Simulation Modelling Practice and Theory 119 (2022) 102590

Contents lists available at ScienceDirect

Simulation Modelling Practice and Theory

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

Improved model and simulation tool for dynamic SOH estimation

and life prediction of batteries used in PV systems
W. Merrouche a, b, *, M. Trari c, L. Djellal b, M. Mammeri d, e, H. Tebibel f, S. Blaifi g,
Lee Wai Chong h, S. Ould-amrouche a, B. Boussaha e
a
Division Energie Solaire Photovoltaïque, Centre de Développement des Energies Renouvelables, CDER, 16340 Bouzareah, Algiers, Algeria
b
Materials and components department, Faculty of Physics, USTHB, BP32, 16111, Algiers, Algeria
c
Laboratory of storage and valorisation of renewable energies L.S.V.E.R, Faculty of Chemistry, USTHB, BP32, 16111, Algiers, Algeria
d
Laboratoire d’étude Physico-Chimique des Matériaux (LEPCM), Elhadj-Lakhdar université de Batna 1, Batna, Algeria
e
Faculté des sciences, Université d’Alger 1, 16000, Algiers, Algeria
f
Division Hydrogène Energies Renouvelables, Centre de Développement des Energies Renouvelables, CDER, 16340 Bouzareah, Algiers, Algeria
g
Research Laboratory of Electrical Engineering and automatics LREA, University of Médéa, Algeria
h
Lee Kong Chian Faculty of Engineering and Science, Universiti Tunku Abdul Rahman, Jalan Sungai Long, 43000 Selangor, Malaysia

A R T I C L E I N F O A B S T R A C T

Keywords: Accurate estimation of the battery state of health (SOH) is necessary for effective monitoring and
State of health prediction of battery performances and useful life in PV systems. This paper proposes an
Battery working zone improvement of a previous SOH model to reflect the battery aging state by monitoring the
Temperature
working zone and temperature. It presents an important contribution for simple, easy-to-
Lifetime estimation
implement, dynamic and non-destructive estimation of battery SOH. Unlike the former model,
Photovoltaics
Software tool the proposed model overcomes the problems associated with the inaccurate SOH estimations
found in the aged state of the battery. This is achieved by completely modifying the SOH equation
and including several important battery characteristics not considered in the traditional model –
mainly by introducing the battery design life parameter in the calculation of the safe working
zone factor. These enhancements allow the user to adjust the new model to different solar battery
types simply from the datasheet information. Using real measurements, model parameters were
computed to reduce the estimation error of the battery SOH. To verify the accuracy of the pro­
posed model, an experimental rig that comprises a solar battery tested for a long time is set up.
The promising part of the method is the improvement shown in the SOH estimation results. The
mean error, in the aged state of the battery, is reduced from 90% to 5%. Furthermore, the pro­
posed model is integrated into a PV designer software for the prediction of battery lifetime and
SOH degradation. Using meteorological data, the software tool shows the ability to predict the
endurance of the solar batteries in a designed PV system for any location. It is expected that this
work will benefit a large number of BMS designers and simulator developers who require a
Battery SOH estimation method with a good compromise between simplicity and accuracy.

* Corresponding author.
E-mail address: [email protected] (W. Merrouche).

https://doi.org/10.1016/j.simpat.2022.102590
Received 29 June 2021; Received in revised form 18 April 2022; Accepted 27 May 2022
Available online 29 May 2022
1569-190X/© 2022 Elsevier B.V. All rights reserved.
W. Merrouche et al. Simulation Modelling Practice and Theory 119 (2022) 102590

1. Introduction

Accurate battery State of health (SOH) information is an efficient tool to give indication of expected battery performance and useful

Nomenclature

Abbreviations
Ah Ampere-hour
DOD Depth-of-discharge
EV Electrical Vehicle
EOL End of Life
LAB Lead Acid Battery
PV Photovoltaic
SOC State-of-Charge
SOH State-of-Health
WZTT Working Zone and Temperature Tracking

Variables
C Capacity
Cmes Measured capacity
Cn Nominal capacity
I Current
nT Temperature health factor
nwz Working zone health factor
SOH0 Initial state of health
SOHexp Experimental SOH
T Temperature
t Time
Tref Reference temperature
V Voltage
Vex Exhaustion threshold voltage
Vg Overcharge threshold voltage
Vn Nominal threshold voltage
Vod Over discharge threshold voltage
Vs Saturation threshold voltage
α Rise in temperature that results in battery SOH reducing by half

life, and the remaining time before the next replacement [1]. Therefore, integrating SOH estimation feature in the new Battery
Management Systems (BMS) and Software Tools has become a hot research topic [2–8].
There are three main indicators of battery SOH: (1) Capacity, the capability to store energy; (2) Internal resistance, the ability to
deliver current; and (3) Self-discharge, revealing mechanical integrity and stress events. Compared with other indicators, measuring
the capacity by discharging a fully charged battery is a leading SOH indicator [9]. Checking battery capacity as a direct method for
SOH estimation is easy-to-implement and provides the most reliable assessment. However, applying a full discharge cycle is
impractical for most electrical systems, requires long time and causes stress [6]. Electric Vehicles (EVs) and Photovoltaic Systems
(PVSs) are the two examples of modern technologies characterized by infrequent battery full charge/discharge cycles or any scheduled
capacity tests, imposed by their dynamic and complex operating conditions [10–13]. Hence it is difficult to directly measure the
battery capacity in the mentioned examples. That’s where particular SOH estimation methods come in, considering battery operation
range. [14–16].
The goal of the present work is to construct a battery SOH model easy to implement in BMS and simulation tools for PVSs. A simple
method is needed that can estimate the battery SOH with a good balance between simplicity and accuracy. The focus of this paper is on
developing a model for Lead Acid Batteries (LABs). However, for the sake of completeness and clarity, we refer to some articles that
cover SOH and life estimation in lithium-ion batteries.
There is a variety of battery models and approaches in the field of battery state and health estimation. SOH models range from the
simplest to implement, to the most complex requiring high computational effort [3,4,7,16,17].
Different complex methodologies for SOH estimation and life prediction through the use of model-based and data-driven based are
proposed in the literature such as in [18–25]. The more complex methods can cover almost all possible scenarios of battery life aging
and can achieve high estimation precision. These models are very useful for making technical and financial decisions [3,26]. The main
drawbacks are that these methods often require huge computational efforts due to algorithm complexity and usually need long
processing time. Only a limited number of articles which present these complex approaches have a practical implementation on

2
W. Merrouche et al. Simulation Modelling Practice and Theory 119 (2022) 102590

Fig. 1. Comparison of some of the most existing methods in terms of computational efforts and required measurements [27,28].

embedded targets [3,8].

