METHOD FOR PERFORMING THERMAL RUNAWAY TESTS ON AN ELECTRIC BATTERY

The method optimizes thermal runaway tests for electric vehicle batteries by reducing the number of destructive tests through a conditional logic sequence, ensuring efficient detection and compliance with regulatory standards.

FR3148301A1Inactive Publication Date: 2024-11-01STELLANTIS AUTO SAS
View PDF 1 Cites 0 Cited by

Patent Information

Application Number
FR2023004194
Authority / Receiving Office
FR · FR
Patent Type
Applications
Current Assignee / Owner
Filing Date
2023-04-26
Publication Date
2024-11-01
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

Conventional thermal runaway tests for electric vehicle batteries are costly and time-consuming, requiring multiple destructive tests to ensure the battery monitoring system can detect thermal runaway events across various charge states.

Method used

A method is developed to optimize thermal runaway tests by implementing a conditional logic sequence involving fewer tests, starting with a fully charged battery and adjusting the state of charge based on previous test outcomes, using a battery monitoring system to determine if a fire and alert are detected, thereby reducing the number of destructive tests required.

Benefits of technology

The method significantly reduces the budget and duration needed for thermal runaway testing while ensuring compliance with regulatory requirements by refining the test sequence to identify critical charge states where thermal runaway occurs and is detected reliably.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 00000000_0000_ABST
    Figure 00000000_0000_ABST
Patent Text Reader

Abstract

The invention relates to a method for performing thermal runaway tests on an electric battery monitored by a battery monitoring system, characterized in that successive tests are carried out on a predefined type of battery, using several examples of batteries of the predefined type, following a test sequencing logic, where the initial conditions of a test to be performed depend on the result of the previously performed test, particularly with regard to the initial state of charge of the battery, for which a bisection logic is used. Figure 4
Need to check novelty before this filing date? Find Prior Art

Description

Title of the invention: METHOD FOR CARRYING OUT THERMAL RUNAWAY TESTS CONCERNING AN ELECTRIC BATTERY

[0001] The invention relates to a method for implementing thermal runaway tests relating to an electric battery.

[0002] In an electrically powered / tracted motor vehicle, there is a traction battery which is designed to store a significant amount of electrical energy in electrochemical form.

[0003] In practice, we are talking about several tens of kWh. A 100% electric vehicle battery has an energy storage capacity typically between 50 kWh and 100 kWh, depending on the target autonomy, weight and consumption of said vehicle.

[0004] The most common traction batteries on electric vehicles are batteries based on Lithium-Ion type electrochemistry.

[0005] Among these, there are two most widespread types, namely NMC (Nickel Manganese Cobalt) and LFP (Lithium Iron Phosphate / LifePo4).

[0006] In an electric vehicle, the battery is protected from mechanical attack by structural elements of the vehicle, of which the floor of the vehicle is generally a part. The battery is also generally protected from the physicochemical environment, for example humidity, fog, liquid flow, etc.

[0007] Batteries are usually composed of a juxtaposition of unit electrochemical cells.

[0008] Furthermore, a battery monitoring system, referred to in the art as a 'BMS' (Battery Management System), is provided with the battery, the monitoring system being configured to acquire various information about the battery and its operating parameters: voltage, current, temperature, degassing, etc.

[0009] It turns out that in practice, for lithium-ion batteries, the risk of a phenomenon called "thermal runaway" cannot be completely eliminated.

[0010] This phenomenon of thermal runaway, also commonly called in practice "thermal runaway", occurs when an electrochemical cell reaches too high a temperature and begins to burn while communicating to the neighboring cells sufficient heat energy to increase the temperature of the neighboring cells so that they will also begin to burn, which gradually leads to the destruction of the battery.

[0011] On motor vehicles, all precautionary measures are taken to avoid at all costs an unwanted inflammatory or incandescent event for the entire vehicle, including the battery.

[0012] To this end, electric motor vehicles are equipped with high-performance cooling systems to cool the electrochemical cells of their battery. Regarding the battery, it is also planned to limit the stress applied to the battery if the temperature approaches a predefined threshold (this is called voluntary performance derating, also called 'derating').

