FR3160471A1 - METHOD FOR DETECTING LITHIUM FORMATION ON AN ELECTRODE IN A LITHIUM-ION BATTERY - Google Patents

Abstract

The invention relates to a method for detecting lithium formation in a lithium-ion battery for at least one charging regime 1a, called the "basic regime," comprising the steps of: applying the basic regime 1a, having a first current I1a, and a charging regime 2a, called the "secondary regime," having a second current I2a; acquiring the voltage and the charge quantity Q of the battery, during the basic regime 1a and the secondary regime 2a; determining a ratio between, on the one hand, the difference between the voltage in the basic regime 1a and that in the secondary regime 2a, and, on the other hand, the difference between the first charging current I1a and the second charging current I2a, as a function of the charge quantity Q; and detecting the appearance of lithium as a function of said ratio. Figure 2

Description

METHOD FOR DETECTING LITHIUM FORMATION ON AN ELECTRODE IN A LITHIUM-ION BATTERY

The invention relates to a method for detecting a formation of lithium, in particular metallic lithium, on an electrode in a lithium-ion battery. The invention further relates to a lithium-ion battery management system configured to implement the method according to the invention.

In the general context of decarbonizing transport, there is a growing demand for electric vehicles that are both quickly rechargeable and have a long range. This demand concerns both light vehicles, such as bicycles, scooters, or other motorized two-wheelers, as well as heavy vehicles, for example cars of all sizes, or trucks. Lithium-ion battery technology (or "Li-ion batteries") is currently the most promising thanks to its performance in terms of mass or volume energy density and power.

Lithium-ion batteries operate on the principle of alternating intercalation of lithium ions between a material forming a positive electrode and a material forming a negative electrode. Typically, during normal operation, the positive electrode has a high electrical potential compared to that of metallic lithium; and the negative electrode has an electrical potential close to that of metallic lithium, but still higher. Lithium ions flow from the positive electrode to the negative electrode, to be integrated into the latter.

However, during fast charging of lithium-ion batteries, the potential of the negative electrode is likely to fall lower than that of metallic lithium: the lithium ions then no longer integrate into the negative electrode and are deposited on its surface. An unwanted deposit of metallic lithium forms on the negative electrode, particularly in the form of dendrites. This phenomenon of lithium plating causes serious damage to the battery if it occurs frequently during charging. In practice, we observe an accelerated loss of battery capacity, and a higher probability of short circuits between the positive and negative electrodes. It is therefore crucial, particularly for manufacturers of electric vehicles, to be able to detect the formation of lithium on the negative electrode in Li-ion batteries throughout their life. This is particularly in order to adjust fast charging programs to avoid lithium formation.

A method for real-time detection of the phenomenon of lithium deposition on the negative electrode is known. It is notably described in patent publication US 10,126,367 B2. The method is based on the charging voltage derivative and consists of monitoring, during charging at a constant charge rate, the variation dU/dQ of the voltage U at the terminals of the battery per unit of charge exchanged as a function of the quantity of charge Q stored in the Li-ion battery. The appearance of a peak on the dU/dQ curve indicates the start of lithium deposition on the negative electrode. However, this method is effective on Li-ion batteries having a regular open circuit voltage or OCV (Open Circuit Voltage). This characteristic is strongly linked to the material of the positive electrode. This is the case for Li-ion batteries whose positive electrode is made of NMC 622, i.e. comprising 60% nickel, 20% manganese and 20% cobalt. On the other hand, Li-ion batteries whose positive electrode is made of certain materials can present several strong variations in slope in the dU/dQ curve making it difficult to identify the significant peak of lithium formation. This is particularly the case when the positive electrode is made of NMC 811. FIG. 1 represents curves acquired on a Li-ion battery having a positive electrode in NMC 811 and a negative electrode in graphite on a full charge for different charge regimes. The curves are acquired for charge regimes having a charge rate of 0.5C, 0.75C, 1C, 1.25C and 1.5C. A charge rate of 1C corresponds to a charge current by which the full charge is reached after 1h. A charge rate of 0.5C corresponds to a charge current by which the full charge is reached after 1/0.5h, or 2h. The formation of lithium corresponds to the first peak P of each curve. However, the dU/dQ curve is subject to several other phenomena than the formation of lithium, in particular to the natural variations of the equilibrium potential ("open circuit voltage (OCV)" in English) of the positive electrode material; which is reflected in particular by the two peaks appearing in the second part of the curve. This can make it difficult to identify the significant P peak of lithium formation. Furthermore, this identification is only sufficiently accurate from a charge rate above 1C.

Therefore, a method for detecting the formation of lithium on the negative electrode during Li-ion battery charging is sought which is sufficiently precise, particularly for Li-ion batteries such as their OCV is irregular.