The simple models link the battery SOH to some parameters which can be easily determined such as Ah-throughput or time used up
during operation. Once a predetermined value of the SOH has been reached, the battery is considered in its End-of-Life (EOL).
Fig. 1 shows a comparison of some of the most existing techniques in literature in terms of computational complexity and required
measurements [27,28]. It is found that simple techniques like Open Circuit Voltage (OCV) or Coulomb Counting methods, require
much less measurements and computational resources than complex model-based approaches, including filter-based and
observer-based methods.
According to the accuracy of the estimation, it is found that, although the complex approaches generally present a lower estimation
error, there are also several relatively simple approaches that give acceptable accuracy [3].
The existing commercial solutions and planning tools usually implement simple methods (such as Coulomb counting) and offer a
wide range of configurability. Indeed, a simpler model with acceptable accuracy is relatively convenient to implement, compared to
more theoretical approaches [3,26,28–30]
The ampere-hour (Ah) integral or Coulomb counting method [31,32] is a tracking method considered in the most used technology
of the moment and implemented in several commercial solutions for SOH and SOC estimation [5,28,33–35]. This is due to its simplicity
and its low computational cost. Compared with other simple methods, the Ah integral method is more straightforward, where the SOH
of the battery is calculated by tracking the charged or discharged Ah using current integration. It has also the pros of being low affected
by some parameters like DOD, C-rate or temperature, which usually have a strong effect on other approaches. This method was
combined with other methods to give a promising approach for better SOH estimation [36–41]. Nevertheless, this approach also counts
with some drawbacks. Its open-loop Ah counting requires sometimes a high storing capacity and high time-consuming. Furthermore,
its accuracy is affected by the accumulative error of the continuous Coulomb counting when the battery capacity is not calibrated [7,
28,41].
Other way of directly estimating the SOH is by determining the actual value of the battery impedance. This method is based on the
principle that battery impedance increases with aging and that degradation mechanisms tend to affect different frequency ranges of the
complex impedance spectrum [28]. In fact, some works define the SOH as the ratio of the actual impedance compared to the
impedance of a healthy battery [42]. Monitoring the internal resistance of the battery through Direct Current (DC) is a simple and fast
method suitable for online applications. However, that resistance method is inherently not very accurate in operating conditions
because it is a one-parameter model affected by the battery SOC and the fast changes in voltage as well as other dynamic battery
parameters [43]. The accuracy may be improved by tracking the battery impedance spectrum instead of the simple internal resistance.
Electrochemical Impedance Spectroscopy (EIS) technique is widely used as an impedance spectrum diagnostic tool to evaluate battery
SOH [44,45]. Nevertheless, EIS methods usually require long testing times, complicated and expensive devices, and sometimes al­
gorithms with high-computational efforts, making them difficult to implement in online and on-board SOH estimation systems. To face
this problem, several recent works have been conducted using simple and low-cost impedance techniques for lithium-ion battery SOH
estimation addressed to embedded applications [45–48]. Among the key factors is the application of “broadband” current signal
excitation such as Pseudo-random sequence (PRS) signal characterized by its low complexity, measurement time, and good accuracy.
Although the above EIS methods provided a promising solution for SOH estimation in lithium-ion batteries, the results were almost
discouraging in the field of lead-acid batteries [43,45]. In fact, the aforementioned studies demand some requirements, such as long
rest time, specific SOC or a fixed current, which cannot be controlled in practical applications.

3
W. Merrouche et al. Simulation Modelling Practice and Theory 119 (2022) 102590

Table 1
comparison of various most existing techniques in terms of accuracy, computational complexity, implementation, strengths and drawbacks.[3,4,7,17,
49,50].
< Accuracy Strengths Drawbacks

- Kalman filter (KF) Very High 95-98% - Accurate - High computational complexity
- Can filter high degree of noise
- Observer High 90-98% - Possible for all chemistries - Only for mature technologies
- Faster than KF - Need powerful controllers
- Heavy calculation burden
- Impedance High 95-97% - Avoids complex calculation when using the - Offline data processing
standalone method - Requires specific current patterns
- Vulnerable to temperature
- High time consumption
- Ah Counting Relatively High - Simplicity - Calibration is required after charge/discharge
90-98% - Fast cycle
- Low computational complexity - Depends on test equipment accuracy
- Easy to implement - Sensitive to initial state
- Low power consumption - When combined with other methods the main
- Online application advantages are gone
- High accuracy if combined with other
methods
- Open Circuit Voltage High 95% - Easy to implement - Cannot operate online
(OCV) - Depends on test equipment accuracy
- Proposed WZTT Relatively High - Simplicity - Calibration is required when new battery is
90-95% - Fast connected
- Low computational complexity - Need more tests
- Easy to implement - Depends on test equipment accuracy
- Low power consumption - Sensitive to initial state
- Acceptable accuracy
- Online application

Most existing techniques are compared in Table 1 from the following aspects: accuracy, computational complexity, implementa­
tion, strengths and drawbacks. [3,4,7,17,49,50].
In the light of the above discussion, and after careful literature search, we found an interesting simple tracking technique worthy to
be studied and enhanced: battery Working Zone and Temperature Tracking (WZTT) method. The mentioned technique was implicitly
used by the CIEMAT model that gives a good dynamic performance of batteries in PVS as it considers the battery working range in
charge and discharge [51–55]. In this model, Copetti et al. [56–59] enhanced the initial Shepherd model for simulation and opti­
mization of ideal PVS. However, their model does not consider the battery ageing and SOH degradation. Guasch and Silvestre [60]
presented an improvement of CIEMAT model. In addition to the automatic extraction of 11 parameters using the Levenberg–Marquardt
algorithm, the authors included a simple SOH method using the WZTT technique to introduce the capacity reduction and the increase
of the self-discharge current caused by the degradation mechanisms. Although the proposed SOH model shows a significant simplicity
and ease of implementation for non-destructive and online SOH estimation, important battery parameters were not considered.
Therefore, inaccurate SOH estimation was found in aged state of the battery as illustrated in the experimental part of our work (see
Section 3).
In previous papers, we have improved the CIEMAT model by automatic parameter extraction based on a genetic algorithm to give
minimized error and good agreement with real measurements [58,59]. Furthermore, a LabVIEW-based real-time interface system
using the enhanced model was proposed for the online estimation and measurements of dynamic battery data in PV systems [59].
In this work, we enhanced the battery SOH model presented in [60] while taking advantage of the WZTT technique for simple,
easy-to-implement, non-destructive, and online estimation of battery SOH. The main original contributions of this paper are sum­
marized as follows:

- The former equation was drastically modified and new parameters were introduced to improve the model accuracy to an acceptable
value (95%).
- The calibration is supported by the enhanced model to correct tracking inaccuracies.
- The design life parameter was added to make the model suitable for a broader range of solar batteries.
- The enhanced SOH model was implemented in PC software.

The proposed approach in this paper is compared with most existing techniques in Table 1.
The main idea of this method can be summarized into three steps. Firstly, collect several kinds of data generated by the battery
during operation, such as voltage, current and temperature. Then make an aging speed coefficient look-up table for each kind of data
and calculate the aging speed coefficient at every moment. Finally, integrate the aging speed coefficient and calculate the aging state of
the battery. This method is very easy to understand and implement.

