Dec. 01, 2017. [Online]. Available:
https://doi.org/10.3390/en10122107
[17] P. Cicconi, D. Landi, A. Morbidoni, and M. Germani, “Feasibility
analysis of second life applications for li-ion cells used in
electric powertrain using environmental indicators,” in 2012
IEEE International Energy Conference and Exhibition (ENERGYCON),
21
power supplies, close to 30 kW (just a few minutes).
However, it can work with lower power supplies (around
5 kW) for several hours, which is very useful in the
energy management of the microgrid. These allow the
storage of energy surplus from distributed generation
systems (wind turbines and photovoltaic systems) and use
that energy when needed, improving microgrid’s efficiency.
Future work would be to recover the second rack by
using the remaining cells and buying new ones to
complete the total of 196 cells. This would double the
storage capacity, in addition to improving the performance
of the entire system by introducing new cells.
6. Declaration of competing interest
We declare that we have no significant competing interests,
including financial or non-financial, professional, or
personal interests interfering with the full and objective
presentation of the work described in this manuscript.
7. Funding
This work was supported by CEDER-CIEMAT.
8. Author contributions
Conceptualization, O.I-M.; Methodology, O.I-M., P.P-C.
and N.A.G.; Software, O.I-M.; writing—original draft
preparation O.I-M. N.A.G. and P.P-C.; writing.
9. Data availability statement
Data were collected at CEDER-CIEMAT from May to
September 2021. We used the BMS of the LFP
battery systems and Ingeteam inverter, connected to
CEDER´s both microgrid energy management control
system developed with Home Assistant to measure power
and voltage.
References
[1] M. H. Lee and D. Shang-Chang, “Allocative efficiency of high-power
li-ion batteries from automotive mode (am) to storage mode (sm),”
Renewable and Sustainable Energy Reviews, vol. 64, Jun. 02, 2016.
[Online]. Available: https://doi.org/10.1016/j.rser.2016.06.002
[2] R. Tresa-Jacob and R. Liyanapathirana, “Technical feasibility in
reaching renewable energy targets; case study on australia,” in
4th International Conference on Electrical Energy Systems (ICEES),
Chennai, India, 2018, pp. 630–634.
[3] P. peña carro, O. Izquierdo-Monge, L. Hernández-Callejo,
and G. Martín-Jiménez, “Estudio e integración de pequeños
aerogeneradores en una microrred periurbana,” Revista Facultad
de Ingeniería, vol. 104, Jul. Sep. 2022. [Online]. Available:
https://doi.org/10.17533/udea.redin.20210845
[4] S. Nesmachnow, G. Colacurcio, D. G. Rossit, J. Toutouh, and
F. Luna, “Optimizing household energy planning in smart cities:
A multiobjective approach,” Revista de Ingeniería, vol. 101, Act.
Dic. 2021. [Online]. Available: https://doi.org/10.17533/udea.redin.
20200587
[5] L. Paoli and T. Gül. (2022, Nov. 30,) Electric cars fend off supply
challenges to more than double global sales, iea: International
energy agency. [Online]. Available: https://tinyurl.com/mr2c8393
[6] M. H. S. M. Haram, J. W. Lee, G. Ramasamy, E. E. Ngua,
S. P. Thiagarajaha, and et al., “Feasibility of utilising second life
ev batteries: Applications, lifespan, economics, environmental
impact, assessment, and challenges,” Alexandria Engineering
Journal, vol. 60, no. 5, Mar. 09, 2021. [Online]. Available: https:
//doi.org/10.1016/j.aej.2021.03.021
[7] E. Hossain, D. Murtaugh, J. Mody, H. M. R. Faruque,
M. S. Haque-Sunny, and et al., “A comprehensive review on
second-life batteries: Current state, manufacturing considerations,
applications, impacts, barriers & potential solutions, business
strategies, and policies,” IEEE Access, vol. 7, May. 20, 2019. [Online].
Available: http://doi.org/10.1109/ACCESS.2019.2917859.
