EP3052953A1

EP3052953A1 - Method and apparatus for evaluating the state of health of a lithium battery

Info

Publication number: EP3052953A1
Application number: EP14780469.4A
Authority: EP
Inventors: Jean-Michel VINASSA; Akram EDDAHECH; Olivier BRIAT
Original assignee: Centre National de la Recherche Scientifique CNRS; Universite de Bordeaux; Institut Polytechnique de Bordeaux
Current assignee: Centre National de la Recherche Scientifique CNRS; Universite de Bordeaux; Institut Polytechnique de Bordeaux
Priority date: 2013-10-01
Filing date: 2014-10-01
Publication date: 2016-08-10
Also published as: WO2015049300A1; JP6502331B2; FR3011393B1; US20160245876A1; FR3011393A1; JP2016539320A; US10036781B2

Abstract

Method for evaluation of the state of health of a lithium battery comprising: • a) a first step of recharging said battery at constant current, until the voltage (U) at its terminals reaches a limit value (Ucv); then • b) a second step of recharging said battery at constant voltage equal to said limit value, until the charging current (I) drops below a threshold value (I _min), a plurality of charging current measurements being acquired during said second recharging step; and • c) a step of estimating the state of health of said battery using at least one characteristic parameter of said second step of recharging at constant voltage, such as the parameter describing the decrease of an exponential function interpolating the charging current measurements, the energy supplied to said battery during said second step of recharging at constant voltage or the duration of said charging step at constant voltage. The invention also relates to an apparatus for implementing such a method and a battery management system comprising such an apparatus.

Description

METHOD AND APPARATUS FOR ASSESSING THE HEALTH STATUS OF A

LITHIUM BATTERY

The invention relates to a method for assessing the "state of health" of a lithium battery, and in particular the loss of capacity caused by the aging of such a battery.

The invention also relates to an apparatus for implementing such a method and to a battery management system incorporating such an apparatus.

The invention applies in particular, but not exclusively, to the field of batteries for powering electric or hybrid land vehicles.

Lithium batteries or accumulators - in their different variants such as "lithium-ion", "lithium-ion-polymer", "lithium-metal-polymer" etc. batteries. - are batteries with the highest energy density and the highest specific energy. It is therefore the technology of choice for powering electric or hybrid vehicles, but also many portable devices. However, it is known that these batteries have a deterioration of their performance - and in particular their capacity - over time, even during periods of non-use (so-called "calendar aging"). Therefore, the State of Health (or SOH) estimate of these batteries - quantified for example by the current reported capacity is either its displayed value ("commercial") or its measured value. in new condition - is one of the most important tasks of battery management systems ("Battery Management

System ", BMS) present in all electric or hybrid vehicles.

This task is not easy. Several techniques have been developed to accomplish it, but none gives full satisfaction.

Electrochemical impedance spectroscopy is a very useful technique for studying battery aging by monitoring the parameters of an impedance model. But it is complex to implement, expensive and does not allow access to capacity. Moreover, it can not be embedded in a BMS. See about it:

T. Hang, D. Mukoyama, H. Nara, N. Takami, T. Momma and T. Osaka, "Electrochemical Impedance Spectroscopy Analysis for Lithium-ion Battery Using Li4Ti5012 Anode," Journal of Power Sources, vol. 222, pp. 442-447, 2013; and

- A. Eddahech, O. Briat, H. Henry, Delétage, J.-Y., E. Woirgard and J.-M. Vinassa, "Aging monitoring of lithium-ion cell during power cycling tests", Microelectronics Reliability Journal, vol. . 51, No. 9-1, pp. 1968- 1971, 201 1.

Other methods, better adapted to online use, exploit the techniques of artificial intelligence, such as neural networks or fuzzy logic. See, for example, W.X. Shen, C.C. Chan, E.W.C.

Lo and K.T. Chau, "Energy Conversion and Management," vol.

43, no. 6, pp. 817-826, 2002.

These methods implement complex algorithms that require significant computing power. In addition, they require a long learning step.