1. appliquer le régime de base, ayant un premier courant de charge, et un régime de charge, dit« régime secondaire », ayant un deuxième courant de charge, la valeur absolue de la différence entre le taux de charge dans le régime de base et le taux de charge dans le régime secondaire étant inférieure à un seuil ;
2. acquérir la tension aux bornes de la batterie et la quantité de charge de la batterie, au cours du régime de base et du régime secondaire ;
3. déterminer un rapport entre d’une part la différence entre la tension aux bornes de la batterie dans le régime de base et celle dans le régime secondaire, et d’autre part la différence entre le premier courant de charge et le second courant de charge, en fonction de la quantité de charge ; et
4. pour le régime de base, détecter l’apparition d’une formation de lithium en fonction de la valeur dudit rapport.

To this end, the invention proposes a method for detecting lithium formation on an electrode in a lithium-ion battery, comprising detecting lithium formation for at least one charging regime, called "basic regime", comprising the steps of:

1. applying the basic regime, having a first charging current, and a charging regime, called “secondary regime”, having a second charging current, the absolute value of the difference between the charging rate in the basic regime and the charging rate in the secondary regime being less than a threshold;
2. acquire the battery terminal voltage and the battery charge quantity, during the basic regime and the secondary regime;
3. determining a ratio between, on the one hand, the difference between the voltage at the terminals of the battery in the basic regime and that in the secondary regime, and on the other hand the difference between the first charging current and the second charging current, as a function of the quantity of charge; and
4. for the basic regime, detect the appearance of lithium formation based on the value of said ratio.

Said ratio has the advantage of being relatively constant at the start of a charge, in particular when the charge begins from a small quantity of charges stored in the battery, then of collapsing when there is formation of lithium on the negative electrode. This strong variation of the ratio makes it possible to better identify the appearance of a lithium deposit. The variation of said ratio has the advantage of being more regular than that of the dU/dQ curve used in the prior art. In addition, the method according to the invention is effective, whatever the OCV of the material of the positive electrode.

According to one embodiment, the appearance of a lithium deposit corresponds to a last maximum, in particular a last local maximum, of said ratio, preceding a fall of at least 10% of said ratio.

According to one embodiment, the absolute value of the difference between the load rate of the basic regime and the load rate of the secondary regime is less than 0.5 and preferably less than 0.25.

According to one embodiment, the application of the basic regime and the secondary regime is carried out alternately during a charge of the lithium-ion battery.

According to one variant, the secondary regime is applied over periods that are negligible compared to the periods over which the basic regime is applied.

According to one variant, the method comprises emitting a lithium formation detection signal when the value of the ratio becomes less than or equal to a threshold.

According to a variant, said threshold is a function of the quantity of charge Q of the battery and is equal to:

where Q _tot is the maximum charge quantity of the battery in mAh, R ₀ is the initial value of the ratio when charging the battery from a state of charge less than 0.1Q _tot , α is a constant between 0.75 and 0.95.

1. lorsque la charge de la batterie a lieu depuis un état de charge inférieur ou égal à 0,1Q_tot:
   • la détermination du paramètre R₀en calculant la moyenne dudit rapport pour des quantités de charge inférieures ou égales à 0,2Q_tot;
   • le stockage du paramètre R₀dans une mémoire ; et
   • l’utilisation du paramètre R₀stocké en mémoire pour la détermination du seuil r_s(Q) pour les quantités de charge supérieurs à 0,2Q_tot,
2. ou, lorsque la charge de la batterie a lieu depuis un état de charge supérieur à 0,1Q_tot, l’utilisation du dernier paramètre R₀stocké en mémoire pour la détection de la formation de lithium au cours de la charge.

According to one variant, the method comprises:

1. when the battery is charged from a state of charge less than or equal to 0.1Q _total :
   • determining the parameter R ₀ by calculating the average of said ratio for load quantities less than or equal to 0.2Q _total ;
   • storing the parameter R ₀ in a memory; and
   • the use of the parameter R ₀ stored in memory for the determination of the threshold r _s (Q) for the quantities of charge greater than 0.2Q _tot ,
2. or, when the battery is charged from a state of charge greater than 0.1Q _tot , the use of the last parameter R ₀ stored in memory for the detection of lithium formation during charging.

1. la comparaison entre les deux dernières valeurs du paramètre R₀déterminées chacune lors d’une charge de la batterie depuis un état de charge inférieur ou égal à 0,1Q_tot, et
2. lorsque la différence entre lesdites deux dernières valeurs du paramètre R₀est supérieure à un seuil, l’émission d’un signal de défaut de la batterie.

According to one variant, the method comprises:

1. the comparison between the last two values of the parameter R ₀ each determined during a charge of the battery from a state of charge less than or equal to 0.1Q _tot , and
2. when the difference between the last two values of parameter R ₀ is greater than a threshold, the emission of a battery fault signal.

According to a variant, the charging of the lithium-ion battery is a rapid charge having decreasing charging current levels, the method comprising detecting the formation of lithium for a plurality of basic regimes, each basic regime corresponding to one of said current levels.