[8] E. Martinez-Laserna, I. Gandiaga, E. Sarasketa-Zabala, J. Badeda,
D. I. Stroee, and et al., “Battery second life: Hype, hope or
reality? a critical review of the state of the art,” Renewable and
Sustainable Energy Reviews, vol. 93, Apr. 14, 2018. [Online]. Available:
https://doi.org/10.1016/j.rser.2018.04.035
[9] J. Neubauer and A. Pesaran, “The ability of battery second use
strategies to impact plug-in electric vehicle prices and serve
utility energy storage applications,” Journal of Power Sources,
vol. 196, no. 23, Jun. 15, 2011. [Online]. Available: https:
//doi.org/10.1016/j.jpowsour.2011.06.053
[10] Q. Liao, M. Mu, S. Zhao, L. Zhang, T. Jiang, and et al., “Performance
assessment and classification of retired lithium ion battery from
electric vehicles for energy storage,” International Journal of
Hydrogen Energy, vol. 42, no. 30, Jun. 06, 2017. [Online]. Available:
https://doi.org/10.1016/j.ijhydene.2017.06.043
[11] V. V. Viswanathan and M. Kintner-Meyer, “Second use of
transportation batteries: Maximizing the value of batteries for
transportation and grid service,” IEEE Transactions on Vehicular
Technology, vol. 60, no. 7, Jun. 23, 2011. [Online]. Available:
http://doi.org/10.1109/TVT.2011.2160378
[12] R. Reinhardt, I. Christodoulou, S. Gassó-Domingo, and
B. Amante-García, “Towards sustainable business models for
electric vehicle battery second use: A critical review,” Journal
of Environmental Management, vol. 245, May. 23, 2019. [Online].
Available: https://doi.org/10.1016/j.jenvman.2019.05.095l
[13] C. S. Ioakimidis, A. Murillo-Marrodán, A. Bagheri, D. Thomas,
and K. N. Genikomsakis, “Life cycle assessment of a lithium iron
phosphate (lfp) electric vehicle battery in second life application
scenarios,” Sustainability, vol. 11, no. 9, Apr. 28, 2019. [Online].
Available: https://doi.org/10.3390/su11092527
[14] Y. Jiang, J. Jiang, C. Zhang, W. Zhang, Y. Gao, and et al., “State of
health estimation of second-life lifepo4 batteries for energy storage
applications,” Journal of Cleaner Production, vol. 205, Sep. 17, 2008.
[Online]. Available: https://doi.org/10.1016/j.jclepro.2018.09.149
[15] F. Salek, S. Resalati, D. Morrey, P. Henshall, and A. Azizi, “Technical
energy assessment and sizing of a second life battery energy
storage system for a residential building equipped with ev charging
station,” Applied Sciences, vol. 12, no. 21, Oct. 31, 2022. [Online].
Available: https://doi.org/10.3390/app122111103
[16] H. C. Hesse, M. Schimpe, D. Kucevic, and A. Jossen, “Lithium-ion
battery storage for the grid—a review of stationary battery storage
system design tailored for applications in modern power grids,”
Energies, vol. 10, no. 12, Dec. 01, 2017. [Online]. Available:
https://doi.org/10.3390/en10122107
[17] P. Cicconi, D. Landi, A. Morbidoni, and M. Germani, “Feasibility
analysis of second life applications for li-ion cells used in
electric powertrain using environmental indicators,” in 2012
IEEE International Energy Conference and Exhibition (ENERGYCON),
21
O. Izquierdo-Monge et al., Revista Facultad de Ingeniería, Universidad de Antioquia, No. 110, pp. 9-22, 2024
Florence, Italy, 2012, pp. 985–990.
[18] H. A. Catherino and F. Feres, “Consequences of deep cycling
24 volt battery strings,” SAE International Journal of Alternative
Powertrains, vol. 4, no. 2, Jul. 2015. [Online]. Available: https:
//www.jstor.org/stable/26169097
[19] O. Izquierdo-Monge, P. Peña-Carro, R. Villafafila-Robles,
O. Duque-Perez, A. Zorita-Lamadrid, and et al., “Conversion
of a network section with loads, storage systems and
renewable generation sources into a smart microgrid,” Applied
Sciences, vol. 11, no. 11, May. 26, 2021. [Online]. Available:
https://doi.org/10.3390/app11115012
22
Florence, Italy, 2012, pp. 985–990.
[18] H. A. Catherino and F. Feres, “Consequences of deep cycling
24 volt battery strings,” SAE International Journal of Alternative
Powertrains, vol. 4, no. 2, Jul. 2015. [Online]. Available: https:
//www.jstor.org/stable/26169097
[19] O. Izquierdo-Monge, P. Peña-Carro, R. Villafafila-Robles,
O. Duque-Perez, A. Zorita-Lamadrid, and et al., “Conversion
of a network section with loads, storage systems and
renewable generation sources into a smart microgrid,” Applied
Sciences, vol. 11, no. 11, May. 26, 2021. [Online]. Available:
https://doi.org/10.3390/app11115012
22