EP3970255A1

EP3970255A1 - Method for restoring capacity of lithium batteries

Info

Publication number: EP3970255A1
Application number: EP20742455.7A
Authority: EP
Inventors: Marco Venturini; Paolo PICCARDO; Roberto SPOTORNO; Marco SMERIERI
Original assignee: Phase Motion Control SpA
Current assignee: Phase Motion Control SpA
Priority date: 2019-05-15
Filing date: 2020-05-14
Publication date: 2022-03-23
Also published as: IT201900006839A1; WO2020230078A1

Abstract

An electrochemical process for restoring the capacity of one or more lithium batteries (4), comprising the following steps: a) discharging (11) the battery (4) until the cut- off voltage value of the battery (4) is reached; b) overdischarging (13) the battery (4) until a voltage value is reached wherein at least a partial destruction of the solid electrolyte interface occurs, this corresponding voltage value being substantially the minimum voltage value (22) reachable by the battery (4) during the overdischarge; c) charging (14) the battery (4) by applying a direct current until the maximum voltage value corresponding to the full charge is reached; d) applying one or more discharge and charge cycles (15).

Description

METHOD FOR RESTORING CAPACITY OF LITHIUM BATTERIES

The present invention relates to an electrochemical process for restoring the capacity of lithium batteries applicable to batteries in use to optimize the control of the overall capacity of the batteries themselves, reducing or eliminating the capacity drop relative to the number of operating cycles .

Lithium-ion batteries are an optimal solution for storing, transporting and distributing energy, whether primary or secondary, with widespread use in a wide variety of applications or wherever the application of electricity is required with high efficiency (e.g. electrical mobility, battery-powered tools, laptops and mobile battery-powered units) .

The application of frequent charging and discharging cycles for the continuous use of batteries, as well as fast charging processes to reach about 60- 80% of total capacity within a few minutes, are considered strategic factors for all instrumentation and equipment that use batteries to operate. In fast charging much of the battery charge is reached in a short time, while the remaining capacity may take a long time to fill.

The number of cycles (i.e. useful battery life) and fast charging are also very important features in electrical mobility in order to enter the market effectively. Each cycle, however, corresponds to a loss of capacity, with an increase in this loss if fast charging is used. In practice, in relation to capacity loss, a fast charge corresponds substantially to a plurality of normal charging cycles.

The capacity reduction is known and included in the data sheets provided by the battery manufacturers as an accompaniment and guarantee of their life cycle. Figure 1 shows an example of a graph of the expected life with relative performance of a widely-used lithium-ion battery with cylindrical geometry of the type 18650. This geometry is a common but not exhaustive standard, as the same problem is encountered in all lithium-ion battery geometries.

This is because the loss of capacity with use (e.g. repeated cycles, fast charges) corresponds to a decrease in the lithium available for the charge transfer due to various reactions including the formation and thickening of the solid electrolyte interface (SEI) or solid electrolyte interface layer between the anode and the insulating membrane placed between anode and cathode. Such SEI promotes the ionic transport and its formation is sought in the early life phases of the battery in which electrolyte, anode and lithium ions interact during the charging process. The development of the SEI, however, does not stop and continues throughout the life of the battery, blocking new lithium ions at each step (and thus reducing the real capacity) and increasing its thickness with a consequent reduction in cell conductivity. The increase in the internal resistance of the battery may cause a charged exhausted battery to present a normal terminal voltage, but at the same time fail to deliver a sufficient current. At the same time an increase in the internal resistance can increase the dissipation of the energy in heat.

Among the other significant phenomena related to the loss of capacity is the constant return of less than 100% of the lithium ions by the anode in the discharging phase. A small fraction of lithium can, in fact, remain in the anode after it has entered it during charging, reducing the actual capacity of the battery during discharging.

The substantially linear trend in capacity loss as the number of cycles increases, appreciable in figure 1, confirms that repeated phenomena occur at each cycle that subtract from the battery part of the capacity thereof. The two examples given are presented for non-exhaustive descriptive purposes of the main causes of ageing and capacity loss of lithium-ion batteries. Many other causes, however, typically intervene in unsuitable uses or abnormal battery situations, while with correct use the two examples reported are the most significant causes of the loss of lithium ion battery capacity.

Capacity loss is therefore a known and currently irreversible phenomenon, culminating in the end of the battery's life and the possible recycling of the battery.