Other techniques are based on the identification of the parameters of a model, for example by Kalman filtering. See, for example:

S. Wang, M. Verbrugge, J. S. Wang and P. Liu, "Multi-parameter battery state estimator based on the adaptive and direct solution of the governing differential equations", Journal of Power Sources, vol. 196, pp. 875-8741, 2011; and

A. Eddahech, O. Briat and JM Vinassa, "Real-Time SOC and SOH Estimation for EV Li-ion Cell Using Online Parameters Identification," in Proc. IEEE Energy Conversion Congress and Exposition, 2012, Raleigh, North Carolina, USA. These techniques use complex identification algorithms, requiring heavy digital processing. In addition, their implementation presupposes that a fine and precise model of the battery is available.

The document US 2001/0022518 teaches a method for estimating the state of health of a battery from the time U / 2 necessary for the charge current of said battery to be halved during the voltage phase. constant of a recharge of the type "constant current - constant voltage". This document shows, however, that the relationship between U / 2 and health status is not univocal.

Document FR 2 977 678 discloses a similar method, in which the state of health is estimated from the time required for the current to cross two thresholds - arbitrarily defined - during said constant voltage charging phase.

The invention aims to overcome the aforementioned drawbacks and to provide a method for evaluating the state of health of a lithium battery that is both simple to implement, reliable and accurate without lengthening the recharging phase nor cause additional aging.

According to the invention, this goal is achieved by estimating the state of health of a battery from the simple observation of the constant voltage step of its recharge. "Charging" refers to the operation of charging in a complete or near-complete manner (for example, 95% or more of available capacity) after a period of use, as opposed to partial "charges" that may occur in use (for example, in the case of an electric vehicle, during regenerative braking).

An object of the invention is therefore a method for evaluating the state of health of a lithium battery comprising:

a) a first step of recharging said constant current battery, until the voltage at its terminals reaches a limit value; then b) a second charging step of said constant voltage battery and equal to said limit value, until the charging current becomes lower than a threshold value; and

c) a step of estimating the state of health of said battery from said load current measurements acquired during said second constant voltage charging step.

The method may comprise in particular the acquisition of a plurality of measurements of the charging current during said second constant voltage charging step. In addition, said step c) may comprise the following substeps:

c1) estimating the decay constant B of a negative exponential function interpolating said load current measurements; and c2) estimating said state of health from said decay constant B.

According to particular embodiments of the invention:

Said sub-step c2 may be implemented by means of a linear function connecting said decay constant B to a loss of capacity of said battery.

Said battery can be a lithium-ion battery.

- More particularly, said battery can be a lithium-ion battery nickel, manganese and cobalt (NMC). In this case, the method may also include:

d) a step of detecting an inflection in a curve representing said load current measurements acquired during said second constant voltage charging step.

Another object of the invention is an apparatus for evaluating the state of health of a lithium battery comprising: a charger of constant current type - constant voltage, adapted to charge a said constant current battery until that the voltage at its terminals reaches a limit value, then at constant voltage and equal to said limit value until the charging current becomes lower than a threshold value; a device for monitoring the charging of said battery; and a treatment device data configured or programmed to cooperate with said charger and with said monitoring device to implement such a method.

Yet another object of the invention is a battery management system comprising such apparatus.

Other characteristics, details and advantages of the invention will emerge on reading the description given with reference to the accompanying drawings given by way of example and which represent, respectively:

Figure 1, the time evolution of the voltage and charging current of a lithium battery during a constant-current charging - constant voltage;

FIG. 2, the succession of steps of a method for evaluating the state of health of a battery according to one embodiment of the invention;

FIGS. 3A, 3B and 3C, the interpolation of the time evolution of the charging current during the constant voltage charging step by a negative exponential function for three battery technologies with different states of aging;

FIGS. 4A, 4B and 4C, the correlation between the decay parameter of said negative exponential function and the loss of capacity for the said three aforementioned battery technologies;

Figure 5, the correlation between the relative discharged capacity and the relative duration of the constant voltage charging phase.

FIG. 6, a graph illustrating the correlation between the constant voltage recharging relative energy and the capacitance loss for an aging lithium battery;

FIG. 7 is a block diagram of an apparatus for evaluating the state of health of a battery according to one embodiment of the invention, integrated in a battery management system; and

FIG. 8, curves illustrating the temporal evolution of the charging current during the charging of NMC batteries at different stages of aging. Lithium batteries are generally charged according to a so-called constant current mode - constant voltage (CC-CV, for the English expression "Constant Current - Constant Voltage"). This modality, illustrated in FIG. 1, consists in carrying out a first recharging phase during which a constant charging current l = 1cc is supplied to the battery, whose voltage U increases to a maximum value U _C v; then to carry out a second recharging phase during which the voltage U is kept constant and equal to Ucv, while the current I decreases. The recharge ends when the current I decreases to a minimum value lmin <lcc specific to the technology of the battery. The inventors have realized that the state of health of a battery can be reliably evaluated from one or more parameters characterizing this second step of constant voltage charging ("step CV"). Thus, the estimation of the state of health is done during the recharge of the battery - necessary for its normal use - without the need for dedicated measuring operations consuming resources and time.