4
W. Merrouche et al. Simulation Modelling Practice and Theory 119 (2022) 102590

Experimental charge/discharge tests, which emulate the typical charge/discharge processes in Stand-Alone PV systems (SAPVS),
were carried out using a solar LAB. The new model was validated using experimental data, and the results were verified by comparing
them with the experimental SOH trend. The results show that the enhanced battery SOH model is able to estimate battery SOH with
acceptable accuracy (95%). The proposed model was implemented in a software tool to estimate the battery lifetime and SOH for
SAPVS applications using real meteorological data.
This case study is limited to the enhancement of a previous model using LAB at low currents and medium temperatures. This
research is expected to serve as a base for future studies, comparisons and improvement for the purpose of the generalization of the
proposed model for lithium-Ion batteries with wider operating conditions.

2. MATERIALS AND METHODS

The different battery operating conditions in PV applications (e.g., cycling, overcharge, deep discharge) affect the health of the
battery. The SOH is an internal parameter that presents the current health of the battery by comparing the actual capacity with the
capacity of a healthy battery. A fully charging/discharging cycle is the best way to calculate the real battery capacity. However, in
dynamic applications such as PVS and EV, batteries are cycled at partial states of charge. Hence, the direct capacity cannot be
determined easily in such systems, as fully charging/discharging cycles rarely occur [61].
As known, a battery model involves establishing rules and formulas to directly calculate the selected output quantity. The initial
obtained or former model is not necessarily ultimate. We can make a study of a very simplistic model based on a fluent technic, and
then we add parameters that were not considered in the first model to try to bring this model closer to real behaviors.
This paper proposes an improvement of a previous simplistic SOH model developed by Guasch et. al [60]. The former model
presents an important contribution for easy-to-implement and non-destructive estimation of battery SOH by monitoring the Working
Zone and Temperature. However, when testing this model using the former parameters, inaccurate SOH estimation was found in aged
state of the battery. After studying the former model, the following limitations were found:

• The values of the aging speed coefficient related to the working zone (nwz ) are not variable according to the change in the design
lifetime of each battery. Obviously, the aging speed of a 2 years design battery lifetime must be higher than that of 10 years.
• The Rule of Thumb, derived from Arrhenius law [6,62–65], affirms that for every 10◦ C increase in temperature the LAB life is
halved (and decreases by a fifth for Li-ion battery). It follows that the overall slope of SOH rate is doubled with 10◦ C rise in
temperature. However, this proportionality and exponential temperature behavior are absent in the former model. This is due to
the use of addition operation between the health factors instead multiplication. In addition, the temperature health factor (nT )
should have an exponential form.
• The initial SOH value that corresponds to ( − ∞) is not defined.

2.1. Enhanced State of Health Model

In order to overcome the limitations discussed previously, the former model was enhanced and modified as follows:
∫t
SOH(t) = SOH0 − (nT nwz )dt (1)
0

The Eq. (1) describes the influence of health factors (aging speed coefficients): nT (s− 1 ) for temperature and nwz (s− 1 ) for working
zone. The two factors are multiplied and the degradation of battery SOH depends on the duration and depth of these factors. SOH is
equal to 100% at a healthy state of the battery and declines gradually with the aging of the battery.
SOH0 is the initial SOH value calculated just before the launch of the battery SOH program. In practical systems, the SOH0 is not
always 100%, but it depends on the initial measured capacity C0 . Add to that, many manufacturers tend to overrate their battery
capacities, knowing that the majority of users do not check the real values. Accordingly, the SOH0 is defined as the percentage of the
initial capacity (C0 ) relative to the battery nominal capacity (Cn ), given by the manufacturer.
C0
SOH0 = 100% (2)
Cn
As with the majority of tracking methods, however, errors accumulate over time because of random and intermittent discharges[6].
This can be solved by applying occasional calibration of the SOH model. The calibration can be made, during battery operation, by
revaluating the actual capacity when a full cycle occurs. In this case, the revaluated SOH is considered as SOH0 and the calculating
algorithm is then initialized. The full cycles provide the most accurate capacity readings and calibrate the SOH estimation to correct
tracking inaccuracies. The estimation error can be reduced to less than 1% at the operating cycle next to the calibration of the SOH.
When the SOH dwells above the user-set threshold that is typically 80%, a manual calibration involving a full discharge/charge
cycle is needed. If the measured capacity is less than 80%, a battery replacement is needed.
The elevated temperatures significantly affect the SOH of batteries. High temperatures accelerate the degradation of active ma­
terials and thereby hasten the loss of permanent capacity (shortening the lifespan). Actually, all batteries achieve optimum cycle life if
operated at 20◦ C or below [6].

5
W. Merrouche et al. Simulation Modelling Practice and Theory 119 (2022) 102590

Fig. 2. Working zones of the battery.

Table 2
Working zone thresholds of battery output voltage at 25◦ C.
Working zone State Zone conditions

Saturation zone Charge I > 0 Vbat > Vs; Vs=2.4V per Cell
Vs is the saturation threshold voltage
Overcharge zone Vg ≤Vbat ≤Vs; Vg=2.3V per Cell
Vg is the overcharge threshold voltage
Safe charge zone Vbat <Vg
Safe discharge zone Discharge I < 0 Vbat > Vod; Vod=1.9 V per Cell
Vod is the over discharge threshold voltage
Deep-discharge zone Vex ≤Vbat ≤Vod; Vex=1.8V per Cell
Vex is the exhaustion threshold voltage
Exhaustion zone Vbat< Vex

The temperature health factor (nT ) represents the SOH degradation related to the battery temperature. According to the Arrhenius
law, the temperature aging speed coefficient should be multiplied by the other degradation factors. The nT coefficient is defined as the
following exponential form:
T− Tref
nT = 2 ∝ (3)

where T is the measured temperature in◦ C, Tref is the reference temperature and α is the rise in temperature that results in the battery
SOH rate reducing by half. According to [6,62–65] α is equal to 10◦ C.
The parameter nwz is the factor of SOH degradation related to the battery working zone. As shown in Fig. 2, there are three working
zones that depend on the battery voltage, which are:

- Safe Zone (Zone 1 - green): it is the safe charging or discharging zone. It is situated between the overdischarge threshold voltage
(Vod) and the overcharge threshold voltage (Vg). The working zone factor is equal to low aging speed coefficient nwz1 .
- Dangerous Zone (Zone 2 - orange): it is the overcharge zone and the overdischarge zone. In overcharge zone, it is situated between
the overcharge threshold voltage (Vg) and the saturation threshold voltage (Vs). In overdischarge zone, it is situated between the
overdischarge threshold voltage (Vod) and the exhaustion threshold voltage (Vex). The working zone factor is equal to nwz2 .
Considering the dangerous effect of overcharging and overdischarging processes on the battery health, nwz2 must be greater than
nwz1 .
- Exhaustion Zone (Zone 3 - red): it is the saturation zone (Where the battery charging voltage is higher than Vs) and the exhaustion
zone (Where the battery discharging voltage is lower than Vex). The working zone factor is equal to nwz3 . The parameter nwz3 must
be greater than nwz2 as the saturation and exhaustion zones are the most dangerous zones and markedly affect the battery SOH.