1. l’application du régime de base s’effectue sur une première charge complète de la batterie, la tension aux bornes de la batterie étant acquise en fonction de la quantité de charge de la batterie,
2. l’application du régime secondaire s’effectue sur une deuxième charge complète de la batterie, la tension aux bornes de la batterie étant acquise en fonction de la quantité de charge de la batterie.

According to one embodiment:

1. the application of the basic regime is carried out on a first complete charge of the battery, the voltage at the terminals of the battery being acquired according to the quantity of charge of the battery,
2. the application of the secondary regime is carried out on a second complete charge of the battery, the voltage at the battery terminals being acquired according to the quantity of charge of the battery.

1. la détection d’un dépôt de lithium pour une pluralité de régimes de base ;
2. le stockage dans une mémoire d’une table comprenant, pour chacun de la pluralité de régime de base, la quantité de charge respectif à partir duquel il y a apparition d’un dépôt de lithium ; et
3. lors d’une charge ultérieure de la batterie, l’utilisation de ladite table pour détecter l’apparition d’une formation de lithium sur une électrode en fonction du courant de charge appliqué et de la quantité de charge de la batterie.

According to one variant, the method comprises:

1. the detection of a lithium deposit for a plurality of basic regimes;
2. storing in a memory a table comprising, for each of the plurality of basic regimes, the respective charge quantity from which a lithium deposit appears; and
3. upon subsequent charging of the battery, using said table to detect the occurrence of lithium formation on an electrode as a function of the charging current applied and the amount of battery charge.

Lithium-ion battery management system configured to implement a method for detecting a formation of lithium on an electrode according to the invention during charging of the lithium-ion battery.

Other characteristics and advantages of the present invention will appear more clearly on reading the description which follows in relation to the following appended figures:

FIG. 1 : there FIG. 1 , already described, illustrates an example of a method according to the prior art;

FIG. 2 : there FIG. 2 represents a current delivered to a Li-ion battery in a first example of a method according to the invention;

FIG. 3 : there FIG. 3 represents the voltage at the terminals of the battery in the first example of the method according to the invention;

FIG. 4 : there FIG. 4 represents the ratio R in the first example of the method according to the invention;

FIG. 5 : there FIG. 5 represents a current delivered to a Li-ion battery in an exemplary method according to the invention;

FIG. 6 : there FIG. 6 represents the voltage across the battery terminals in the process example illustrated in FIG. 5 ;

FIG. 7 : there FIG. 7 represents the ratio R in several examples of the method according to the invention;

FIG. 8 : there FIG. 8 represents the negative potential of the battery in several examples of the method according to the invention;

FIG. 9 : there FIG. 9 represents a current delivered to a Li-ion battery in an exemplary method according to the invention;

FIG. 10 : there FIG. 10 represents the voltage at the terminals of the battery as a function of the quantity of charges stored, for several examples of the method according to the invention;

FIG. 11 : there FIG. 11 represents the ratio R for several examples of the method according to the invention;

FIG. 12 : there FIG. 12 represents the negative potential of the battery for several examples of the method according to the invention.

A first example of a method according to the invention will be described in relation to figures 2 to 4. The method makes it possible to detect the formation of lithium, in particular metallic lithium, on a negative electrode of a lithium ion battery.

For this purpose, the method comprises detecting the formation of lithium on a charging regime, called the basic regime 1a, having a first charging current I _1a . As for example illustrated in FIG. 2 , the basic regime 1a is applied to the Li-ion battery with the first charging current I _1a . Another charging regime, called "secondary regime" 2a, having a charging current I _2a is further applied to the Li-ion battery. In FIG. 2 , the y-axis represents the values of the current i(t) in mA and the values of the charge quantity Q(t) in mAh.

The secondary regime 2a is chosen so that the charge rate C1 of the base regime 1a and the charge rate C2 of the secondary regime 2a are close. In this respect, the secondary regime 2a may have a higher or lower charge current I _2a than the charge current I _1a in the base regime 1a. In other words, the base regime 1a and the secondary regime 2a are chosen so that the absolute value of the difference C2-C1 between the charge rate C1 in the base regime 1a and the charge rate C2 in the secondary regime 2a is less than a threshold.

As for example illustrated in FIG. 3 , the voltage U _1a at the battery terminals is acquired during the basic regime 1a, as is the voltage U _2a at the battery terminals during the secondary regime 2a. As for example illustrated in FIG. 2 , the charge quantity Q of the battery is also acquired, in particular in parallel with the acquisition of the voltages U _1a , U _2a at the terminals of the battery during the basic regime 1a and the secondary regime 2a. From the acquired values, we determine, for a state of charge Q of the battery, a ratio R(Q) between on the one hand the difference ΔU between the voltage U _1a at the terminals of the battery during the basic regime 1a and the voltage U _2a at the terminals of the battery during the secondary regime 2a, and on the other hand the difference ΔI between the first charge current I _1a and the second charge current I _2a :

There FIG. 4 illustrates the evolution of the ratio R as a function of the quantity of charges Q of the battery. The shape of the graph makes it possible to detect the appearance of a lithium deposit on the negative electrode. In particular, when the quantity of charges Q increases, the curve remains relatively stable, in particular to within 0.1x10 ^-3 Ω, before falling relatively sharply, which is significant of a formation of lithium on the negative electrode of the battery. Thus, the quantity of charges for which there is formation of lithium is determined as a function of the value of the ratio R.

