GB2583510A

GB2583510A - A method and apparatus for detecting defective cells within a battery

Info

Publication number: GB2583510A
Application number: GB1906150.6A
Authority: GB
Inventors: John Hughes Timothy
Original assignee: Siemens PLC
Current assignee: Siemens PLC
Priority date: 2019-05-02
Filing date: 2019-05-02
Publication date: 2020-11-04
Anticipated expiration: 2039-05-02
Also published as: GB2583510B; DE102020111788A1; GB201906150D0

Abstract

A method of detecting defective cells in a battery by measuring a characteristic of a plurality of cells as a time-series at multiple times, arranging results for a certain time in a histogram, comparing histograms of different times and identifying cells results that have changed positions or that assume extreme positions in the distribution as deteriorating. The amount a cell needs to move to qualify as degrading may be based on a threshold in sigma, percentage change in peak value or parameter value or position within a ranking according to magnitude. Cells identified as deteriorating may be replaced or repaired. The characteristic measured may be open-circuit voltage, internal resistance, cell current, charge capacity, state of charge, temperature, or an electrochemical parameter. Deteriorating cells may be identified when the distribution of a histogram changes from a normal to a log-normal distribution. The method may be executed by an apparatus with monitoring equipment for measuring cell parameters, a processor to generate a histogram, a memory, and a comparison unit that outputs an indication of cells that are identified as deteriorating.

Description

A METHOD AND APPARATUS FOR DETECTING DEFECTIVE CELLS WITHIN A BATTERY.

The present invention relates to methods for detecting the 5 deterioration of cells within a battery, so that preventative maintenance or replacement of individual cells can be planned.

By detecting the gradual deterioration of a cell, that cell may be repaired or replaced at a convenient time, before the 10 deteriorating cell causes malfunction of the whole battery.

Energy storage in electrical batteries is currently of increasing importance in applications such as grid stabilisation, cars, buses, trains, ferries. Battery packs 15 used in such applications are made up of very many cells.

Such cells degrade over time, typically either exhibiting increasing internal resistance, or decreasing charge storage capacity.

For this reason, a battery designed for a specific task must be allocated an expected lifetime, and designed such that its ability to store energy at the end of its planned lifetime is sufficient to meet a predetermined demand. Conventionally, this has meant that batteries tend to be over-specified at the start of their life cycle, to ensure that they remain sufficient at the end of their life cycle. Lifecycle estimation and lifecycle metrics accordingly are of great significance in battery business cases, to enable a designer to determine the required capacity at the start of a battery's life, such that it remains sufficient at the end of the battery's life.

In conventional methods and systems for battery condition monitoring, a battery management system measures the capacity of the whole battery, and measures cell voltages of each cell, so as to indicate when a cell has a voltage falling outside of a certain desired range, or when the battery as a whole has a charge capacity which has fallen below a certain minimum tolerable level. Alternatively, or in addition, the battery management system may measure the temperature of each cell, and may indicate when a temperature of a cell exceeds a certain maximum value.

The present invention aims to provide a battery condition monitoring system which provides an earlier indication of a degrading cell than was possible in the prior art. Indeed, the present invention aims to identify a degrading cell even before the cell performance falls outside of an acceptable range. The present invention also aims to provide a self-calibrating method for identifying failing cells within a battery.

The present invention accordingly provides methods and 15 apparatus as defined in the appended claims.

The above, and further, objects, characteristics and advantages of the present invention will become more apparent on consideration of the embodiments described below, by way of non-limiting examples only, together with the appended drawings, in which: Figs. 1A and 1B illustrate histograms of cell parameter measurements taken at different times during the useful life 25 of a battery, to illustrate a method according to an embodiment of the present invention; Fig. 2 illustrates a histogram of cell parameter measurements taken at different times during the useful life of a battery, to illustrate a method according to another embodiment of the present invention; Figs. 3A and 3B illustrate histograms of cell parameter measurements taken at different times during the useful life of a battery, to illustrate a method according to a further embodiment of the present invention; and Fig. 4 schematically illustrates battery condition monitoring apparatus 40 according to an embodiment of the present invention.