6
W. Merrouche et al. Simulation Modelling Practice and Theory 119 (2022) 102590

Table 3
Enhanced working zone parameter nwz .
Working zone nwz (s− 1 )

Zone1: Safe charge or discharge zone 6.34 × 10− 9

nwz1 =
ld
Zone2: Overcharge or deep-discharge zone nwz2 = x1 × nwz1
Zone3: Saturation or exhaustion zone nwz3 = x2 × nwz1

Fig. 3. Diagram of the test bench.

SOH degradation is accelerated in zones 2 and 3 because of the gravity of degradation mechanisms, particularly sulfation in
discharge and corrosion in charge of LAB. For an in-depth understanding of the impact of aging mechanisms on the battery working
zone, the reader is referred to the references [58,66–72].
The working zone thresholds of the battery output voltage are listed in Table 2 [56,57]. Table 3 shows the modified values of the
working zone parameter nwz in the three working zones. The nwz1 base equation is the slope of SOH degradation from 100% to 80%
during the design lifetime ld given by the battery manufacturer. Usually, manufacturers give the battery lifetime in years, for safe
working zones at the reference temperature Tref (20 or 25◦ C) to a specified SOH minimum threshold (commonly 80%) [6,73–75]. By
applying the “trial-and-error” method based on the experimental data, it is found that the model parameter nwz2 is nearly thirty times
(x1 = 30) greater than nwz1 , and nwz3 is twenty times greater than nwz2 and thus six hundred times nwz1 (x2 = 600).

3. Results and discussion

3.1. Experimental model validation

As mentioned previously, this case study is limited to the improvement of the estimation accuracy and parameters of the model
proposed in [60] based on the WZTT technique using LAB at low currents.
The method uses a tracking technique very easy to understand and implement. It can be summarized into three steps. Firstly, the
user includes the initial capacity into the tracking program that starts to collect voltage, current and temperature measurements
generated by the battery during operation. Then Table 3 is used as a look-up table to calculate the corresponding aging speed coef­
ficient nwz at every moment. At the same time, the aging speed coefficient nT is calculated using (Eqn 3). Finally, the aging speed
coefficients are integrated in the model equation (Eqn 1) and the new SOH of the battery is calculated.
Experimental charge-discharge tests were carried out on a LAB designed for PV applications. The battery was cycled under five
levels of charge /discharge currents at extreme voltage thresholds. The chosen variation of low currents aims to reproduce the change
of the battery charge/discharge rate in SAPVS [56,76]. Overcharge and deep discharge states of the battery have been achieved in
order to cover the various expected working zones.
Fig. 3 illustrates the battery test bench used for the cycling tests. It consists of a battery testing system, data logger, computer,
thermocouple, and 12V/100Ah solar LAB. The battery testing system is a reversible converter BENNING - DNF2-72/5 that operates in
two modes: charging and discharging. It is equipped with digital display and indicators for voltage and current. The voltage can be
regulated between 1 and 72V, and the current between 1 and 150A. The battery testing system charged and discharged the LAB

7
W. Merrouche et al. Simulation Modelling Practice and Theory 119 (2022) 102590

Table 4
Battery characteristics presented by the manufacturer.
Nominal Voltage (V) Nominal Capacity (AH) 1.8V/Cell- 25◦ C Number of Cells Design Lifetime (years) 25◦ C

12 V C10=100AH; I10=10A 6 3 years

Fig. 4. Variation of the Charge/Discharge Current.

Fig. 5. Variation of (a) battery voltage and thresholds during the tests and (b) corresponding working zone parameter nwz .

according to the planned cycling procedure and measured the battery voltage (Vbat) and current (I). The Datalogger is conFig.d to
collect the battery voltage, current and temperature data and save them on the host PC. The characteristics of the tested battery are
listed in Table 4.
The tests were carried out under charge/discharge currents of 5, 10, 15, 20, and 25A including the rest periods as shown in Fig. 4.
Fig. 5(a) shows the measured battery voltage per cell and the covered working zones. The voltage thresholds are compensated by
multiplying by the temperature compensation factor of -5mV◦ C− 1 per cell [56]. The measured temperatures are illustrated in Fig. 6(a).
The measured voltage, current and temperature are the model inputs used to track the variation of the working zone parameter nwz
(Fig. 5(b)) and the temperature parameter nT (Fig. 6(b)). The model parameters related to the tested battery are shown in Table 5.
The capacity was measured in every full charge/discharge cycle to calculate the related experimental battery State of Health
SOHexp . This latter is defined as the ratio between the measured capacity Cmes and the nominal capacity Cn :

Cmes (t)
SOHexp (t) = (4)
Cn
The battery was cycled until it achieved sufficient SOH degradation. Fig. 7 shows the comparison between simulated and exper­
imental SOHs. The markers indicate the data-points of the SOHexp and the solid blue line indicates the trend of the SOHexp maximum
values. The former SOH model (in yellow) shows a quick battery degradation as compared with the real SOH trend (experimental). It

8
W. Merrouche et al. Simulation Modelling Practice and Theory 119 (2022) 102590

Fig. 6. Variation of (a) temperature during the tests and (b) corresponding temperature parameter nT .

Table 5
Parameters of the enhanced model.
parameter value Comment

SOH0 1.2 The initial measured capacity is C0(10A)=120Ah

nwz1 2.11 × 10− 9 The battery lifetime is 3 years
nwz2 6.34 × 10− 8 x1 = 30
nwz3 1.27 × 10− 6 x2 = 600
Tref 25◦ C
α 10∘ C

estimates that the battery is out of service in a few days! but in fact, the actual state of the battery remains healthy. The improved
parameters of the new model overcome this issue, and the enhanced model simulation (solid sky blue line) presents a good match to the
experimental SOHexp even in the aged state of the battery. Fig. 7 illustrates also the expected trend of the battery SOH only in one
working zone namely in zone 1 (safe), zone 2 (dangerous), and zone 3 (exhaustion). These trends demonstrate that the working of the
battery in the safe zone is the best for its SOH preservation. However, when the battery is supposed to operate only in the exhaustion
zone (zone 3), the SOH is zero within a few days. The expectation for battery working in the dangerous zone (zone 2) is almost the same
as the experimental trend. This means that the general operating of the tested battery is around zone 2. We also notice that the expected
trend of the SOH without including the temperature effect is not far from the experimental trend. This is because the average test
temperatures are around the reference temperature.
As depicted in Fig. 8, the error and the mean error of the enhanced model are lower as compared with those of the former model. In
the aged state of the battery (60% of SOH), the mean error of the enhanced model is less than 7% compared with over 90% for the
former model. At 80% of SOH, the common threshold of battery EOL, the error of the improved model is less than 5% (4.07%) which
means accuracy of more than 95%. This result confirms the significant improvement and reliability of the new SOH model. The SOH
estimation error can be significantly reduced if the SOH algorithm is frequently calibrated by reevaluating the actual capacity of the
battery.