In particular, the amount of charge for which a lithium deposit appears corresponds to a last maximum, in particular a last local maximum, of the ratio R preceding a fall of at least 10% of said ratio. By considering a low fall value, the formation of lithium can be detected as early as possible, but there is then the risk of false detections due to intrinsic variations in the ratio curve. By taking a higher fall value, a false detection can be avoided, but possibly to the detriment of the accuracy of the detection of lithium formation.

The charge rate C1 of the basic regime 1a and that C2 of the secondary regime 2a are preferably close so that the detection of lithium formation by the ratio R is meaningful. For example, the absolute value of the difference between the charge rate C1 of the basic regime 1a and the charge rate C2 of the secondary regime 2a is less than or equal to 0.5 and preferably less than 0.25. However, the absolute value of the difference between the charge rate C1 of the basic regime 1a and the charge rate C2 of the secondary regime 2a is preferably large enough to give a meaningful result, taking into account the precision with which the voltages U _1a U _2a are acquired.

The first method example 3 is according to a first embodiment in which the basic regime 1a and the secondary regime 2a are applied alternately during a charge of the battery. In other words, during the same charge, the basic regime 1a and the secondary regime 2a are applied successively. Thus, by monitoring the ratio R between the basic regime 1a and the secondary regime 2a, it is possible to detect the appearance of a lithium deposit during the charge itself. In particular, the method according to this embodiment allows real-time detection of the appearance of a lithium deposit during the charge. In particular, in the context of the present application, the term “charge” means the uninterrupted application of an electric current to the terminals of the Li-ion battery in order to increase the quantity of charges stored in the battery.

In particular, the secondary regime 2a is applied over negligible durations compared to the durations over which the basic regime 1a is applied. In particular, the basic regime 1a is a nominal charging regime. The secondary regime 2a is therefore sufficiently short so as not to disturb the charging of the battery with the basic regime. In particular, the ratio between a duration of application of the secondary regime and a duration of application of the basic regime is less than or equal to 10 ^-2 . The application of the secondary regime 2a takes in particular the form of pulses during the basic regime 1a. For example, an occurrence of the secondary regime 2a has a duration of between 1s and 10s. Thus, in FIG. 2 , the duration of a pulse corresponding to the secondary regime 2a is for example equal to 5s.

Preferably, the secondary regime 2a is applied periodically. Thus, the ratio R can be determined regularly during charging, and then allow regular detection of lithium deposition on the negative electrode. For example, the secondary regime 2a is applied with a period Ts between 50s and 1000s. In particular, a period Ts shorter than 50s could lead to disturbances of the basic regime 1a, while periods greater than 1000s would not allow sampling precise enough to detect the moment of lithium deposition.

Preferably, the difference between the load rate C1 of the basic regime 1a and the load rate C2 of the secondary regime 2a is less than or equal to 0.5 and preferably less than 0.25. Thus, in FIG. 2 , the difference ΔI between the load current I2a in the secondary regime 2a and the load current I1a in the basic regime 1a is for example equal to 10mA.

As for example illustrated in FIG. 3 , each application of the secondary regime 2a induces a voltage U _2a greater than the voltage U _1a of the basic regime 1a. By calculating the voltage difference ΔU between the value of the voltage U _1a before the application of the secondary regime 2a and the value of the voltage U _2a during the application of the secondary regime 2a, and by relating it to the difference between the load current I _2a in the secondary regime 2a and the load current I _1a in the basic regime 1a, we obtain the ratio illustrated in FIG. 4 . To avoid applying the secondary regime 2a for too long and thus disrupting the battery charge, only one voltage value U _2a is preferably measured per application of the secondary regime 2a. For example, the measurement is carried out one second after applying the secondary regime 2a. However, it could be considered to take the average of several measurements of the voltage U _2a .

In FIG. 7 , the ratio R is shown in the first process example 3 with a base regime 1a having a charge rate of 1C and a secondary regime 2a having a charge rate of 1.25C. For comparison, the ratio R is also shown in a second process example 4 according to the first embodiment with the base regime having a charge rate of 0.75C and the secondary regime 2a having a charge rate of 1C; and the ratio R in a fast charge example 5 described later. The second process example 4 according to the first embodiment is identical to the first example 3, except for the charge rates in the base regime and the secondary regime.