[0013] Despite all the appropriate precautions taken, the risk of local heating or even a thermal runaway type phenomenon cannot be completely excluded, even if the probability is extremely low.

[0014] According to the regulatory aspects in force, it is necessary that if such an event occurs, the battery monitoring system ('BMS') can detect it in order to alert in time the occupants and / or owners of the vehicle in question, so that they can take all relevant and appropriate actions in view of the situation.

[0015] For approval, electric vehicle manufacturers are required to carry out real tests to show that either the thermal runaway phenomenon cannot occur, or if it does occur, that it is accompanied by a sufficiently early warning issued by the BMS.

[0016] Furthermore, the actual tests in question must provide the expected demonstration, regardless of the state of charge of the battery. To carry out such an actual test, one of the cells of the plurality of cells making up a battery is overheated, said cell being heated beyond its ignition temperature (called the 'onset' temperature).

[0017] Conventionally, around ten tests are carried out, namely for the following different battery charge states: 100%, 90%, 80%, 70%, 60%, 50%, 40%, 30%, 20%, 10%, which means the implementation of ten tests to test a type of battery with a type of BMS.

[0018] Conducting ten potentially more or less destructive tests represents a costly and time-consuming activity to implement.

[0019] A need arose to optimize the logic of implementing these tests in order to reduce their quantity and consequently reduce the budget to be mobilized. This led to the genesis of the present invention.

[0020] To this end, the invention proposes a method for implementing thermal runaway tests for an electric battery monitored by a battery monitoring system, characterized in that the tests are carried out on a predefined type of battery, having available several examples of batteries having the type of predefined battery, the following test sequence: - carry out an initial thermal stress test on a first battery sample with a state of charge greater than 95%, then: - if no fire occurs, then the test output is positive and the test sequence is declared complete, - if a fire occurs and no alert is declared by the battery monitoring system, then the test is declared failed and the test sequence is declared complete, - if a fire occurs while a related alert is declared by the battery monitoring system, then the test sequence is continued, - carry out a second thermal stress test, on a second battery example presenting a first intermediate state of charge, the second thermal stress test providing a result according to a conditional logic defined according to the four clauses which follow: — if no fire occurs while a related alert is declared by the battery monitoring system, then the test output is positive and the test sequence is declared complete, - if no fire occurs, and no alert is declared by the battery monitoring system, then the test sequence is continued with a new intermediate state of charge value higher by one increment, - if a fire occurs and no alert is declared by the battery monitoring system, then the test is declared failed and the test sequence is declared complete, - if a fire occurs while a related alert is declared by the battery monitoring system, then the test sequence is continued, with a new intermediate state of charge value lower by one decrement, - carry out a third thermal stress test, on a third battery example presenting a higher or lower intermediate state of charge depending on the output of the conditional logic of the previous test, by applying the conditional logic again, as a first loop, - carry out a fourth thermal stress test, on a fourth battery example presenting a higher or lower intermediate state of charge depending on the output of the conditional logic of the previous test, by applying the conditional logic again, as a second loop.

[0021] With such a tree logic arrangement, the desired result on the approval is obtained by implementing only four destructive tests or less, instead of ten destructive tests according to the conventional method. The budget is therefore significantly reduced as well as the time required.

[0022] The inventor noticed that the propagation of the thermal flux from the stressed critical cell was all the easier the higher the state of charge of the battery was. The inventor also noted that detecting a runaway phenomenon is easier when the battery's state of charge is high. In other words, the ease of thermal runaway increases with the state of charge and the reliability of detection also increases with the state of charge of the battery.

[0023] It is the two hypotheses above which made it possible to develop the test sequencing logic presented above.

[0024] Thus, for example, if for a given state of charge, the thermal stress of the critical cell does not cause the phenomenon of thermal runaway of the complete battery, then it is not necessary to carry out the test with lower states of charge because the phenomenon of thermal runaway will not take place either.

[0025] It is noted that the test sequence is continued when the runaway phenomenon occurs but it is correctly signaled by the BMS, the test sequence is continued with a lower charge level. Conversely when the runaway phenomenon does not occur and there is no alert either, the test sequence is continued with a higher charge level than the previous test.