According to an aspect of the present invention, characteristics of the various cells of a battery are measured, each as a time-series. For each time-point, and for each characteristic in examples where more than one characteristic is measured, the results are arranged within a histogram. The evolution of that histogram over time is used to identify failing cells, preferably even before the failing cell even falls outside of acceptable operational values.

Embodiments of the present invention detect when a cell's measured value of a characteristic changes position within the histogram, or when the histogram changes shape. If the changes in position or shape are significant enough (as defined by one or more predetermined parameter), then the method and/or system of the present invention will identify the cells contributing most significantly to the change, and will indicate those cells as deteriorating. A maintenance operation may then be scheduled for a convenient time to repair or replace the affected cells, while ensuring that the battery remains in an 20 acceptable state for usage at all times.

Certain specific embodiments will be described, by way of non-limiting examples only.

In certain embodiments, the cell voltage and cell current are measured for each cell. The state of charge for each cell is also recorded. Each of these measurements is repeatedly performed, and the results then collated into a time-ordered series. In certain embodiments, these time series of data are used, together with respective cell state-of-charge and respective cell open circuit voltage to estimate cell circuit parameters such as charge capacity, internal cell resistance. Parameter estimation techniques may be used, such as Kalman filter, extended Kalman filter, Sigma Point Kalman Filter.

Once calculated, the parameter values for each parameter and each cell, at respective time points, are arranged within a histogram. In some embodiments, the histogram is fitted to a distribution.

Fig. lA shows an example histogram for illustrating this stage in the method of the present invention. This histogram represents measured parameter values of respective cells in a battery at the beginning of their service life. In this example, the histogram of parameter values (in this example, internal resistance) fits a normal distribution with a standard 10 deviation interval of x sigma.

The resulting histogram can be fitted to a distribution. It has been found that, normally, at the start of a battery's useful life, the cell parameters will be arranged in a normal distribution. As the cells age during operation, the distribution can change in mean, variance or other Parameters, or can even change in overall shape: for example, a normal distribution may change into a log-normal distribution. Such changes indicate that the cells are not all aging at a same rate, but some cells are behaving differently from the behaviour of the general population.

The parameter value for each cell (here, internal resistance) is remeasured periodically and the histogram replotted. During the lifetime of the battery, the respective values of the measured parameter of the cells will change. According to the present invention, the change in position within the histogram of a measurement corresponding to a certain cell is detected and taken as an indication that the associated cell is deteriorating. While internal resistance is considered here, as an example parameter for monitoring according to the present invention, other parameters may of course, alternatively or in addition, be employed. Suitable such parameters include cell open-circuit voltage, cell charge capacity, cell current, cell state-of-charge, cell temperature, or model parameters from a cell equivalent circuit model. The attached appendix explains example equivalent circuit models, and parameters from such models may be used as the measured parameter in a method according to the present invention. For example, parameters such as R1, R2, Cl, C2 of a Warburg impedance model, as explained in the appendix. The document forming the appendix may be found at-java.uccs.c Other parameters which could be monitored as the paramenter in a method of the present invention include electrochemical parameters such as lithium inventory.

Fig. 1B shows a histogram corresponding to the histogram of Fig. 1A, but taken after a certain amount of ageing by the battery. The normal distribution has widened; the peak distribution at the mean parameter value has decreased. On both the upper and lower extremes of the parameter value (here, internal resistance), the parameter values for certain cells are found outside of the x sigma interval defined by the histogram of Fig. 1A.

Those certain cells are taken to be degrading, since they exhibit parameter values outside of the interval defined for the battery cells at the beginning of their service life. The corresponding cells can be identified by their parameter value, and a maintenance operation may be scheduled to repair or replace those certain cells.