Comparing the curves with each other, we note in particular that the higher the charge rate, the more the drop in the R ratio, indicating lithium formation, is marked and appears earlier. In particular, the drop in the R ratio remains sufficiently marked for a basic regime having a charge rate greater than or equal to a threshold, for example 0.75C, in particular for an example of an NMC 811-graphite type battery. However, the minimum charge rate for which the drop in the R ratio remains significant depends on the type of Li-ion battery and its state of aging.

In particular, the battery is capable of receiving a maximum charge quantity Q _tot . In particular, for quantities of stored charges Q less than 0.2Q _tot , the ratio R is substantially constant for a given charge quantity regardless of the charging regime considered. This is particularly the case under normal operating conditions, i.e. outside of any formation of lithium on the negative electrode. For example, in FIG. 7 , we thus observe that for a quantity of stored charges less than 0.2Q _tot , the R ratios are relatively close. We also observe that the R ratios decrease slightly on average from an initial value R ₀ , for example 4.2x10 ^-3 Ω, whatever the charge regime considered, before reaching the lithium formation zone.

Such a decrease is for example represented in FIG. 4 and 7 by a straight line r, for example with the equation:

Where Q _tot is the maximum charge quantity of the battery in mAh,
R ₀ is the initial value of the ratio when charging the battery from a state of charge less than 0.1Q _tot .

When lithium begins to form on the negative electrode, there is a relatively sharp drop in the R ratio. The last maximum before the drop in the R ratio is notably around 8 mAh for the first process example 3. However, to avoid false detections, one can choose to trigger the emission of a lithium formation detection signal when the value of the R ratio becomes less than or equal to a threshold r _s . The threshold r _s is for example defined by the expression:

Where α is a constant between 0.75 and 0.95.

Thus, the threshold r _s is in particular a translation of the straight line r corresponding to the shape of the ratio R under normal operating conditions, with a reduction factor α. For example, the reduction factor α is equal to 0.90. In particular, crossing the threshold r _s will cause the emission of a signal detecting a lithium deposit on the negative electrode. In the event of heating of the battery during charging, the shape of the ratio R may vary, the threshold r _s can then be adapted. In particular, heating of the battery results in a greater average decreasing slope of the ratio R before the lithium formation zone. However, a sudden drop in the ratio R will still be significant of lithium formation. The threshold r _s is preferably corrected to take into account the heating.

The detection of the appearance of a lithium deposit can be done by means other than crossing a threshold. For example, it can be done by determining the local slope of the R ratio, in particular after noise filtering. Or, by the detection by artificial intelligence of an S-shaped profile of the R ratio, said profile being significant of the appearance of the lithium deposit.

The initial value R ₀ of the ratio R can be determined during charging. This is particularly the case when charging starts from a battery charge quantity that is less than 0.1Q _tot . The initial value R ₀ can then correspond to the average of the ratio R for charge quantities less than or equal to 0.2Q _tot . The initial value R ₀ thus determined is then stored in a memory, particularly a memory of a battery management system, to be used in determining the threshold r _s for charge quantities greater than 0.2Q _tot .

Alternatively, the initial value R ₀ used in the charge may be a value stored in memory, in particular in a memory of a battery management system. This is particularly the case when the battery is charged from a battery charge quantity that is greater than 0.1Q _tot . In particular, the initial value R ₀ then corresponds to the last value stored in memory during a previous charge from a charge quantity less than 0.1Q _tot .

The determination of the initial value R ₀ can be carried out at each charge of the battery starting from an initial quantity of stored charges less than or equal to 0.1Q _tot . The determination of the initial value R ₀ can also be carried out every X charges starting from an initial quantity of stored charges less than or equal to 0.1Q _tot , X being a natural number.

The evolution of the initial value R ₀ during battery charging can be significant of battery aging. For example, we can compare the last two values of the parameter R ₀ each determined during a battery charge. A large difference between these last two values of the parameter R ₀ can be significant of a sudden and significant deterioration of the battery. Thus, we can predict that when the difference between the last two values of the parameter R ₀ is greater than a threshold, a battery fault signal is emitted, in particular by a battery management system.

The first method example 3 has been described with a basic regime 1a, having a constant first load current I _1a , and a secondary regime 2a having a constant load current I _2a , as for example illustrated in FIG. 2 . However, the basic regime 1a could have a variable charging current I _1a , as could the secondary regime 2a, with a difference C2-C1 between the charging rate C1 of the basic regime 1a and the charging rate C2 of the secondary regime 2a remaining below a threshold as described previously, and in particular remaining constant.