The article "Mechanism of the entire overdischarge process and overdischarge-induced internal short circuit in lithium-ion batteries" (R. Guo et al . , Nature 2016) investigates the physical-chemical phenomena that occur inside a lithium battery when it is discharged below the state of charge (SOC) of 0%, i.e. in an overdischarge . An overdischarge may occur, for example, in abnormal conditions in which a battery management system BMS is forced to force the current demand to a battery pack. Overdischarge is, however, a potentially very damaging condition for the state of the battery, which may result in permanent damage. If the cut-off voltage is defined, indicated by each manufacturer specifically for the battery in use, as the voltage value at which the entire useful charge for an operating cycle is released starting from a charged battery, it is possible to identify the null state of charge, i.e. , 0% SOC at this voltage. It should therefore be considered that at cut-off voltage the normal battery management systems already consider the battery electronically, although not yet chemically, discharged. For each battery, therefore, a cut-off voltage value is indicated by the manufacturer, indicating the lower limit voltage at which the battery discharge is considered complete. The cut-off voltage is usually chosen in order to obtain the maximum useful battery capacity. The cut-off voltage is different from one battery to another and is strongly dependent on the type of battery and the type of use to which it is subjected.

The aforementioned article investigates the phenomena that intervene in the application of an overdischarge that goes beyond the elimination of all the available charge (indicated with 0 SOC/%) until reaching a charge removal equal to twice the nominal one (i.e. SOC -100%). This overdischarge leads first to the complete destruction of the SEI and then to the non-reversible destructive aggression of the copper used as an electron collector on which the anode is deposited. The article does not in any way indicate how to apply the knowledge of these phenomena in solving the technical problem of capacity drop as a function of the number of operating cycles .

The document US 2011/236751 A1 describes the application of overcharges to a lithium battery to restore its capacity. The overcharges are such as to ensure that the active material of the positive electrode is somehow stabilized by inserting a greater amount of original lithium in the active material of the positive electrode itself during the overcharge. The overdischarges are therefore not in any way related to the chemical-physical behaviour of the SEI.

Similarly, the document US 2014/197799 A1 describes a method for decreasing the internal resistance of a lithium battery by maintaining the battery for a predetermined time in an overcharged condition. Also in this case, the overdischarge is not related to the chemical-physical behaviour of the SEI.

Therefore, there is currently an unmet need for a process that allows to limit or reverse the capacity drop, regenerating the battery substantially to initial conditions of its operational life.

The object of the present invention is the development and application of an electrochemical discharge and controlled charge protocol that allows to recover the capacity of the battery lost during the life process of the battery itself, restoring it.

The present invention achieves the above object with an electrochemical process for restoring lithium battery capacity comprising the following steps:

a) discharging the battery until reaching the cut off voltage value of the battery; b) overdischarging the battery until a voltage value is reached in which at least a partial destruction of the solid electrolyte interface occurs, this corresponding voltage value being substantially the minimum voltage value reachable by the battery during the overdischarging;

c) recharging the battery by applying a direct current until the maximum voltage value corresponding to the full charge is reached;

d) application of one or more discharge and charge cycles .

The process therefore provides for bringing the battery to an overdischarge condition such as to destroy at least in part and preferably completely the solid electrolyte interface, a phenomenon that occurs approaching the minimum voltage value reachable by the battery, and at that point carrying out a charging process aimed at restoring the full functionality of the battery, similar to that which occurs in the usual methods of forming the solid electrolyte interface for a newly built lithium ion battery not yet operated.

The minimum voltage value that can be reached by the battery is not necessarily the condition of equivalence between the poles, i.e., 0 V, but is the minimum voltage value that an induced discharge causes without the alteration of the current collectors occurring, and in particular, the ionization of the copper. This value is the minimum achievable because when the aggression of the current collectors is irreversibly triggered, the voltage returns to rise precisely for the input in solution of the ions constituting the material of the collectors. From a state of charge point of view, the applied overdischarge corresponds to an SOC below zero corresponding to the minimum observable in a voltage discharge (V) - state of charge (SOC) curve. Said voltage value can then also fall below zero, to negative values, i.e., performing a polarization inversion, as has been observed in most cases in the tests carried out. The overdischarge is typically carried out thanks to the insertion of the battery into a Battery Management System (BMS) type circuit, which can force the extraction of power from the battery even beyond the condition of pole equivalence.