3.2. Implementation of the SOH model in a PC software

A PC software was developed to predict the SOH and lifespan of solar batteries using the weather data. It is based on the enhanced
WZTT model. It allows the user to have a prediction on the endurance of the battery in a selected PV system at any location on Earth.
Additionally, this software has the ability to process multiple use cases in different periods through the year, which allows the program
to work on every situation. It uses a quick and simple procedure (Figs. 9–11).
The main steps for achieving an estimation of the expected battery lifetime and SOH degradation are as follows:

- Specify the battery parameters (Capacity, design life, reference temperature and sizing).
- Set the charging mode of the solar controller.

9
W. Merrouche et al. Simulation Modelling Practice and Theory 119 (2022) 102590

Fig. 7. Simulated SOH compared with the experimental SOH.

Fig. 8. Matching error and mean error between simulated and experimental SOH.

- Call weather data by location name, postcode or geographic coordinates.

- Set the thermal conditions of the battery room.
- Select the Battery working time (annual, seasonal, weekly).

Using the programmed specifications, the software provides the expected battery lifetime and SOH declination with age. The SOH
degradation model is directly linked to the detection of the battery EOL. Once a predetermined SOH threshold is exceeded, the battery

10
W. Merrouche et al. Simulation Modelling Practice and Theory 119 (2022) 102590

Fig. 9. First software window – battery/controller parameters.

Fig. 10. Second software window – site and battery working temperature.

11
W. Merrouche et al. Simulation Modelling Practice and Theory 119 (2022) 102590

Fig. 11. Third software window – battery working time.

Table 6
Parameters and conditions of the tested battery.
parameter value

Battery capacity 100Ah

Initial capacity 100Ah
Battery design life (25◦ C) 8years
Site Algiers
Battery sizing Accurate
Working temperature conditions Ambient
Working time seasonal- summer

Fig. 12. Temperature variation for one year in the selected site.

12
W. Merrouche et al. Simulation Modelling Practice and Theory 119 (2022) 102590

Fig. 13. Battery aging in a PV system without solar controller: (a) SOH and expected lifetime, (b) Capacity degradation.

is considered to be at the EOL [26].The first software window (Fig. 9) is to fill the battery characteristics, the panel sizing and the
control type. The panel Sizing and Control type have a direct impact on the working zone of the battery.
The second window (Fig. 10) is designed to automatically call weather data based on the location of the PV system. This data is used
for calculations in the WZTT model. In this window, the user can fix the battery temperature in the case of a battery room with a
controlled temperature.
The third window (Fig. 11) allows the user to choose a precise period of battery operation in the PV system.
Table 6 shows an example of conditions and specifications used for the battery SOH prediction. From a recent weather call, Fig. 12
presents a one-year temperature variation in the selected location (Algiers). It is noticed that the average temperature is less than the
reference temperature (25◦ C). Fig. 13, Fig. 14 and Fig. 15 illustrate the expected battery SOH, capacity degradation and lifetime at
three variants of PV systems.
Fig. 13 shows the SOH and the capacity degradation of the battery at the ambient temperature without a charge controller.
Compared with the design life (8 years), the expected battery life is decreased to 6 years and 4 months. In fact, without a charge
controller the battery is not protected from dangerous working zones.
On the other hand, Fig. 14 shows the SOH and the capacity degradation of the battery at the ambient temperature using a charge
controller. The expected life is extended to 9 years and 9 months. The use of a charge controller decelerates the battery aging and thus
prolongs the battery lifetime. This is because it protects the battery from dangerous working zones. As mentioned previously, the
average ambient temperature in the selected site is slightly less than the reference temperature 25◦ C (Fig. 12), thus the expected
lifespan is higher than the design life given in the reference temperature (Arrhenius law).
Fig. 15 shows the battery SOH and capacity degradation using a charge controller but at a fixed room temperature (25◦ C). In this
case, the battery will work in the safe zone and at the reference temperature. Therefore, the estimated lifetime will be close to the

13
W. Merrouche et al. Simulation Modelling Practice and Theory 119 (2022) 102590

Fig. 14. Battery aging in a PV system with solar controller: (a) SOH and expected lifetime, (b) Capacity degradation.

design life.

4. Conclusion

This paper presents an enhancement and modification of a previous SOH model, based on battery Working Zone and Temperature
Tracking WZTT, supported by experimental validation. The enhanced model resolves many problems seen in the former model by
taking into account the manufacturer design lifetime, the Arrhenius law, and the determination of the initial SOH. The enhanced SOH
model is validated using experimental data and also compared with the former SOH model. The improved SOH model presents a good
match with the experimental SOH trend and shows better accuracy as compared to the former model. The small mean error, less than
5%, enables the model to generally predict the battery SOH pattern in real operating conditions. The main advantage of this model is
that it is easy to be implemented as it presents a simple and non-destructive method to track the battery health degradation throughout
its lifetime.
The model was used and tested in a software tool developed for the life prediction of solar batteries in PV systems using meteo­
rological data. The software user has the ability to forecast the endurance of his batteries in a designed PV system at any location. More
research is necessary to investigate either further modification/extension of the model or a combination with other methods to
improve the SOH estimation of a wider range of battery technologies and current rates. It is expected that this work will benefit a large
number of simulator developers and BMS designers who require an accepted method for battery SOH estimation with a good
compromise between simplicity and accuracy.

14
W. Merrouche et al. Simulation Modelling Practice and Theory 119 (2022) 102590

Fig. 15. Battery aging at fixed temperature (25◦ C) in a PV system with solar controller: (a) SOH and expected lifetime, (b) Capacity degradation.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to
influence the work reported in this paper.

Acknowledgments

The authors would like to thank the Direction Générale de la Recherche Scientifique et du Développement Technologique
(DGRSDT) and the Centre de Développement des Energies Renouvelables (CDER) for providing financial assistance under Socio-
economic impact project No: DGRSDT/ CDER/178/2017.

References

[1] N. Beganovic, D. Söffker, Estimation of remaining useful lifetime of lithium-ion battery based on acoustic emission measurements, J. Energy Res. Technol.
(2019) 141, https://doi.org/10.1115/1.4042234.
[2] Deutsche Gesellschaft für Sonnenenergie (DGS), Planning and Installing Photovoltaic Systems, 3rd Ed., Routledge, London, 2013 https://doi.org/10.4324/
9781849776998.