The inventors are able to propose a probable explanation of the fact - found experimentally - that the observation of the constant voltage charging step provides sufficient information to evaluate the state of health of a battery. Indeed, one of the major mechanisms responsible for the degradation of the capacity of a lithium battery over time is the formation of a solid electrolyte interface (SEI, for "Solid Electrolyte Interface") which is an obstacle to the intercalation of lithium ions in the material of the anode and the cathode. However, this intercalation occurs essentially during the constant voltage step of the recharge.

FIG. 2 illustrates the succession of steps of a method according to the invention:

The first step of constant current charging;

- The second step of constant voltage charging; Measuring, determining or estimating at least one characteristic parameter of this second step of constant voltage charging;

The determination of at least one SOH indicative of the state of health of the battery (for example, its capacity related to the capacity of said battery in new condition, or its advertised capacity) from said or at least one said parameter. This last step is made possible by a prior calibration step in which a relationship is established between said or each parameter and the state of health of the battery. Calibration requires a reference method for determining the state of health of the battery. This method can be, for example, a capacity measurement performed during a complete discharge of the battery (measurement of "discharged capacity").

For the determination of the state of health of the battery is reliable, it is supposed that the recharging is carried out at a controlled temperature (for example 25 ° C) or at least known (in this latter case, the calibration must allow to take into account the effect of temperature on the relationship existing between the parameter characterizing the step of charging at constant voltage and the state of health).

According to one embodiment of the invention, the parameter characterizing the step of charging at constant voltage is the decay parameter B of a negative exponential function

l (t) = Ae ^"Bt + C

interpolating the evolution of the load current measured I as a function of time t (t = 0 corresponding to the beginning of the step of charging at constant voltage). FIGS. 3A-3C make it possible to check the quality of such an interpolation, the continuous curves representing the interpolation function being practically superimposed on the measurement points. The three curves presented in these figures relate to three lithium-nickel cobalt aluminum battery technologies (NCA, Fig. 3A), nickel cobalt nickel (NMC, Fig. 3B), lithium manganese oxide (LMO, Fig. 3C) - each being in a distinct state of health. The parameter B can be connected to the loss of capacity of the battery, expressing its state of health, by a linear function, as illustrated by FIGS. 4A, 4B and 4C which correspond to the three curves of FIGS. 3A, 3B and 3C respectively . These three examples are not limiting. In any case, the quality criterion R ² of the least squares identification is very close to 1, which is very satisfactory.

The parameter B proves to be a much better indicator of the state of health of a battery than the time U / 2 necessary to divide the charging current by 2, used for example in the document US 2001/0022518 mentioned above. Indeed, as mentioned above, the relationship between U / 2 and the state of health of the battery is neither linear nor unambiguous (two different health states can be associated with the same value of ti / ₂ ).

In the case of a battery type LFP (lithium iron phosphate), the state of health can more simply be determined from the measured duration of the step of charging constant voltage. More precisely, it is found that this measured duration, relative to its duration during the initial recharge, is proportional to the state of health of the battery. Figure 5, which relates to the case of a Lithium Iron Phosphate (LFP) technology battery, is a graph showing:

a first curve representing the evolution, during aging, of the state of health SOH of such a battery - expressed by the discharged capacity of the battery compared to its value when new; and

a second curve representing the evolution of the duration Tcv of the constant voltage phase of the refill - also referred to its value when new.

We can verify that the curves are very close. Therefore, in this case, an estimate of the state of health can be deduced directly from the duration of the constant voltage charging phase relative to its value when new.

Whatever the embodiment considered, the relationship between the characteristic parameter of the step of charging at constant voltage (parameter B or duration of the step) need not necessarily be expressed by a linear or nonlinear mathematical function; it can also be, for example, a correspondence table.

