Korea opinion for “Fire resistance test” Sep. 15, 2015 Korea Transportation Safety Authority (TS) Korea Automobile Testing & Research Institute (KATRI) Seulki Lee Senior Researcher UN ECE / WP29 / GRSP/ 9 th EVS Informal Meeting in Changchun China (5 th TFG-7 F2F Meeting) EVS TF-04-XXe

Contents 1.Open issues of TF-7 2.Review of TOR 3.Review of current research results 4.Review of test purpose 5.Review of other regulations 6.Korea opinion for ‘Fire resistance test’

Open issues( EVS-08-22e) 1. Open issues of TF-7 Part 1. Short Duration Fire Resistance Test 1.Gasoline pool test vs. LPG burner test – equivalency etc. 2.Necessity to define a single or multiple test methods 3.Necessity to provide temperature profile regarding to starting condition 4.How to define the temperature sensor location 5.Need to measure heat flux of LPG burner test 6.Necessity to define a specification of test equipment 7.How to justify performing the vehicle test on a mock up structure with the combustible materials removed. ☞ # 3, 4, 5, 6, 7 : Should be considered after making a decision for # 1, 2 above Part 2. Long Duration Fire Resistance Test 1.Define the proper test method for long duration test 2.Basic issues of the fire resistance test such as purpose, definition of long term, exposure time, etc. 3.Detailed working plan and timeline by US and Canada 1/22

Main issue 1. Open issues of TF-7 The task of TF-7 is to study and discuss on fire resistance test procedure of REESS which has been proposed by Korea as an alternative. Korea proposal is based on component level. But, possibility of vehicle based test was considered by request. So we have 4 types of test method for fire resistance test optionally. 4 test methods are a burden to manufacturer. Depending on the purpose of GTR, it needs to define a single method or reduce the number of test methods. 2/22 Test methodGasoline pool fire testLPG burner fire testRemark Component based test Thermal Energy : High Flame control : Impossible Flame temperature : Uneven Occurred waste water, Gasoline smoke : Plenty Thermal Energy : Lower than Gasoline Flame control : Possible Flame temperature : Even No waste water Combustion smoke : Few High severity compared to vehicle level Vehicle based test Temp. of DUT : Lower than LPG Temp. of DUT : Low Need to adjust the condition of temp. by shape of vehicle body Removal combustible materials in vehicle Remark - It is possible to make thermal energy equivalency with gasoline test if exposure time is extended.

The purpose of 1998 Agreement 2. Review of TOR CONCERNING THE ESTABLISHING OF GLOBAL TECHNICAL REGULATIONS FOR WHEELED VEHICLES, EQUIPMENT AND PARTS WHICH CAN BE FITTED AND/OR BE USED ON WHEELED VEHICLES 1.1. The purpose of this Agreement is: To establish a global process by which Contracting Parties from all regions of the world can jointly develop global technical regulations regarding the safety, environmental protection, energy efficiency, and anti-theft performance of wheeled vehicles, equipment and parts which can be fitted and/or be used on wheeled vehicles; To ensure that, in developing global technical regulations, due and objective consideration is given to the existing technical regulations of Contracting Parties, and to the UN/ECE Regulations; To ensure that objective consideration is given to the analysis of best available technology, relative benefits and cost effectiveness as appropriate in developing global technical regulations; To ensure that the procedures used in developing global technical regulations are transparent; To achieve high levels of safety, environmental protection, energy efficiency, and anti-theft performance within the global community, and to ensure that actions under this Agreement do not promote, or result in, a lowering of these levels within the jurisdiction of Contracting Parties, including the subnational level; To reduce technical barriers to international trade through harmonizing existing technical regulations of Contracting Parties, and UN/ECE Regulations, and developing new global technical regulations governing safety, environmental protection, energy efficiency and anti-theft performance of wheeled vehicles, equipment and parts which can be fitted and/or be used on wheeled vehicles, consistent with the achievement of high levels of safety and environment protection and the other above-stated purposes; and To ensure that, where alternative levels of stringency are needed to facilitate the regulatory activities of certain countries, in particular developing countries, such needs are taken into consideration in developing and establishing global technical regulations This Agreement is to operate in parallel with the 1958 Agreement, without affecting the institutional autonomy of either Agreement. 3/22