[0026] It is noted that the demonstration concerns on the one hand the type of battery concerned but also the monitoring software of the monitoring system which goes with the battery concerned. It is therefore a pair (battery + BMS) for which the approval is proven.

[0027] For each test, we start with a battery sample in new condition, i.e. one that has not undergone one of the previous tests.

[0028] According to one embodiment, the method provides that the first intermediate state of charge is between 45% and 55%. According to one example, it is approximately 50%. This is the beginning of a dichotomy process, the continuation of which will be seen later. We place ourselves in the middle of the interval between 100% and 0%.

[0029] According to another embodiment, one could choose as the first intermediate value a value a little higher than 50% by focusing on the range of charge states far from zero.

[0030] According to one embodiment, the method provides an increment or respectively a decrement of increasing size for the new intermediate state of charge values ​​at each looping of the conditional logic. The logic is advanced step by step by refining the precision to identify the case where a fire actually occurs for a different state of charge.

[0031] According to one embodiment, the method provides that the increment or respectively the decrement has a value of half the value of the increment or respectively the decrement of the previous loop of the conditional logic. A pure dichotomy logic is thus applied. The following test is carried out in the middle of the uncertainty interval.

[0032] According to one embodiment, the method provides that the output of the fourth test is considered positive if the test sequence has not been declared terminated before by the conditional logic. According to this comma configuration the test is not continued beyond, and the sanction of the branches which would have led to continuing the test is declared positive.

[0033] According to one embodiment, the method may provide for a fifth test, similar to the previous tests, carried out following the fourth test. The results can thus be further refined in the area of ​​charge states between 75% and 100%.

[0034] According to one embodiment, the tests are carried out while the batteries under test are placed in the open air. The behavior of the battery can be more easily observed and only the battery is damaged at the end of the test.

[0035] According to an alternative embodiment, the tests are implemented while the batteries under test are placed inside a vehicle. The runaway test can in this case be more representative of the real life of the vehicle and its battery.

[0036] According to one embodiment, the method may provide that for the first, second, third and fourth tests, the battery is heated in a temperature range between 40°C and 50°C and a so-called critical cell is overheated in a temperature range between 130 and 140°C. The temperature known as the 'onset temperature' is thus exceeded to cause local ignition.

[0037] According to one embodiment, the so-called critical cell is located in an intermediate part of the battery or in a less well-cooled portion of the battery. This increases the representativeness of the runaway test.

[0038] The invention will be further detailed by the description of non-limiting embodiments, and on the basis of the appended figures illustrating variants of the invention, in which: [Fig.l] is a schematic representation giving an example of an electric vehicle with a traction battery, [Fig.2] is a schematic representation of an example of a battery formed by a juxtaposition of electrochemical cells, [Fig.3] shows a graphical example illustrating the dependence of the thermal runaway probability and the detection reliability on the battery state of charge. [Fig.4] represents an example of a flowchart of the test sequencing.

[0039] A motor vehicle 1 with electric propulsion / traction is now described with reference to [Fig.l]. The driving power can be transmitted to the front axle and / or to the rear axle of the vehicle.

[0040] The vehicle comprises a main battery pack marked 2 which is located in the floor structure of the vehicle. The main battery pack is simply called “battery” in the following. A charging socket 22 is provided through which it is possible carry out a standard recharge of the battery or in some cases a rapid recharge of the battery which places high thermal stress on the battery cells.

[0041] The vehicle includes a battery monitoring system 3.

[0042] The battery monitoring system 3 is referred to in the art as 'BMS' (of English: Battery Management System). The monitoring system is configured to acquire various information about the battery and its operating parameters: voltage, current, temperature, degassing, etc. The monitoring system is configured to deliver one or more alerts about the battery. There may be a first level of alert that prompts the driver to stop the vehicle. There may be a second level of alert that prompts the vehicle's occupants to leave the vehicle.