15
W. Merrouche et al. Simulation Modelling Practice and Theory 119 (2022) 102590

[3] L. Ungurean, G. Cârstoiu, M. v. Micea, V. Groza, Battery state of health estimation: a structured review of models, methods and commercial devices, Int. J.
Energy Res. 41 (2017) 151–181, https://doi.org/10.1002/er.3598.
[4] J. Tian, R. Xiong, W. Shen, A review on state of health estimation for lithium ion batteries in photovoltaic systems, ETransportation 2 (2019), 100028, https://
doi.org/10.1016/j.etran.2019.100028.
[5] M.V. Micea, L. Ungurean, G.N. Cârstoiu, V. Groza, Online state-of-health assessment for battery management systems, IEEE Trans. Instrum. Meas. 60 (2011)
1997–2006, https://doi.org/10.1109/TIM.2011.2115630.
[6] I. Buchmann, Batteries in a portable world : a handbook on rechargeable batteries for non-engineers, 4th edition, Cadex Electronics Inc., Canada, 2017. https://
batteryuniversity.com/buy/ (accessed July 7, 2020).
[7] Y. Wang, J. Tian, Z. Sun, L. Wang, R. Xu, M. Li, Z. Chen, A comprehensive review of battery modeling and state estimation approaches for advanced battery
management systems, Renewable Sustainable Energy Rev. 131 (2020), 110015, https://doi.org/10.1016/j.rser.2020.110015.
[8] F. Díaz-González (Serra Húnter Fellow)., D. Heredero-Peris, S. Galceran-Arellano, Design methodology for a dc–dc power conversion system with EIS capability
for battery packs, Simul. Modell. Pract. Theory 87 (2018) 15–34, https://doi.org/10.1016/j.simpat.2018.05.008.
[9] A. Barré, B. Deguilhem, S. Grolleau, M. Gérard, F. Suard, D. Riu, A review on lithium-ion battery ageing mechanisms and estimations for automotive
applications, J. Power Sources 241 (2013) 680–689, https://doi.org/10.1016/j.jpowsour.2013.05.040. Review.
[10] E. Meissner, G. Richter, Battery monitoring and electrical energy management precondition for future vehicle electric power systems, J. Power Sources 116
(2003) 79–98, https://doi.org/10.1016/S0378-7753(02)00713-9.
[11] G. Piłatowicz, H. Budde-Meiwes, J. Kowal, C. Sarfert, E. Schoch, M. Königsmann, D.U. Sauer, Determination of the lead-acid battery’s dynamic response using
Butler-Volmer equation for advanced battery management systems in automotive applications, J. Power Sources 331 (2016) 348–359, https://doi.org/10.1016/
j.jpowsour.2016.09.066.
[12] L.W. Chong, Y.W. Wong, R.K. Rajkumar, R.K. Rajkumar, D. Isa, Hybrid energy storage systems and control strategies for stand-alone renewable energy power
systems, Renewable Sustainable Energy Rev. 66 (2016) 174–189, https://doi.org/10.1016/j.rser.2016.07.059.
[13] S.B. Sarmah, P. Kalita, A. Garg, X.D. Niu, X.W. Zhang, X. Peng, D. Bhattacharjee, A review of state of health estimation of energy storage systems: challenges and
possible solutions for futuristic applications of li-ion battery packs in electric vehicles, J. Electrochem. Energy Convers. Storage 16 (2019), https://doi.org/
10.1115/1.4042987.
[14] D.F. Specht, Probabilistic neural networks and the polynomial adaline as complementary techniques for classification, IEEE Trans. Neural Networks (1990),
https://doi.org/10.1109/72.80210.
[15] M. Danko, J. Adamec, M. Taraba, P. Drgona, Overview of batteries state of charge estimation methods. Transportation Research Procedia, Elsevier B.V., 2019,
pp. 186–192, https://doi.org/10.1016/j.trpro.2019.07.029.
[16] S. Yang, C. Zhang, J. Jiang, W. Zhang, L. Zhang, Y. Wang, Review on state-of-health of lithium-ion batteries: characterizations, estimations and applications,
J. Cleaner Prod. 314 (2021), 128015, https://doi.org/10.1016/j.jclepro.2021.128015.
[17] L. Tao, J. Ma, Y. Cheng, A. Noktehdan, J. Chong, C. Lu, A review of stochastic battery models and health management, Renewable Sustainable Energy Rev. 80
(2017) 716–732, https://doi.org/10.1016/j.rser.2017.05.127.
[18] T.T. De Sousa, V.T. Arioli, C.S. Vieira, S.R. Dos Santos, A.P. França, Comparison of different approaches for lead acid battery state of health estimation based on
artificial neural networks algorithms, in: Proceedings of the 2016 IEEE Conference on Evolving and Adaptive Intelligent Systems, EAIS 2016, Institute of
Electrical and Electronics Engineers Inc., 2016, pp. 79–84, https://doi.org/10.1109/EAIS.2016.7502495.
[19] H. Chaoui, S. Miah, A. Oukaour, H. Gualous, State-of-charge and state-of-health prediction of lead-acid batteries with genetic algorithms, in: 2015 IEEE
Transportation Electrification Conference and Expo, ITEC 2015 105, 2015, pp. 21–26, https://doi.org/10.1109/ITEC.2015.7165782.
[20] J. Lu, L. Wei, M.M. Pour, Y. Mekonnen, A.I. Sarwat, Modeling discharge characteristics for predicting battery remaining life, in: 2017 IEEE Transportation and
Electrification Conference and Expo, ITEC 2017, Institute of Electrical and Electronics Engineers Inc., 2017, pp. 468–473, https://doi.org/10.1109/
ITEC.2017.7993316.