In the case of a battery NMC type (lithium-ion nickel, manganese and cobalt), in addition, an advanced state of aging can be identified by detecting a deviation of the curve l (t) in phase CV compared to a decreasing exponential. This difference takes the form of an inflection (change of convexity) or "hump" which occurs after a phase of rapid decay of the current. This effect is illustrated in FIG. 8, where the curves CUO, CU1, CU2 and CU3 correspond to different aging states of a Li-ion battery NMC. More precisely, the curve CUO corresponds to the new battery, the curve CU1 to the battery after a storage of 770 days at maximum load and at a temperature of 45 ° C, the curve CU2 to the battery after a storage of 920 days in the same conditions and the curve CU3 to the battery after a storage of 1060 days, always under the same conditions. The curves have been temporally offset to facilitate their identification. Note that, in the case of the new battery (CUO curve), the decay of the charging current during the CV phase is well described by an exponential law characterized by a certain value of the decay parameter B. As and when As the battery ages, the parameter B decreases (the decrease in current is slower), but especially we notice the appearance of a "bump" increasingly pronounced, which makes the exponential law less relevant. It can be considered that the curve CU3 - for which said exponential law is clearly not a satisfactory approximation - corresponds to the end of the useful life of the battery.

Thus - in addition to or in replacement of the estimation of the state of health of the battery from the parameter B - an advanced state of aging can be identified by detecting the appearance of an inflection in the curve l (t) during the CV phase of the refill. This detection can be performed by simple observation of the curve or, preferably, automatically, by calculating a parameter representative of the deviation of said curve of a decreasing exponential and comparing the value of this parameter with a value of reference. This parameter can be, for example, a quadratic difference between the measured current and its interpolation by an exponential function.

It can be assumed that the inflection in the decay of the current during the CV phase of the recharge is due to a slowdown in the insertion of lithium ions caused by the growth of a solid electrolyte layer at the level of the 'anode.

FIG. 7 illustrates the block diagram of an apparatus according to one embodiment of the invention, integrated in a BMS battery management system, for example in an electric or hybrid vehicle. The device includes a conventional constant current type CHG charger - constant voltage, charging a BATT lithium battery; a DSC load monitoring device and a data processing device, or processor, PR. The DSC monitoring device comprises, for example, a current sensor, for measuring the load current I, and a voltage sensor, for measuring the voltage U at the terminals of the battery BATT. This device can be integrated with the CHG charger or the BATT battery. The data processing device PR (preferably a digital processor or an electronic card comprising such a processor) is programmed and / or configured to receive the measurements from the load monitoring device and use it to calculate a SOH indicator the state of health of the battery as described above. The processor PR can also control the charger CHG, for example by controlling the transition between the first constant-current charging step and the second constant-voltage charging step, as well as stopping the charge when l (t) reaches its peak. minimum value Un-

Claims

1. A method of assessing the health status of a lithium battery (BATT) comprising:

a) a first charging step of said constant current battery, until the voltage (U) at its terminals reaches a limit value (U _C v); then

b) a second step of recharging said battery with constant voltage and equal to said limit value, until the charging current (I) becomes less than a threshold value (l _min ), a plurality of measurements of the current charging being acquired during said second recharging step; and

c) a step of estimating the state of health of said battery from said load current measurements acquired during said second constant voltage charging step;

characterized in that said step c) of estimating the state of health of said battery comprises the following sub-steps:

2. Method according to claim 1 wherein said substep c2 is implemented by means of a linear function connecting said decay constant B to a loss of capacity of said battery.

3. Method according to one of the preceding claims also comprising a prior calibration step, comprising the determination of a relationship linking said or each said characteristic parameter of said second step of constant voltage charging to said state of health of the battery.

4. Method according to one of the preceding claims, wherein said battery is a lithium-ion battery.

5. The method of claim 4 wherein said battery is a lithium-ion battery nickel, manganese and cobalt, the method further comprising:

6. Apparatus for evaluating the state of health of a lithium battery comprising:

a charger (CHG) of constant current type - constant voltage, adapted to charge a said constant current battery (BATT) until the voltage at its terminals reaches a limit value, then at constant voltage and equal to said value limit until the charging current becomes lower than a threshold value;

a monitoring device (DSC) for recharging said battery; and

- A data processing device (PR) configured or programmed to cooperate with said charger and with said monitoring device to implement a method according to one of the preceding claims.

Battery management system (BMS) comprising an apparatus according to claim 6.