[0043] The vehicle 1 comprises an instrument cluster 4 which makes it possible to deliver graphic and / or textual messages to the driver and the vehicle may further comprise an audio system which makes it possible to deliver messages more generally to the occupants of the vehicles. It should also be understood that the vehicle can deliver a message remotely via the connectivity system of the automotive telematics system, to a connected portable terminal such as for example a smartphone.

[0044] The messages in question correspond to first or second level alerts generated by the battery monitoring system.

[0045] [Fig.2]

[0046] according to the example illustrated, the battery 2 is formed by a juxtaposition of unit electrochemical cells 5.

[0047] Among these cells, a so-called critical cell 51 is chosen to which an overheating will be applied, which is supposed to represent a potential source of thermal runaway of the battery.

[0048] As illustrated in [Fig.3], the ease of propagation of a thermal flux from the stressed critical cell is all the easier as the state of charge of the battery is high. This is represented in [Fig.3] curve 61. The reason is that cells containing high electrochemical energy are easier to trigger in terms of thermal runaway.

[0049] Similarly, but not identically, it turns out that the detection of a runaway phenomenon by the BMS battery monitoring system is all the easier the higher the state of charge of the battery. This is shown in [Fig.3] curve 62. The reason for this is that degassing, for example, is greater on a more charged cell, and is therefore easier to detect.

[0050] Expressed differently, the ease of thermal runaway increases with the state of charge and the reliability of detection also increases with the state of charge of the battery. However, curves 61, 62 cannot be known theoretically by simulation and in practice their position depends on a multiplicity of parameters, which makes the requirement to carry out real tests entirely legitimate.

[0051] The design of the battery, in particular the thermal insulation between the cells, as well as the software functions for detecting a runaway phenomenon, must be such that the curves are not too far apart or even better that the curve at 62 passes above the curve 61.

[0052] [Fig.4] illustrates the logic of implementing the tests. We start by carrying out a test with a battery sample in a fully charged state, e.g. in practice we check that the state of charge is at least 95%.

[0053] The first test is called TEST1 and to the right of the box in [Fig.4], the charge rate of the battery at the start of the test is indicated (here 100). The entire battery has a general temperature between 40°C and 50°C. One of the cells in the battery is caused to overheat, in this case a cell called a critical cell, which is heated to a temperature higher than its own runaway temperature (called 'onset temperature'). For example, this overheating temperature is between 130°C and 140°C.

[0054] In one example, the battery is in the open air and can be freely observed. In another example test configuration, the battery is located in a target vehicle with the target vehicle's battery monitoring system in question.

[0055] Thermal energy is applied to the critical cell using an electrical resistance carrying a current.

[0056] When the critical cell reaches the prescribed temperature, what happens at the level of the complete battery is monitored, and this over a long period of time, e.g. at least one hour, but in practice it is monitored for several hours.

[0057] The test sanction is as follows depending on what is observed.

[0058] In [Fig.4], a positive output of the tests is symbolized by a square surrounding a comma called "check" and failure was symbolized by a square surrounding a cross.

[0059] If no fire occurs (branch ST1-3), then the test output is positive and the test sequence is declared complete. There is no need to perform an additional test.

[0060] If no fire occurs (branch ST1-4), and in addition an alert is declared by the BMS 3, then the test output is also positive and the test sequence is also declared complete. There is no need to carry out an additional test. It is noted that it is possible to logically group the two aforementioned branches ST 1-3 and ST 1-4 into a single conditional output.

[0061] If a fire breaks out and no alert is declared by the monitoring system battery (ST1-1 branch), then the test is declared failed and the test sequence is declared complete. The pair (battery + monitoring system) does not pass the approval barrier.

[0062] If a fire occurs while a related alert is declared by the battery monitoring system (ST 1-2 branch), then the test sequence is continued by a second test which will be detailed in the following paragraphs. The runaway phenomenon has occurred but an alert has been correctly signaled, the additional test will be intended to verify that this occurs correctly for lower charge states.

[0063] We therefore take a new example of a battery in new condition of the type considered, and we subject it to a test called here second test, identified by the box TEST2. First, we prepare this second battery so that it has a prescribed state of charge, here at 50%. More generically, the second test is carried out with an intermediate state of charge noted SOCint.