Fig. 2 shows an analysis performed according to a method of another embodiment of the present invention. Similarly to the embodiment illustrated with reference to Figs. lA and 1B, a histogram of parameter values of respective cells is formed, for each time point. At a first time point, in this example at the beginning of the service life of the battery cells, the parameter value for a certain cell (in this example, the internal resistance of the cell labelled 107) appears at the mean value, in this example being at the peak of the distribution curve.

A similar histogram, formed of parameter values of respective cells, is formed for values measured after some deterioration in the battery. The measurement associated with the cell labelled 107 is no longer located at the mean value; at the peak of the distribution curve, but is now located in the tail of the distribution. According to an embodiment of the present invention, this relative movement of the measurement for a particular cell may be taken to indicate a deterioration in the associated cell, which may then be scheduled for replacement. The overall distribution may remain a normal distribution, and the cell labelled 107 may not have a parameter value falling outside of the x sigma interval determined by the histogram at the start of the battery's service life, but the fact that the associated parameter value has moved location within the associated histogram may be taken as an indication that the cell is degrading, and replacement or repair may be scheduled as appropriate.

With the method of this embodiment of the invention, there is no need to consider any change in the standard deviation interval of the distribution or peak frequency value of the histogram -a simple comparison is made of the relative position of the measurement for each cell, within the distribution as a whole, and any significant movement in position of the measurement for a particular cell within the distribution is taken as an indication of the deterioration of that cell. Replacement of any cells identified in this manner may be planned for a convenient time. Such detection and replacement of a deteriorating cell may be performed without the battery's condition falling outside of predefined permissible range. The definition of "significant" movement may be indicated in terms of sigma, in terms of percentage change in peak value or earlier value of the measured parameter, or in a position of a particular cell within cell measurements ranked in order of magnitude.

Figs. 3A, 3B illustrate a method according to another embodiment of the present invention. In this embodiment, there is no need to identify the position of respective measurements for each particular cell in the original histogram distribution (Fig. 3A) which represents the state of the cells at the start of their useful life. At time intervals during the useful life of the cells, the parameter of the cells is re-measured (in the illustrated example, the measured parameter is internal resistance). Those measurements are again plotted in a histogram, and that histogram fitted to several different distributions, to find a distribution which matches the re-measured histogram. In the example shown in Fig. 3B, the histogram of internal resistance measurements has changed 10 shape compared to the histogram of Fig. 3A at the start of the useful life of the cells.

As illustrated in Figs. 3A and 3B, the distribution fitting to the histogram has changed from a normal distribution in Fig. 3A to something of a long-tailed log-normal distribution in Fig. 3B. Those cells represented by measurements most remote from the peak histogram value (shaded in Fig. 3B), which may be represented by a sigma value in excess of a certain threshold, may be identified for repair or replacement. Even if operation of the battery as a whole may not have exceeded defined limits, failing cells may be identified and repaired or replaced at a convenient time, so as not to interrupt operation of the battery.

According to the present invention, cells within a battery are identified for intervention, typically repair or replacement, by measuring a certain characteristic and ordering the measured values of the cells into a histogram. The measuring and ordering into a histogram are repeated at certain time intervals, and cells whose position within the histogram change over time are identified for intervention. This enables cells which are deteriorating to be identified before the deterioration has become so bad that overall operation of the battery falls outside of an acceptable range. Contrary to conventional methods and equipment for cell health monitoring, the accuracy of cell health calculation is not so relevant. The characteristics of the cells are compared against the corresponding characteristics of other cells, and it is the relative movement of the cell values within the histogram which is used to identify deteriorating cells for intervention, which may be replacement or repair of the identified cells. Such measurement is in effect self-calibrating as only a relative measurement is involved.

Fig. 4 schematically illustrates battery condition monitoring apparatus 40 according to an embodiment of the present invention, which may be used to implement a method according 10 to an embodiment of the present invention.