Figures 5 and 6 show a third example of fast charging 5, in which the charging current varies during charging. The example of fast charging 5 is implemented in particular on a Li-ion battery of the NMC 811-graphite type. In FIG. 5 , the y-axis represents the values of the current i(t) in mA and the values of the charge quantity Q(t) in mAh. In particular, as for example illustrated in FIG. 5 , charging starts with a relatively high current, then the current is decreased in stages during charging. Such a charging profile makes it possible to adapt the charging current to avoid lithium formation on the negative electrode. The charging current profile can be determined at the beginning of the battery life. However, over time such a current profile may become ineffective due to battery aging. The method according to the invention can be applied for each current stage and thus make it possible to verify in real time that lithium formation is not taking place, and thus make it possible to verify that the charging current profile of the fast charge 5 is still effective in avoiding lithium formation in the Li-ion battery.

In particular, the third method example is applied with a first basic regime 1a, having a constant load current I _1a corresponding to the first current level. Then, the third method example is applied with a second basic regime 1b, having a constant load current I _1b corresponding to the second current level, and so on for the other current level(s).

The third method example can be implemented with a secondary regime 2a having a charging current I _2a greater than the charging current I _1a of the base regime 1a, or with a secondary regime 2a' having a charging current I _2a' less than the charging current I _1a of the base regime 1a. In particular, charging of the battery can be implemented with the third method example using only positive pulses, i.e. with a secondary regime 2a having a charging current greater than that of the base regime 1a; or only negative pulses, i.e. with a secondary regime 2a' having a charging current I _2a' less than that of the base regime 1a; or both.

There FIG. 6 represents the voltage U across the battery terminals. The voltage increases gradually during charging, with jumps corresponding to changes in the current level. The voltage exhibits other variations corresponding to the application of the secondary regimes. In particular, positive current pulses induce a voltage U _2a in the secondary regime 2a; negative current pulses induce a voltage U _2a' in the secondary regime 2a'. Thus, in particular, each application of the secondary regime 2a induces a voltage U _2a higher than the voltage U _1a of the basic regime 1a or a voltage U _2a' lower than the voltage U _1a of the basic regime 1a, depending on the polarity of the pulse.

There FIG. 7 represents the evolution of the R ratio as a function of the battery charge for the fast charge example 5. For this graph, only positive pulses are taken into account, the results being similar for negative pulses. We note that curve 5 only reaches the r _s threshold at the end of charging, indicating however that the last current level should be adjusted to avoid the formation of lithium on the negative electrode.

There FIG. 8 represents the potential Un of the negative electrode during charging for the fast charging example 5 and the first example 3 of the method according to the first embodiment with a charging rate of 1C for the basic regime, and the second example 4 of the method according to the first embodiment with a charging rate of 0.75C for the basic regime. The passage of the potential Un of the negative electrode below the zero value is indicative of the appearance of a lithium deposit on the negative electrode. It can be seen that the passage below the zero value corresponds to the last maximum of the ratio R, preceding a sharp drop in the latter. In normal use of a battery, the potential Un of the negative electrode is generally not accessible. In particular, obtaining such a potential Un requires an invasive device in the battery, such as a reference electrode, which impacts the size of the battery. In addition, the reliability of the battery during its life may no longer be guaranteed.

Another example of a method according to the invention will now be described in relation to Figures 9 to 12. These figures show graphs acquired on a Li-ion battery of the NMC 811-graphite type. However, similar results can be obtained on a Li-ion battery of another type. The method makes it possible to detect the formation of lithium, in particular metallic lithium, on a negative electrode of a lithium-ion battery.

For this purpose, the method involves detecting the formation of lithium on a charging regime, called “basic regime” Ia. As for example illustrated in FIG. 9 , the basic regime Ia is applied to the Li-ion battery with the first charging current I _{I a} constant. Another charging regime, called “secondary regime” IIa, having _a constant charging current I _{II a} is further applied to the Li-ion battery.

The secondary regime IIa is chosen so that the charge rate C1 of the base regime Ia and the charge rate C2 of the secondary regime IIa are close. In this respect, the secondary regime IIa may have a higher or lower charge current I _2a than the charge current I _1a in the base regime Ia. In other words, the base regime Ia and the secondary regime IIa are chosen so that the absolute value of the difference C2-C1 between the charge rate C1 in the base regime Ia and the charge rate C2 in the secondary regime IIa is less than a threshold. For example, in FIG. 9 , the basic regime Ia has a charge rate of 1C and the secondary regime IIa has a charge rate of 1.25C.

The voltage U _Ia at the terminals of the Li-ion battery is acquired during the base regime Ia, as is the voltage U _IIa at the terminals of the battery during the secondary regime IIa. The charge quantity Q of the battery is also acquired, in particular in parallel with the acquisition of the voltages U _Ia , U _IIa at the terminals of the battery during the base regime Ia and the secondary regime IIa. As for example illustrated in FIG. 10 , we obtain in particular the voltage U _Ia at the terminals of the battery during the basic regime Ia and the voltage U _IIa at the terminals of the battery during the secondary regime IIa, as a function of the charge Q of the battery. From the acquired values, we determine, for a state of charge Q of the battery, a ratio R(Q) between on the one hand the difference ΔU between the voltage U _Ia at the terminals of the battery during the basic regime Ia and the voltage U _IIa at the terminals of the battery during the secondary regime IIa, and on the other hand the difference ΔI between the first charging current I _{l a} and the second charging current I _{ll a} :