[0064] As described above, the critical cell is overheated. Then what happens to the entire battery is monitored over a long period of time, e.g. at least one hour, but in practice monitoring is carried out for several hours.

[0065] The sanction of the second test is as follows depending on what is found. The following is called the four-clause conditional logic, it is used not only for the second test but also for subsequent tests.

[0066] If no fire occurs while a related alert is declared by the battery monitoring system (LC2-4 branch), then the test output is positive and the test sequence is declared complete. There is no need to carry out an additional test.

[0067] If no fire occurs, and no alert is declared by the battery monitoring system (branch LC2-3), then the test sequence is continued with a new intermediate state of charge value higher by one increment (INC2),

[0068] If a fire occurs and no alert is declared by the battery monitoring system (branch LC2-1), then the test is declared failed and the test sequence is declared complete.

[0069] If a fire occurs while a related alert is declared by the battery monitoring system (LC2-2 branch), then the test sequence is continued, with a new intermediate state of charge value lower by one decrement (DEC2).

[0070] Here INC2 and DEC2 are worth 25%. So for the third tests, we will have on one side SOCint+INC2 or 50+25 = 75 and on the other side SOCint-DEC2 or 50-25=25.

[0071] Generally speaking, an index i can be used to designate the date of the current test; in the example illustrated, the index i can be 2, 3 or 4. This index i is used for the naming of the branches and for the designation of the increments and decrements.

[0072] In the case where the sequence of tests is continued, a third test is then carried out. A new example of a battery of the type considered is therefore taken and it is subjected to a test called the third test here, identified by the box TEST3-1 or TEST3-2 depending on the result of the second test. First, this third battery is prepared so that it has a prescribed state of charge, here at 25% for the TEST3-1 branch and at 75% for the TEST3-2 branch, the two branches in question being mutually exclusive and one being chosen depending on the result of the second test.

[0073] As described previously, the critical cell 51 is overheated. Then what happens to the entire battery is monitored, over a long period of time, e.g. at least one hour, but in practice monitoring is carried out for several hours.

[0074] The sanction of the third test is as follows according to the conditional logic already explained above.

[0075] If no fire occurs while a related alert is declared by the battery monitoring system (LC3-4 branch), then the test output is positive and the test sequence is declared complete. There is no need to carry out an additional test.

[0076] If no fire occurs, and no alert is declared by the battery monitoring system (branch LC3-3), then the test sequence is continued with a new intermediate state of charge value higher by one increment (INC3),

[0077] If a fire occurs and no alert is declared by the battery monitoring system (branch LC3-1), then the test is declared failed and the test sequence is declared complete.

[0078] If a fire occurs while a related alert is declared by the battery monitoring system (LC3-2 branch), then the test sequence is continued, with a new intermediate state of charge value lower by one decrement (DEC3).

[0079] As apparent in [Fig.4], there are, at the end of the third possible tests, four configurations which lead to carrying out a fourth thermal runaway test.

[0080] Here INC3 and DEC3 are worth 12.5%. We therefore have for the step of the fourth test: a first possibility of fourth test called TEST4-1, with a state of charge for the test of 12.5%, a second possibility of fourth test called TEST4-2, with a state of charge for the test of 27.5%, a third possibility of fourth test called TEST4-3, with a state of charge for the test of 62.5%, a fourth possibility of fourth test called TEST4-4, with a state of charge for the test of 87.5%.

[0081] Of course, the fourth battery to be tested must be prepared so that it has a state of charge in accordance with the prescribed value resulting from the result of the third test (namely 82.5% or 62.5% or 37.5% or 12.5%).

[0082] As described above, the critical cell is overheated. Then what happens to the entire battery is monitored over a long period of time, e.g. at least one hour, but in practice monitoring is carried out for several hours.

[0083] The sanction of the fourth test is similar to what was presented previously, still based on the conditional logic already explained above.

[0084] Branch LC4-4 corresponds to the condition: If no fire occurs while a related alert is declared by the battery monitoring system.

[0085] Branch LC4-3 corresponds to the condition: If no fire occurs, and no alert is declared by the battery monitoring system.