At this minimum voltage value, the complete removal of the lithium present in the anode and the lithium present in the SEI occurs, which is thus disintegrated and largely dissolved in the liquid electrolyte. The process then operates by maximizing the cathode recovery of the lithium available between anode and SEI with a deep discharge process even beyond the pole equivalence limit and with a subsequent reconstruction of the complete cell functionality.

The battery is preferably charged by applying a direct current until the maximum voltage value corresponding to the full charge of the anode is reached and a new SEI is formed at the interface between the separation membrane and the anode. This charging has the purpose of returning the battery to a full state of charge (100% SOC) with a capacity in mAh close to the initial one.

The electrochemical protocol therefore provides for a deep controlled discharge and consecutive controlled charge for the restoration of a capacity completely similar to the initial operating life conditions, and is operable for lithium batteries of any shape, capacity, voltage and that have been used according to the charging and discharging process recommended by the manufacturer, including fast charging, which thus undergo a regeneration process.

It is possible to bring the battery to a discharge level corresponding to the minimum voltage value achievable by the battery by measuring the battery voltage or state of charge in real time and comparing them with their respective reference values. The reference voltage values vary from battery to battery, however, and the measurement may not be accurate enough. The reference values for the state of charge, moreover, are for example SOC -11%, but the state of charge of the battery may not be easily known.

According to a preferred embodiment, therefore, the voltage value in which at least partial destruction of the solid electrolyte interface occurs is detected by calculating the derivative of the battery voltage with respect to the time, the measurement of the battery voltage and the calculation of the derivative being carried out in real time.

In this way it is not necessary to exactly know the state of charge of the battery in real time, nor to accurately compare the measured voltage with threshold values, but it is sufficient to analyse the trend of the voltage over time by means of the derivative in order to identify the minimum point. In particular, it is possible to identify as a minimum point the point for which the derivative is null.

The overdischarge preferably occurs in a controlled manner, i.e., a direct current supply from the battery is forced. In this way it is possible to replace the voltage-state of charge graph with a voltage-time graph in an equivalent manner, i.e., finding the same type of curve.

According to a refinement, the voltage value in which at least partial destruction of the solid electrolyte interface occurs is identified at a predetermined minimum threshold value of the derivative in absolute value.

In this way it is possible to stop the overdischarge process as close as possible to the minimum point without however reaching the minimum point itself, thus reducing as much as possible the risk of entering the non-reversible destructive aggression phase of the copper used as an electron collector on which the anode is deposited.

According to a refinement, a voltage range is defined in which to stop the overdischarge process, which range is comprised between a value slightly higher than the minimum voltage value, for example 20% with respect to such minimum voltage value, and the minimum value identified by the derivative.

In one embodiment, the derivative calculation is performed only below a threshold voltage value.

This precaution avoids the evaluation of variations of the derivative in areas of the voltage time graph that have bending points, which could be exchanged for minimum points if measured with instruments that are not sufficiently accurate.

In one embodiment, a battery charging step is provided prior to step a) until the maximum voltage value is reached.

Only once the battery is fully charged, therefore, is the discharge proceeded with up to the cut-off voltage and the subsequent steps of the process. This is particularly advantageous in order to have a precise idea of the residual capacity of the battery before proceeding with the overdischarge and the consequent regeneration, both to assess the state of health of the battery and to define the residual capacity between the maximum voltage with the battery charged and the minimum voltage of the battery indicated by the manufacturer as the cut-off voltage. Thanks to this value it is possible to define: 1) the capacity lost by the battery due to ageing; 2) the amount of additional capacity recovered during the overdischarge process; 3) the efficiency of the regeneration process by comparing the residual capacity with the capacity acquired as a result of the complete regeneration process .

In one embodiment, step d) provides for a measurement of the state of charge of the battery and the discharge and charge cycles end when the maximum state of charge at the end of two consecutive charge cycles is unchanged or otherwise stable.

At the end of the overdischarge , the battery must be brought to its maximum capacity as occurs for newly produced batteries through a series of discharge and charge cycles. When these cycles are applied, the capacity gradually rises and then settles to a maximum value. The application of additional cycles therefore decreases capacity, as in normal battery operation. Tests have shown that during the discharge and charge cycles the capacity grows asymptotically up to a maximum which, once reached, corresponds to the maximum possible regeneration of capacity. Thus by detecting an unchanged maximum state of charge value between two successive cycles in the increasing phase of the curve, the final condition of the regeneration process can be identified, in which the battery capacity is again maximized.

