Washington, Sept 29th, 2014

1. Informal working group: Large Lithium batteries testing RECHARGE & SAFT Washington, Sept 29th, 2014

2. Content: • 38-3 Definitions. • Large batteries tests limit of 6200 Wh. • Limit of tolerance for mass in shock test.. • Amendment for large cells in small batteries (if needed). • Amendment for T5 test completion conditions: precisions • .

3. 38-3 DEFINITIONS table Table describing the existing testing requirements: 1- A single cell battery is different from a cell only if it is equipped with an overcharge protection: in this case it should be tested with T7. In other cases, the tests applicable are the one of a cell. This should be reflected in the definition. -2- A component cell transported separately from battery is tested exactly as a cell, so there is no need to create a specific case. On the contrary, it is a specific case when the component cells are transported only in a battery. It should be clarified in the definition that a component cell cannot be proposed for transport separately from a battery.

4. 38-3 DEFINITIONS 38-3-2-1

5. 38-3 DEFINITIONS 38-3-2-3

6. Situation to be consdered for 38.3.3 (f) and last § of 38.3. Existing situation 6200 Wh limit? 12 kg limit? • Is it a need to maintain 2 specifiedlimits? (to beconsidered if removed the weightlimit is from the shock tests?)

7. Issues and proposal for change of 38.3.3 (f) ISSUES: 1- Transport safety protection is unevenly specified for Battery assemblies depending on their energy: • Over 6200 Wh, is specified a monitoring system (most of the time not active during transport), but no test. • Under 6200 Wh, a complementary test is specified, but no monitoring system. 2- Significant cost and duration of cycling for assembled batteries under 6200Wh, for a test which is duplicating the test being already requested at battery level. Proposal:suppress the cycling requirement for assembled batteries. “(f) When testing a battery assembly in which the aggregate lithium content of all anodes, when fully charged, is not more than 500 g, or in the case of a lithium ion battery, with a Watt-hour rating of not more than 6200 Watt-hours, that is assembled from batteries [or single cell batteries] that have passed all applicable tests, one assembled battery assembly in a fully charged state shall be tested under tests T.3, T.4 and T.5, and, in addition, test T.7 in the case of a rechargeable battery assembly. A rechargeable battery shall have been cycled at least 25 cycles.”.

8. Issue and Proposals for change of 38.3.2.2 ISSUE: In case of application in T5 of an acceleration depending on the battery mass, there is a need to define a mass threshold allowing the qualification by equivalence for batteries of similar weight. Proposal:add in 38.3.2.2 (c ): a battery mass threshold for qualification by equivalence. 38.3.2.2 (c) A change that would lead to failure of any of the tests, shall be considered a new type and shall be subjected to the required tests. NOTE: The type of change that might be considered to differ from a tested type, such that it might lead to failure of any of the test results, may include, but is not limited to: (a) A change in the material of the anode, the cathode, the separator or the electrolyte; (b) A change of protective devices, including hardware and software; (c) A change of safety design in cells or batteries, such as a venting valve; (d) A change in the number of component cells; and (e) A change in connecting mode of component cells. (f ) A change in the battery mass of more than 20%. In the event that a cell or battery type does not meet one or more of the test requirements, steps shall be taken to correct the deficiency or deficiencies that caused the failure before such cell or battery type is retested.

9. Proposal for change of definition of “large batteries”, (depending on shock test discussion) Issues for T4: Shock test Example battery of 3 cells with a mass > 500g and the total mass of the battery < 12kg: it is a “small battery” assembled with “large cells”. • Anomalies between the shock test parameters between cells and battery: acceleration for small batteries not adapted for large cells.(explain #g acce 50g 150gl) Proposal: amendment of 38.3.2. complement in the definition of « large batteries »= « large battery means a lithium metal battery or lithium ion battery with a gross mass of more than 12kg. » Any battery assembled with large cells shall be considered as a large battery for testing purpose.

10. Proposal for change of the T.5: external short circuit. • AmendmentProposal: for §38.3.4.5.2 Test procedure. The cell or battery to be tested shall be heated to reach an homogenous stabilized temperature of stabilized so that its external case temperature reaches57 oC ±4 oC ,measured on the external case. The time required for achieving a stabilized temperature shall be assessed and documented depending on its size and design of the cell or battery. If this assessment is not feasible, the exposure time shall be at least [6]hours for small cells and small batteries, and [12] hours for large cells and large batteries. Then the cell or battery at 57oC ±4 oC shall be subjected to a one short circuit condition with a total external resistance of less than 0.1 ohm at 55 ±2 oC. This short circuit condition is continued for at least one hour after the cell or battery external case temperature has returned to 57oC ±4 oC, or in the case of the large batteries, has decreased of half of the maximum increase observed during the test, provided the temperature during the cooling phase has been steadily decreasing [not re-increased by more than 5°C].”. Textcommentededitiorial in red

11. Concerns raised at UN SCTDG about T.5 amendents: • UK: measuring on the external case cannot say if it is homogeneous: other place for measurement? • Proposal to explain that internal measurement is not feasible. But homogeneity can be controlled through minimum time and/or surface temperature stabilization. Add ”for a minimum time” in the text. • Netherlands: Is it risk of fooling the test, as the ambient temperature is not mentioned when the short circuit test is started? • Proposal to explain that because the temperature is homogeneous when the battery is removed out of the thermal cabinet, then the surface temperature will be the first cooling place of the battery. Consequently, as the surface temperature is specified, there is no risk to fool the test. No text change needed? • Belgium/China: propose for simplification to remove the part of the sentence about « temperature re-increase«. • Agreed

12. Proposal for T.5 amendents: The cell or battery to be tested shall be heated for a minimum time to reach an homogenous stabilized temperature of stabilized so that its external case temperature reaches57 oC ±4 oC ,measured on the external case. The time required for achieving a stabilized temperature shall be assessed and documented. This time depends on the size and design of the cell or battery. If this assessment is not feasible, the exposure time shall be at least [6]hours for small cells and small batteries, and [12] hours for large cells and large batteries. Then the cell or battery at 57oC ±4 oC shall be subjected to a one short circuit condition with a total external resistance of less than 0.1 ohm at 55 ±2 oC. This short circuit condition is continued for at least one hour after the cell or battery external case temperature has returned to 57oC ±4 oC, or in the case of the large batteries, has decreased of half of the maximum increase observed during the test, provided the temperature during the cooling phase has been steadily decreasing [not re-increased by more than 5°C].”.

13. Data about thermal behavior in T5 test.

14. Data about thermal behavior in T5 test.

15. Data about thermal behavior in T5 test. Duration: Case A= 11h Case B= 9 h