[21] K. Khodadadi Sadabadi, P. Ramesh, P. Tulpule, Y. Guezennec, G. Rizzoni, Model-based state of health estimation of a lead-acid battery using step-response and
emulated in-situ vehicle data, J. Energy Storage 36 (2021), 102353, https://doi.org/10.1016/j.est.2021.102353.
[22] K. Khodadadi Sadabadi, P. Ramesh, P. Tulpule, G. Rizzoni, K.K. Sadabadi, P. Ramesh, P. Tulpule, G. Rizzoni, K. Khodadadi Sadabadi, P. Ramesh, P. Tulpule,
G. Rizzoni, Design and calibration of a semi-empirical model for capturing dominant aging mechanisms of a PbA battery, J. Energy Storage 24 (2019), 100789,
https://doi.org/10.1016/j.est.2019.100789.
[23] N. Wassiliadis, J. Adermann, A. Frericks, M. Pak, C. Reiter, B. Lohmann, M. Lienkamp, Revisiting the dual extended Kalman filter for battery state-of-charge and
state-of-health estimation: A use-case life cycle analysis, J. Energy Storage 19 (2018) 73–87, https://doi.org/10.1016/j.est.2018.07.006.
[24] B. Saha, K. Goebel, S. Poll, J. Christophersen, Prognostics methods for battery health monitoring using a Bayesian framework, IEEE Trans. Instrum. Meas. 58
(2009) 291–296, https://doi.org/10.1109/TIM.2008.2005965.
[25] S.B. Vilsen, D.I. Stroe, Battery state-of-health modelling by multiple linear regression, J. Cleaner Prod. 290 (2021), 125700, https://doi.org/10.1016/j.
jclepro.2020.125700.
[26] H. Wenzl, I. Baring-Gould, R. Kaiser, B.Y. Liaw, P. Lundsager, J. Manwell, A. Ruddell, V. Svoboda, Life prediction of batteries for selecting the technically most
suitable and cost effective battery, J. Power Sources 144 (2005) 373–384, https://doi.org/10.1016/j.jpowsour.2004.11.045.
[27] J. Remmlinger, M. Buchholz, K. Dietmayer, Methods for monitoring the state of batteries in automotive applications, in: Proc. VDE-Kongress E-Mobility, 2010.
[28] M. Berecibar, I. Gandiaga, I. Villarreal, N. Omar, J. van Mierlo, P. van den Bossche, Critical review of state of health estimation methods of Li-ion batteries for
real applications, Renewable Sustainable Energy Rev. 56 (2016) 572–587, https://doi.org/10.1016/j.rser.2015.11.042.
[29] T.M. Layadi, G. Champenois, M. Mostefai, D. Abbes, Lifetime estimation tool of lead–acid batteries for hybrid power sources design, Simulation Modell. Practice
Theory 54 (2015) 36–48, https://doi.org/10.1016/j.simpat.2015.03.001.
[30] K. Vignarooban, X. Chu, K. Chimatapu, P. Ganeshram, S. Pollat, N.G. Johnson, A. Razdan, D.S. Pelley, A.M. Kannan, State of health determination of sealed lead
acid batteries under various operating conditions, Sustain. Energy Technol. Assessments 18 (2016) 134–139, https://doi.org/10.1016/j.seta.2016.10.007.
[31] J. Schiffer, D.U. Sauer, H. Bindner, T. Cronin, P. Lundsager, R. Kaiser, Model prediction for ranking lead-acid batteries according to expected lifetime in
renewable energy systems and autonomous power-supply systems, J. Power Sources 168 (2007) 66–78, https://doi.org/10.1016/j.jpowsour.2006.11.092.
[32] D.U. Sauer, H. Wenzl, Comparison of different approaches for lifetime prediction of electrochemical systems—Using lead-acid batteries as example, J. Power
Sources 176 (2008) 534–546, https://doi.org/10.1016/j.jpowsour.2007.08.057.
[33] Texas Instruments, bq78412 Pb-Acid Battery State-of-Charge Indicator With Run-Time Display Check for Samples: bq78412 1FEATURES, 2010. www.ti.com.
accessed June 20, 2021.
[34] Texas Instruments, BQ78350-R1A BQ78350-R1A CEDV Li-Ion Gas Gauge and Battery Management Controller Companion to the BQ769x0 Battery Monitoring
AFE Simplified Schematic, 2019. www.ti.com. accessed June 21, 2021.
[35] Maxim Integrated, MAX17047-MAX17050 1-Cell Fuel Gauge with ModelGauge m3, 2016. www.maximintegrated.com. accessed June 21, 2021.
[36] I. Baccouche, S. Jemmali, A. Mlayah, B. Manai, N.E. ben Amara, Implementation of an improved Coulomb-counting algorithm based on a piecewise SOC-OCV
relationship for SOC estimation of Li-ion battery, Int. J. Renew. Energy Res. 8 (2018).
[37] A. Ganeshan, R. Shanmughasundaram, Estimation of soc and soh using mixed neural network and coulomb counting algorithm, Int. J. Innovative Technol.
Explor. Eng. 8 (2019), https://doi.org/10.35940/ijitee.J1308.0881019.
[38] K.S. Ng, C.S. Moo, Y.P. Chen, Y.C. Hsieh, Enhanced coulomb counting method for estimating state-of-charge and state-of-health of lithium-ion batteries, Appl.
Energy 86 (2009), https://doi.org/10.1016/j.apenergy.2008.11.021.
[39] A. Gismero, E. Schaltz, D.I. Stroe, Recursive state of charge and state of health estimation method for lithium-ion batteries based on coulomb counting and open
circuit voltage, Energies (Basel) (2020) 13, https://doi.org/10.3390/en13071811.
[40] S. Zhang, X. Guo, X. Dou, X. Zhang, A rapid online calculation method for state of health of lithium-ion battery based on coulomb counting method and
differential voltage analysis, J. Power Sources 479 (2020), https://doi.org/10.1016/j.jpowsour.2020.228740.