In a further embodiment increasing C-rate discharge and charge cycles are applied in step d) .

The C-rate is a magnitude indicative of the charge or discharge rate of the battery by a measurement relative to reaching the maximum charge or discharge capacity in 1 hour. This means that the C-rate is equal to 1 when 1 hour is necessary to charge or discharge the battery. The C-rate is calculated as the inverse of fractions of an hour, thus if the charge occurs in 30 minutes or 0.5h the C-rate is equal to 2, while if it occurs in 2 hours it is equal to 0.5. When the C- rate is low, i.e. less than 1, the system is allowed to adjust step by step and thus maximize the efficiency of the discharge or charge. For example, if a battery is discharged with C-rate = 0.2 the measured capacity is close to that available, while if it is discharged at C-rate = 2 the capacity is lower because the kinetics are detrimental to the quality of the work. Consequently, in order to reconstruct the SEI, it is essential to start with a low C-rate that is then gradually increased cycle by cycle. An example of an ideal final value is C-rate = 0.5. This makes it possible to optimize the process while maintaining reasonable times .

In one embodiment, in step d) the temperature of the battery is kept constant between 20 °C and 50 °C, preferably between 30 °C and 40°C.

According to one embodiment, at least five discharge and charge cycles are applied in step d) . This achieves battery stability, which can then be put back into operation.

Variations of the described process may be envisaged in order to adapt the process to the specific type of lithium-ion battery in use.

An object of the present invention is also a battery management system (BMS) , i.e. an electronic system that manages one or more rechargeable batteries, for example by protecting the batteries from operating outside the safe operating area, monitoring their status, calculating secondary data, reporting such data, controlling their environment, validating it and/or returning it to optimal conditions. The BMS object of the present invention is configured to implement the process described herein. In particular, the BMS comprises a control unit, discharge means of one or more batteries , charge means of one or more batteries, voltage measurement means, and said control unit is configured to implement the method described herein .

The BMS preferably applies the regeneration process by detecting the residual capacity value and comparing it with: 1) the initial nominal capacity of the battery; 2) the capacity removed during the overdischarge ; 3) the capacity acquired as a result of the regeneration process.

The process is applied by the BMS at a given stage of battery capacity reduction in an operator-driven manner or automatically by software based on an assessment of the performance status achieved.

According to an embodiment, means are provided for detecting the capacity decrease of one or more batteries as a function of the number of operating cycles, and means are provided for comparing the value of the detected decrease with a threshold value such that when the detected decrease exceeds said threshold value, said battery management system implements said process on said one or more batteries.

This automates the performance of the capacity restoration process, which is thus applied to one or more batteries when they exhibit reduced capacity for a predetermined rate.

The BMS can also implement an automation of the process described above in an adaptable manner by the operator based on specific values related to the battery on which it works.

In one embodiment said comparison means compare the detected capacity value with the initial capacity value and the capacity values of the previous cycles in order to determine the slope of the ageing curve. Preferably, two thresholds are set: a first threshold is relative to the slope of the charge capacity-cycle curve, which is compared with the slope defined according to the indications in the data sheet of the battery manufacturer and any internal checks; a second threshold is relative to the achievement of a predefined residual capacity in the programming step. Reaching at least one of the two threshold values causes the battery management system to implement the regeneration process for one or more batteries.

In one embodiment the battery management system is configured for managing a set of batteries or battery pack and said control unit is configured for dividing said set into sub-sets of batteries and for implementing said process only on one or more sub-sets. This allows to automate the implementation of the process only on a part of the entire set of batteries, while the other batteries can be used normally.

In one embodiment a predetermined number of additional subsets are provided maintained in storage conditions and redundant with respect to the declared power of the battery pack in order to maintain the nominal power of the battery pack unchanged even during the regeneration of one or more sub-sets.

In a further embodiment, alternatively or in combination, said control unit is configured to periodically implement said process consecutively on each individual sub-set of batteries.

This allows to schedule a periodic restoration of each individual battery sub-set, which cyclically covers the entire battery pack to continuously restore the capacity of all batteries, in particular by using the remaining battery sub-sets to manage the regeneration process to support or replace, if absent, the mains supply.

The electrochemical process described heretofore is intended to be applied to lithium-ion batteries for the maintenance of capacity over the operating life of the component with a consequent increase in the number of cycles. The process can be applied at any time on functioning batteries that have undergone charge and discharge cycles. The application of fast charges is included among the charging processes indicated in the cycles .