[0086] Branch LC4-1 corresponds to the condition: If a fire breaks out and no alert is declared by the battery monitoring system.

[0087] Branch LC4-2 corresponds to the condition: If a fire breaks out while a related alert is declared by the battery monitoring system.

[0088] The four outputs LC4-4 correspond to a positive output and the test sequence is declared complete.

[0089] All four LC4-1 outputs correspond to a negative output and the test is declared failed.

[0090] For the four LC4-2 outputs and the four LC4-3 outputs, it is decided by convention not to continue the sequence with additional tests and to accept the result as positive (symbolized by a circle in a square in [Fig.4]).

[0091] Optionally, it may be provided, as shown in dotted lines in [Fig.4], to continue the test sequence with a fifth test, in particular for the range of the highest charge states, namely in the 60% to 100% zone.

[0092] According to one option, the critical cell 51 is located in an intermediate part of the battery. According to one option, the critical cell 51 is located in a less well-cooled portion of the battery.

Claims

1. Claims Method for implementing thermal runaway tests for an electric battery (2) monitored by a battery monitoring system (3), characterized in that the following test sequence is carried out on a predefined battery type, having several examples of batteries having the predefined battery type available: - carry out an initial thermal stress test (TEST1), on a first battery sample with a state of charge greater than 95%, then: - if no fire occurs (STl-3, STl-4), then the test output is positive and the test sequence is declared complete, - if a fire occurs and no alert is declared by the battery monitoring system (ST1-1), then the test is declared failed and the test sequence is declared complete, - if a fire occurs while a related alert is declared by the battery monitoring system (ST 1-2), then the test sequence is continued, - carry out a second thermal stress test (TEST2), on a second battery example presenting a first intermediate state of charge, the second thermal stress test providing a result according to a conditional logic defined according to the four clauses which follow: - if no fire occurs while a related alert is declared by the battery monitoring system, then the test output is positive and the test sequence is declared complete, - if no fire occurs, and no alert is declared by the battery monitoring system, then the test sequence is continued with a new intermediate state of charge value higher by one increment, - if a fire occurs and no alert is declared by the battery monitoring system, then the test is declared failed and the test sequence is declared complete, - if a fire occurs while a related alert is declared by the battery monitoring system, then the test sequence is continued, with a new intermediate state of charge value lower by one decrement, - carry out a third thermal stress test (TEST3-1, TEST3-2), on a third battery example having a higher or lower intermediate state of charge (75; 25) depending on the output of the conditional logic of the previous test, by applying the conditional logic again, as a first loopback, - carry out a fourth thermal stress test, on a fourth battery example having a higher or lower intermediate state of charge (87.5; 62.5; 27.5; 12.5) depending on the output of the conditional logic of the previous test, by applying the conditional logic again, as a second loopback.

2. Method according to claim 1, characterized in that the first intermediate state of charge is between 45% and 55%.

3. Method according to any one of claims 1 to 2, characterized in that an increasingly smaller increment or decrement is provided for the new intermediate state of charge values ​​at each looping of the conditional logic.

4. Method according to claim 3, characterized in that the increment or respectively the decrement has a value of half the value of the increment or respectively the decrement of the previous loop of the conditional logic.

5. Method according to any one of claims 1 to 4, characterized in that the output of the fourth test is considered positive if the test sequence has not been declared finished before by the conditional logic.

6. Method according to any one of claims 1 to 5, characterized in that a fifth test is provided, similar to the previous tests, carried out following the fourth test.

7. Method according to any one of claims 1 to 6, characterized in that the tests are carried out while the batteries under test are placed in the open air.

8. Method according to any one of claims 1 to 6, characterized in that for the first, second, third and fourth tests, the battery is heated in a temperature range between 40°C and 50°C and a so-called critical cell (51) is overheated in a temperature range between 130°C and 140°C.

9. Method according to claim 8, characterized in that the so-called critical cell (51) is located in an intermediate part of the battery or in a less well-cooled portion of the battery.

Citation Information

Patent Citations

  • Method and device for quantitatively evaluating thermal runaway risk of lithium batteries in different charge states

    CN114859232A