There FIG. 11 illustrates the evolution of the ratio R as a function of the battery charge Q. Each curve is identified by the charge rate of the basic regime used to determine the ratio R. The shape of the graph makes it possible to detect the formation of a lithium deposit on the negative electrode for the basic regime. In particular, towards an increasing charge quantity, the curve corresponding to the basic regime 1C undergoes variations before reaching a local maximum P, then falling relatively sharply, which is significant of a formation of lithium on the negative electrode of the battery. Thus, we can determine the charge quantity from which there is lithium formation as a function of the value of the ratio R.

In particular, the amount of charge for which a lithium deposit appears corresponds to a last maximum, in particular a last local maximum, of the ratio R preceding a drop of at least 10% of said ratio. By considering a low drop value, the formation of lithium can be detected as early as possible, but there is then the risk of false detections due to intrinsic variations in the ratio curve. By taking a higher drop value, a false detection can be avoided, but possibly to the detriment of the accuracy of the detection of lithium formation.

The charge rate C1 of the basic regime Ia and that C2 of the secondary regime IIa are preferably close so that the detection of lithium formation by the ratio R is meaningful. For example, the absolute value of the difference between the charge rate C1 of the basic regime Ia and the charge rate C2 of the secondary regime IIa is less than 0.5 and preferably less than 0.25. However, the absolute value of the difference between the charge rate C1 of the basic regime Ia and the charge rate C2 of the secondary regime IIa is preferably large enough to give a meaningful result, taking into account the precision with which the voltages U _{I a} , U _{II a} are acquired.

The exemplary method is according to a second embodiment in which the basic regime Ia is applied to a first complete charge of the battery; and the secondary regime IIa is applied to a second complete charge of the battery. In other words, in the second embodiment, the basic regime Ia and the secondary regime IIa are applied during different charges. On the first charge, the voltage U _Ia at the terminals of the battery is acquired as a function of the charge quantity Q of the battery. On the second charge, the voltage U _{I I a} at the terminals of the battery is acquired as a function of the charge quantity Q of the battery. The method according to the second embodiment is particularly advantageous when it is possible to carry out a calibration of the Li-ion battery, either at the end of the battery production line, or on a vehicle test bench when the battery is mounted therein. Thus, with the method according to the second embodiment, it is possible to determine for the basic regime Ia, the battery charge from which the lithium deposit appears. This information may then be stored in a memory, in particular in a memory of a battery management system, for use during the life of the battery. For example, the information may be used to stop or change the charging rate, when the battery is charged with the basic rate and the battery charge reaches the previously identified charge, from which a lithium deposit is formed.

Preferably, the exemplary method is implemented for several basic regimes. For example, the FIG. 10 represents the voltage at the terminals of the Li-ion battery as a function of the quantity of charge Q stored in the battery, for several charging regimes. In particular, the FIG. 10 illustrates charging regimes with charging rates of 0.1C, 0.5C, 0.75C, 1C, 1.25C and 1.5C. For each charging regime applied, the ratio R is determined by taking as the basic regime the charging regime itself, and as the secondary regime, the lower or higher charging regime with the closest charging rate. For example, the ratio R for each charging regime illustrated in FIG. 10 , is represented in FIG. 11 , except for the regime with a charge rate of 0.1C.

In particular, the exemplary lithium formation detection method is effective for a charging regime having a charging rate greater than or equal to a threshold. For example, in FIG. 11 , the ratio R for the 0.5C charging regime does not exhibit a sufficiently marked drop in ratio to be significant, unlike the 0.75C, 1C, 1.25C, and 1.5C charging regimes. In particular, the efficiency threshold for an example of a Li-ion battery of the NNLIC 811-graphite type is 0.75C, whereas with the prior art method illustrated in FIG. 1 , detection of lithium formation is only possible for a charging regime having a charging rate greater than or equal to 1C. The exemplary method according to the invention therefore allows detection of lithium formation for a wider range of charging regimes than in the prior art. However, this efficiency threshold depends on the architecture of the battery, in particular the composition of the electrodes, and its state of aging.

For comparison, the FIG. 12 represents the potential Un of the negative electrode during charging for the charging regimes illustrated in FIG. 10 . The passage of the potential Un of the negative electrode below the zero value is indicative of the beginning of the formation of lithium on the negative electrode. We note that the passage below the zero value corresponds to the last maximum of the ratio R, preceding a sharp fall in it.

Thus, for each of the charging regimes applied, the respective charge quantity from which a lithium deposit is formed is obtained from the shape of the ratio R. This information can be stored in the form of a table in a memory, in particular a memory of a battery management system.