Batteries which have operated canonically with charge and discharge cycles including also fast charge cycles that have been subjected to the process described hereinabove may be restored to normal service and, following further operation, may again be subjected to the process object of the invention.

These and other features and advantages of the present invention will become clearer from the following description of some non-limiting exemplary embodiments illustrated in the attached drawings in which :

fig. 1 shows a graph of the capacity drop in a lithium-ion battery relative to the number of operating cycles ;

fig. 2 shows a diagram of an embodiment of the process ;

fig. 3 shows a graph of the voltage of a lithium- ion battery in the overdischarge step as a function of time and the trend of the voltage derivative;

fig. 4 shows the capacity increase in a battery recovered according to the process;

fig. 4 shows a block diagram of an embodiment of the battery management system.

As mentioned at the beginning, figure 1 shows a graph of the capacity of a lithium ion battery as the number of operating cycles increases . Each operating cycle represents a complete discharge and a complete charge. The graph refers to a PANASONIC NCR18650E battery. Nevertheless, batteries of other types were also tested, always obtaining satisfactory results. The graph clearly shows an almost linear decrease in battery capacity as the number of operating cycles increases, losing about 30% of the capacity at 500 cycles .

The present invention relates to an electrochemical process to reverse this decay by restoring the lithium battery capacity, an example of which is shown in figure 2 and is described below with reference to the PANASONIC NCR18650E battery mentioned in relation to figure 1. However, the process is applicable to lithium-ion batteries of any shape and size, which have been used and exhibit any level of capacity reduction, but which are intact or not exhausted due to internal short circuits or similar damage .

The process outlined in figure 2 is described in relation to the graph of the terminal voltage trend of the battery with respect to time, and to the calculation of the derivative of this voltage, illustrated in figure 3. The voltage curve is indicated with 24, while the derivative is reported on a logarithmic scale and is indicated with 25.

The process starts with the charge 10 of the battery until the maximum voltage provided for the type of battery used is reached, through a standard charging process. For the model NCR18650E, the maximum voltage is 4.2V.

A discharge 11 is then carried out until the cut off voltage value indicated by the specifications of the battery used is reached. For the model NCR18650E the minimum or cut-off voltage is 2.75V.

The voltage is therefore at the cut-off point 20 corresponding to a 0% state of charge.

The battery is then forced to deliver additional current, to enter an overdischarge step until the complete destruction of the solid electrolyte interface, which occurs at the minimum voltage point 22.

The battery enters an initial overdischarge step 12, up to a threshold voltage value. This step is optional and allows to avoid, in the analysis for the identification of the minimum voltage point 22, abnormal points such as the inflection point 21, which is normally exhibited by all batteries.

Falling below the threshold voltage, the battery enters a controlled overdischarge step through the voltage value as a function of the elapsed time, until the minimum voltage point 22 is reached in the graph of the voltage relative to process time. The minimum is evaluated mathematically and on the process in progress, with the application of a control function on the voltage values measured in real time. In particular, the derivative dV/dt of the voltage with respect to time is calculated in real time, for which the null value or the absolute minimum value is sought, for example by identifying a minimum threshold value or using a calibrated derivative. This is possible thanks to the absorption of a direct current output from the battery, in which the output charge is therefore constant with respect to time. However, other methods of finding the minimum voltage 22 can be used in the voltage-time graph.

If the overcharge continues beyond the minimum voltage point 22, the battery would enter a phase of irreversible modifications that would affect its restoration or even just its operation. In particular, beyond the minimum voltage point 22, a non-reversible destructive aggression step 23 is entered of the copper used as an electron collector on which the anode is deposited.

It is therefore preferable to stop the overdischarge at a minimum value in absolute value of the derivative dV/dt such as to bring the voltage as close as possible to the minimum point 22, as can be seen in figure 3, without however reaching it.

The battery is then charged with the application of a fixed and controlled current until the maximum voltage value is reached as indicated by the battery specifications, corresponding to the full charge. As written above, for the model NCR18650E, the maximum voltage is 4.2V.

Then the application of a number of discharge and charge cycles follow with capacity control to achieve battery stability. Experimental tests have shown that a number of five cycles is sufficient to achieve this stability, in particular for NCR18650E type batteries.