16
W. Merrouche et al. Simulation Modelling Practice and Theory 119 (2022) 102590

[41] J.S. Goud, R. Kalpana, B. Singh, An online method of estimating state of health of a Li-Ion Battery, IEEE Trans. Energy Convers. 36 (2021), https://doi.org/
10.1109/TEC.2020.3008937.
[42] A. Farmann, W. Waag, A. Marongiu, D.U. Sauer, Critical review of on-board capacity estimation techniques for lithium-ion batteries in electric and hybrid
electric vehicles, J. Power Sources 281 (2015) 114–130, https://doi.org/10.1016/J.JPOWSOUR.2015.01.129.
[43] S. Jiang, Z. Song, A review on the state of health estimation methods of lead-acid batteries, J. Power Sources 517 (2022), 230710, https://doi.org/10.1016/J.
JPOWSOUR.2021.230710.
[44] H. Blanke, O. Bohlen, S. Buller, R.W. de Doncker, B. Fricke, A. Hammouche, D. Linzen, M. Thele, D.U. Sauer, Impedance measurements on lead-acid batteries for
state-of-charge, state-of-health and cranking capability prognosis in electric and hybrid electric vehicles, J. Power Sources 144 (2005) 418–425, https://doi.org/
10.1016/J.JPOWSOUR.2004.10.028.
[45] K.M. Carthy, H. Gullapalli, K.M. Ryan, T. Kennedy, Review—use of impedance spectroscopy for the estimation of Li-ion battery state of charge, state of health
and internal temperature, J. Electrochem. Soc. 168 (2021), 080517, https://doi.org/10.1149/1945-7111/AC1A85.
[46] E. Locorotondo, V. Cultrera, L. Pugi, L. Berzi, M. Pierini, G. Lutzemberger, Development of a battery real-time state of health diagnosis based on fast impedance
measurements, J. Energy Storage 38 (2021), 102566, https://doi.org/10.1016/j.est.2021.102566.
[47] E. Locorotondo, F. Corti, L. Pugi, L. Berzi, A. Reatti, G. Lutzemberger, Design of a wireless charging system for online battery spectroscopy, Energies 2021 14
(2021) 218, https://doi.org/10.3390/EN14010218. Page14218.
[48] F. Díaz-González (Serra Húnter Fellow)., D. Heredero-Peris, S. Galceran-Arellano, Design methodology for a dc–dc power conversion system with EIS capability
for battery packs, Simul. Modell. Pract. Theory 87 (2018) 15–34, https://doi.org/10.1016/J.SIMPAT.2018.05.008.
[49] S. Park, J. Ahn, T. Kang, S. Park, Y. Kim, I. Cho, J. Kim, Review of state-of-the-art battery state estimation technologies for battery management systems of
stationary energy storage systems, J. Power Electron. 20 (2020), https://doi.org/10.1007/s43236-020-00122-7.
[50] M.S.H.S.H. Lipu, M.A.A. Hannan, A. Hussain, M.M.M. Hoque, P.J. Ker, M.H.M.H.M. Saad, A. Ayob, A review of state of health and remaining useful life
estimation methods for lithium-ion battery in electric vehicles: challenges and recommendations, J. Cleaner Prod. 205 (2018) 115–133, https://doi.org/
10.1016/j.jclepro.2018.09.065.
[51] W. Wang, W. Chen, W. Yao, D. Chen, Y. Zhang, Z. Lu, Mathematical analysis of dynamic safe operation area of very large capacity lead-acid battery, J. Energy
Storage 36 (2021), 102389, https://doi.org/10.1016/j.est.2021.102389.
[52] Z. Cabrane, M. Ouassaid, M. Maaroufi, Battery and supercapacitor for photovoltaic energy storage: a fuzzy logic management, IET Renew. Power Gener. 11
(2017) 1157–1165, https://doi.org/10.1049/iet-rpg.2016.0455.
[53] W. Wang, W. Yao, W. Chen, D. Chen, Z. Lu, Failure warning at the end of service-life of lead-acid batteries for backup applications, Appl. Sci. (Switzerland) 10
(2020) 5760, https://doi.org/10.3390/APP10175760.
[54] M. Zaibi, G. Champenois, X. Roboam, J. Belhadj, B. Sareni, Smart power management of a hybrid photovoltaic/wind stand-alone system coupling battery
storage and hydraulic network, Math. Comput. Simul 146 (2018) 210–228, https://doi.org/10.1016/j.matcom.2016.08.009.
[55] S. Blaifi, Energy Storage and Photovoltaic Systems. Advanced Structured Materials, Springer, 2020, pp. 139–164, https://doi.org/10.1007/978-3-030-43473-1_
8.
[56] J.B. Copetti, F. Chenlo, Lead/acid batteries for photovoltaic applications. Test results and modeling, J. Power Sources 47 (1994) 109–118, https://doi.org/
10.1016/0378-7753(94)80054-5.
[57] J.B. Copetti, E. Lorenzo, F. Chenlo, A general battery model for PV system simulation, Prog. Photovoltaics 1 (1993) 283–292, https://doi.org/10.1002/
pip.4670010405.
[58] S. Blaifi, S. Moulahoum, I. Colak, W. Merrouche, An enhanced dynamic model of battery using genetic algorithm suitable for photovoltaic applications, Appl.
Energy 169 (2016) 888–898, https://doi.org/10.1016/j.apenergy.2016.02.062.
[59] S. Blaifi, S. Moulahoum, I. Colak, W. Merrouche, Monitoring and enhanced dynamic modeling of battery by genetic algorithm using LabVIEW applied in
photovoltaic system, Electrical Eng. 100 (2018) 1021–1038, https://doi.org/10.1007/s00202-017-0567-6.
[60] D. Guasch, S. Silvestre, Dynamic battery model for photovoltaic applications, Prog. Photovoltaics Res. Appl. 11 (2003) 193–206, https://doi.org/10.1002/
pip.480.
[61] K.M. Tsang, W.L. Chan, State of health detection for Lithium ion batteries in photovoltaic system, Energy Convers. Manage. 65 (2013) 7–12, https://doi.org/
10.1016/j.enconman.2012.07.006.
[62] J. Bizouard, V. Schneider, M. Boost, A battery node for power plant monitoring systems, in: INTELEC - Twentieth International Telecommunications Energy
Conference (Cat. No.98CH36263, 1998, pp. 607–616, https://doi.org/10.1109/INTLEC.1998.793607.
[63] A. Jossen, J. Garche, D.U. Sauer, Operation conditions of batteries in PV applications, Sol. Energy 76 (2004) 759–769, https://doi.org/10.1016/j.
solener.2003.12.013.
[64] A.H. Anbuky, P.E. Pascoe, Apparatus, methods and computer program products for estimation of battery reserve life using adaptively modified state of health
indicator-based reserve life models, (2007). https://www.google.com/patents/US7199557%0Ahttp://patents.justia.com/patent/7199557.
[65] M. Ecker, J.B. Gerschler, J. Vogel, S. Käbitz, F. Hust, P. Dechent, D.U. Sauer, Development of a lifetime prediction model for lithium-ion batteries based on
extended accelerated aging test data, J. Power Sources 215 (2012) 248–257, https://doi.org/10.1016/j.jpowsour.2012.05.012.
[66] J.J. LANDER, Further studies on the anodic corrosion of lead in solutions, J. Electrochem. Soc. 103 (1956) 1–8, https://doi.org/10.1149/1.2430227.
[67] B. Culpin, D.A.J. Rand, Failure modes of lead/acid batteries, J. Power Sources 36 (1991) 415–438, https://doi.org/10.1016/0378-7753(91)80069-A.
[68] C. Armenta-Deu, T. Donaire, Determination of an ageing factor for lead/acid batteries. 1. Kinetic aspects, J. Power Sources 58 (1996) 123–133, https://doi.org/
10.1016/S0378-7753(96)02371-3.
[69] P. Ruetschi, Aging mechanisms and service life of lead-acid batteries, J. Power Sources (2004) 33–44, https://doi.org/10.1016/j.jpowsour.2003.09.052.
[70] C. Suozzo, Lead-Acid Battery Aging and State of Health Diagnosis, Thesis, 2008.
[71] D.U. Sauer, Optimization of the Use of Lead-Acid Batteries in Photovoltaic Hybrid Systems With Special Consideration of Battery Aging, University of Ulm,
2003, https://doi.org/10.18725/OPARU-1816.
[72] W. Merrouche, N. Achaibou, B. Bouzidi, M. Kasser, M. Trari, Lead-acid battery degradation mechanisms in photovoltaic systems PVS, in: The 3rd International
Workshop on Integration of Solar Power into Power Systems SIW2013, At London, UK, 2013.
[73] IEEE, in: IEEE Recommended Practice for Maintenance, Testing, and Replacement of Valve-Regulated Lead- Acid (VRLA) Batteries for Stationary Applications,
2006, https://doi.org/10.1109/IEEESTD.2014.6853300.
[74] Y. Mekonnen, H. Aburbu, A. Sarwat, Life cycle prediction of sealed lead acid batteries based on a Weibull model, J. Energy Storage 18 (2018) 467–475, https://
doi.org/10.1016/j.est.2018.06.005.
[75] M.R. Palacin, A. de Guibert, M.R. Palacín, A. de Guibert, Batteries: why do batteries fail? Science 351 (2016) (1979), 1253292 https://doi.org/10.1126/
science.1253292. –1253292.
[76] N. Achaibou, M. Haddadi, A. Malek, Lead acid batteries simulation including experimental validation, J. Power Sources 185 (2008) 1484–1491, https://doi.org/
10.1016/j.jpowsour.2008.06.059.

17