In particular, during a subsequent charge of the battery, the information stored in the table makes it possible to determine the occurrence of lithium formation as a function of the charging current, i.e. the charging rate, and the quantity of charge stored in the battery. Thus, before reaching the quantity of charge from which lithium formation occurs, a change can be made to a more advantageous charging regime, i.e. a charging regime for which the quantity of charge for which lithium formation occurs is higher.

The graphs illustrated in the figures are acquired at a constant temperature, such as 25°C. However, the temperature could be variable while still allowing application of the method according to the invention. The example methods have been described with loading rate values which are illustrative. Other loading rate values can be used while remaining within the spirit of the invention.

Claims (13)

Method for detecting lithium formation on an electrode in a lithium-ion battery, comprising detecting lithium formation for at least one charging regime 1a, Ia, called “basic regime”, comprising the steps of:

1. applying the basic regime 1a, Ia, having a first charging current I _1a , I _Ia , and a charging regime 2a, IIa, called “secondary regime”, having a second charging current I _2a , I _IIa , the absolute value of the difference between the charging rate C1 in the basic regime 1a, Ia and the charging rate C2 in the secondary regime 2a, IIa being less than a threshold;
2. acquire the voltage U at the terminals of the battery and the quantity of charge Q of the battery, during the basic regime 1a, Ia and the secondary regime 2a, IIa;
3. determining a ratio R between on the one hand the difference ΔU between the voltage U _1a , U _{I a} at the terminals of the battery in the basic regime 1a, Ia and that U _{2 a} , U _{II a} in the secondary regime 2a, IIa, and on the other hand the difference ΔI between the first charging current I _1a , I _Ia and the second charging current I _{2 a} , I _{I I a} , as a function of the quantity of charge Q; and
4. for the basic regime 1a, Ia, detect the appearance of lithium formation as a function of the value of said ratio R.

Method according to claim 1, in which the appearance of a lithium deposit corresponds to a last maximum P, in particular a last local maximum, of said ratio R, preceding a fall of at least 10% of said ratio R. Method according to claim 1 or 2, in which the absolute value of the difference between the load rate C1 of the basic regime 1a, Ia and the load rate C2 of the secondary regime 2a, IIa is less than 0.5 and preferably less than 0.25. Method according to one of the preceding claims, in which the application of the basic regime 1a and the secondary regime 2a is carried out alternately during a charge of the lithium-ion battery. Method according to the preceding claim, in which the secondary regime 2a is applied over negligible durations compared to the durations over which the basic regime 1a is applied. Method according to one of claims 4 to 5, comprising the emission of a lithium formation detection signal when the value of the ratio R becomes less than or equal to a threshold r _s . Method according to the preceding claim, in which said threshold r _s is a function of the quantity of charge Q of the battery and is equal to:

Where Q _tot is the maximum charge quantity of the battery in mAh, R ₀ is the initial value of the ratio when charging the battery from a state of charge less than 0.1Q _tot , α is a constant between 0.75 and 0.95. Method according to the preceding claim, comprising:

1. when the battery is charged from a state of charge less than or equal to 0.1Q _total :
   • determining the parameter R ₀ by calculating the average of said ratio R for quantities of charge Q less than or equal to 0.2Q _total ;
   • storing the parameter R ₀ in a memory; and
   • the use of the parameter R ₀ stored in memory for the determination of the threshold r _s for load quantities greater than 0.2Q _tot ,
2. or, when the battery is charged from a state of charge greater than 0.1Q _tot , the use of the last parameter R ₀ stored in memory for the detection of lithium formation during charging.

Method according to the preceding claim, comprising:

1. the comparison between the last two values of the parameter R ₀ each determined during a charge of the battery from a state of charge less than or equal to 0.1Q _tot , and
2. when the difference between the last two values of parameter R ₀ is greater than a threshold, the emission of a battery fault signal.

Method according to one of claims 4 to 9, in which the charging of the lithium-ion battery is a rapid charge having decreasing charging current levels, the method comprising the detection of the formation of lithium for a plurality of basic regimes 1a, 1b, each basic regime 1a, 1b corresponding to one of said current levels. Method according to one of claims 1 to 3, in which:

1. the application of the basic regime Ia is carried out on a first complete charge of the battery, the voltage U at the terminals of the battery being acquired as a function of the quantity of charge Q of the battery,
2. the application of the secondary regime IIa is carried out on a second complete charge of the battery, the voltage U at the terminals of the battery being acquired as a function of the quantity of charge Q of the battery.

Method according to the preceding claim, comprising:

1. the detection of a lithium deposit for a plurality of basic regimes;
2. storing in a memory a table comprising, for each of the plurality of basic regimes, the respective charge quantity from which a lithium deposit appears; and
3. upon subsequent charging of the battery, using said table to detect the occurrence of lithium formation on an electrode as a function of the charging current applied and the amount of battery charge.

Lithium-ion battery management system configured to implement a method for detecting a formation of lithium on an electrode according to one of claims 1 to 12 during a charge of the lithium-ion battery.