Measurements of the capacity of a battery used and subjected to the restoration process of the present invention are illustrated in figure . For each cycle there is a full charge and a full discharge, with battery capacity determination (BCD) . The first value indicated with BCD corresponds to the measured capacity of the battery before undergoing the restoration process. The following values relate to the battery capacity following the application of the restoration process. The measured capacity values increase from about 2000 mAh to about 2100 mAh, confirming the presence of a significant increase in the capacity guaranteed by the restoration process.

Figure 5 shows the diagram of an embodiment of a BMS 40 configured to implement the process described in figure 2. A BMS 40 of the present invention can also manage a single battery 4. In the example shown, the BMS 40 manages a set of batteries 4, or battery pack 41. The BMS is configured to apply the process to a given stage of battery 4 capacity reduction in an operator-driven manner or automatically by software based on an assessment of the performance status achieved. In particular, a capacity decrease detection unit 43 of one or more batteries 4 is provided as a function of the number of operating cycles, which performs repeated measurements at each cycle or periodically upon reaching a predetermined number of cycles, to monitor the capacity drop. Furthermore, a comparison unit 44 is provided of the detected decrease with a threshold value. In this way, when the detected decrease exceeds the threshold value, the BMS 40 implements the process on batteries that need to be restored to their capacity.

In the advantageous example in the figure, moreover, the BMS 40 is configured for subdividing the battery pack 41 into sub-sets 42 of batteries 4 for the selective implementation of the capacity restoration process on only one or more sub-sets 42. Advantageously, the BMS 40 is configured to periodically implement the process, sequentially acting on each individual sub-set of batteries 42. This automates the implementation of the process only on a sub-set 42 of the entire battery pack 41 at a time, while the other batteries can be used normally, but cyclically covering the entire battery pack 41 to continuously restore the capacity of all batteries 4.

Claims

1. An electrochemical process for restoring the capacity of one or more lithium batteries (4) ,

characterized in that

it comprises the following steps:

a) discharging (11) the battery (4) until the cut off voltage value of the battery (4) is reached;

b) overdischarging (13) the battery (4) until a voltage value is reached wherein at least a partial destruction of the solid electrolyte interface occurs, this corresponding voltage value being substantially the minimum voltage value (22) reachable by the battery (4) during the overdischarge;

c) charging (14) the battery (4) by applying a direct current until the maximum voltage value corresponding to the full charge is reached;

d) applying one or more discharge and charge cycles (15) .

2. The process according to claim 1, wherein the voltage value in which at least partial destruction of the solid electrolyte interface occurs is identified by calculating the derivative of the battery voltage with respect to the time, the measurement of the battery voltage and the derivative being carried out in real time.

3. The process according to claim 2, wherein the voltage value in which at least partial destruction of the solid electrolyte interface occurs is identified at a predetermined minimum threshold value of the derivative in absolute value.

4. The process according to claim 3, wherein the derivative calculation is performed only below a threshold voltage value.

5. The process according to one or more of the preceding claims, wherein a step of charging (10) the battery (4) is provided before step a) , until the maximum voltage value is reached.

6. The process according to one or more of the preceding claims, wherein step d) provides for a measurement of the state of charge of the battery and the discharge and charge cycles end when the maximum state of charge at the end of two consecutive charge cycles is unchanged.

7. The process according to one or more of the preceding claims, wherein at least five discharge and charge cycles are applied in step d) .

8. The process according to one or more of the preceding claims, wherein discharge and charge cycles at increasing C-rates are applied in step d) .

9. The process according to one or more of the preceding claims, wherein in step d) the temperature of the battery is kept constant between 20° C and 50° C, preferably between 30° C and 40° C.

10. A battery management system (40), comprising a control unit, discharge means of one or more batteries , charge means of one or more batteries , voltage measurement means,

characterized in that

said control unit is configured to implement the process according to one or more of claims 1 to 9.

11. The battery management system (40) according to claim 10, wherein means are provided for detecting (43) the capacity decrease of one or more batteries (4) as a function of the number of operating cycles, and means are provided for comparing (44) the detected decrease with a threshold value such that when the decrease detected exceeds said threshold value, said battery management system (40) implements said process on said one or more batteries (4) .

12. The battery management system (40) according to claim 10 or 11, which battery management system (40) is configured for managing a set of batteries (41) and said control unit is configured for dividing said set (41) into sub-sets of batteries (42) and for implementing said process only on one or more sub-sets (42) .

13. The battery management system (40) according to claim 12, wherein the control unit is configured to periodically implement said process consecutively on each individual sub-set of batteries (42) .