As illustrated in Fig. 4, a battery of cells 42 is provided. Monitoring equipment 44 takes current and voltage measurements from the cells within the battery. The monitoring equipment uses these current and voltage measurements, and possibly other measurements which may also be taken, and uses them to calculate parameter values which are then supplied to a data processor 46. The data processor 46 obtains the parameter values and generates a corresponding histogram 48. Histogram 48 may be stored in a memory 50. Histogram 48 may also be compared by a comparison unit 52 to an earlier histogram which has already been stored in the memory 50. This comparison unit performs a comparison method according to an embodiment of the present invention and outputs an indication 54, indicating any cells which are determined to be deteriorating according to the method applied by the comparison unit 52.

Preferably, the battery condition monitoring apparatus 40 as described above with reference to Fig. 4 is implemented in a programmable digital computer, and the present invention extends to such programmable digital computers programmed to perform the method of the present application, or to embody a battery condition monitoring apparatus according to an embodiment of the present invention. The present invention also extends to data storage devices loaded with data representing a computer program which, when loaded onto such a programmable digital computer, will cause said programmable digital computer to perform the method of the present application, or to embody a battery condition monitoring apparatus according to an embodiment of the present invention.

Claims

CLAIMS1. A method for detecting defective cells within a battery, wherein a characteristic of each of a plurality of cells of a battery is measured, as a time-series of values at a plurality of time-points; for each time-point, the results are arranged within a histogram; cells whose measurement significantly changes position within the histogram between time-points are identified as deteriorating.
2. A method according to claim 1, in which a cell's measurement is deemed to have changed significantly if a change is detected in excess of a predetermined threshold in terms of sigma; in terms of percentage change in peak value or earlier value of the measured parameter; or in a position of a particular cell within cell measurements ranked in order of magnitude.
3. A method for detecting defective cells within a battery, wherein a characteristic of each of a plurality of cells of a battery is measured, as a time-series of values at a plurality of time-points; for each time-point, the results are arranged within a histogram; where the histogram changes shape between time-points, cells whose measurement assume an extreme position within the histogram are identified as deteriorating.
4. A method according to claim 1, further comprising the step of repairing or replacing the cells identified as deteriorating.
5. A method according to any preceding claim, wherein the characteristic is one of: cell open-circuit voltage; cell current; internal cell resistance; cell charge capacity, cell current, cell state-of-charge, cell temperature, parameters from an equivalent circuit model; or an electrochemical parameter.
6. A method according to claim 1, wherein a histogram is compared to an earlier histogram, and cells are identified as deteriorating where the parameter values are found outside of an x sigma interval defined by the earlier histogram.
7. A method according to claim 1, wherein a histogram is compared to an earlier histogram, and cells are identified as deteriorating where the associated parameter value has moved location within the associated histogram.
8. A method according to claim 3, wherein the distribution fitting to the histogram is determined to have changed from a normal distribution to a long-tailed log-normal distribution, and those cells represented by measurements most remote from the peak histogram value, are identified as deteriorating.
9. A battery condition monitoring apparatus for implementing a method according to any preceding claim, comprising: - monitoring equipment (44) for measuring current and voltage in cells within a battery and for calculating parameter values; -a data processor (46) arranged to obtain parameter values from the monitoring equipment (44) and to generate a corresponding histogram (48); - a memory (50) for storing the histogram; - a comparison unit (52) arranged to perform a comparison method according to an embodiment of the present invention and arranged to outputs an indication (54) to indicate any cells which are determined to be deteriorating according to a method according to any preceding claim, applied by the comparison unit (52).
10. A battery condition monitoring apparatus according to claim 9, further arranged such that the histogram (48) may also be compared by the comparison unit (52) to an earlier histogram stored in the memory (50).
11. A battery condition monitoring apparatus according to claim 9 or claim 10, implemented in a programmable digital computer, programmed to perform the method of any of claims 1-8.
12. A data storage device loaded with data representing a computer program which, when loaded onto a programmable digital computer, will cause said programmable digital computer to perform a method according to any one of claims 1-8 or to embody a battery condition monitoring apparatus according to any of claims 9-10.