TOR of EVS IWG 2. Review of TOR Terms of Reference for the informal group on Electric Vehicle Safety (EVS) 2. OBJECTIVE OF THE PROPOSAL The GTR will address the unique safety risks posed by EVs and their components. It will be performance-based to the extent possible so as not to restrict future technology development. It will be preceded by an exchange of information on current and future planned domestic regulatory safety requirements for electric vehicles based on section C of the formal proposal, (ECE/TRANS/WP.29/2012/36 and its Corr1) including the underlying scientific and technical basis and research. The GTR will cover high voltage electrical safety, electrical components such as electric connectors and inlets, and REESS-- in particular those containing flammable electrolyte. The provisions will address the safety of electric vehicles, both in-use and post-crash. The key items would be as follows: a.In-use--normal operation of the vehicle excluding maintenance and repair: c) Safety requirements for REESS risks, including thermal shock, thermal cycling, mechanical shock, over-discharge, isolation resistance, over-charge, vibration, fire resistance and short circuit, etc. b. During and post-crash: To the extent possible, the experts of the subgroup will develop the GTR using the following processes: ∙ Identify potential safety risks specific to EVs ∙ Develop and evaluate the requirements by reviewing analyses and evaluations conducted to support the requirements;  Develop and validate test procedures using existing evaluations and research; and ∙ Avoid design-restrictive requirements and provisions that are not technically supported 4/22

Difference of Certification System Type Approval  Type approval is the confirmation that production samples of a design will meet specified performance standards. The specification of the product is recorded and only that specification is approved. Self-certification  Vehicle manufacturers, assemblers and importers need to assure (self-certification) that their vehicles are compliant with the rules on safety standards by themselves.  Government do defects investigation after vehicle sales. Type ApprovalSelf-certification CountryMost of countries (EU, China, Japan, Australia etc.)Korea, Canada, USA Post ManagementCOP(Conformity of Product) Compliance test Recall 2. Review of TOR 5/22 Comment  Countries have a different stance which certification system is adopted.  Reproducibility is one of the important factor for compliance test in self-certification system.  There should be the same result whenever the test is performed.

6/22 Comparison flame temperatures LPG burning with small mockupGasoline free burning (small pool)Gasoline burning with small mockup LPG burning with large mockupGasoline free burning (large pool)Gasoline burning with large mockup Gasoline pool fire test shows a large deviation of flame temperature at the same height. Otherwise LPG burner fire test shows a small deviation of flame temperature at the bottom of DUT stably. 3. Review of current research results Compare LPG burner with Gasoline pool fire test

Canada Research Result 25L Gasoline Propane Heat Release Rate Flame Temperature Flame temperature in Canada test result using propane sand burner shows similar trend with Korean result. Heat release rate of propane sand burner was maintained about 2,000kW. Heat release rate of gasoline pool is higher than propane sand burner. But It does not entirely affect to DUT. 7/22 3. Review of current research results

8/22 LPGDirect exposure, 120[s]Pre-heating30[s] Direct exposure expand 30[s] SUM Parameter Heat Flux [kW/m 2 ] Thermal Energy [kJ/m 2 ] Thermal Energy [kJ/m 2 ] Total Energy [kJ/m 2 ] Total Energy [kJ/m 2 ] Small Mockup27532,9924,1248,248 37,116 (45,364) Large Mockup11814,1581,7703,540 15,928 (19,468) Total Energy per unit area by CFD simulation GasolinePhase B, 70[s]Phase C, 60[s]SUM Parameter Heat Flux [kW/m 2 ] Thermal Energy [kJ/m 2 ] Heat Flux [kW/m 2 ] Thermal Energy [kJ/m 2 ] Total Energy [kJ/m 2 ] Small Pool43330, ,33353,667 Large Pool16011, ,35018,550 Comment  The thermal energy per unit area of gasoline pool fire test is 44% higher than LPG burner fire test for the small, and 16% higher than LPG burner fire test for the large.  To make a thermal energy per unit area equivalency between LPG burner and gasoline pool, direct exposure time of LPG burner test should be extended about 30 seconds additionally. 800 ℃ Reaching Time Comparison thermal energy 3. Review of current research results

9/22 Test Result Comparison component based test (PRIUS REESS) Temp. distribution - Avg. Temp. (150 sec from ignition) ㆍInside : 287 ℃ ㆍBottom : 800 ℃ - Avg. Temp. (40 min from ignition) ㆍInside : 159 ℃ - Max. Temp. : 887 ℃ Temp. distribution - Avg. Temp. (150 sec from ignition) ㆍInside : 408 ℃ ㆍBottom : 755 ℃ - Avg. Temp.(40 min from ignition) ㆍInside : 154 ℃ - Max. Temp. : 872 ℃  Gasoline’s inside temperature is ℃ higher than LPG during fire exposure time.  After remove the fire resource, both DUTs burned themselves showing different temperature trend until they burn out eventually. Comment 3. Review of current research results

10/22 Test Result Inside appearance after test Gasoline pool fire testLPG burner Test Comment  The DUT of Gasoline pool fire test damaged more than LPG burner test comparatively due to higher thermal energy  The point where arcing happened is damaged more severely  Both DUTs occurred arching within exposure time and after remove the fire source, they burned themselves. The damage of DUTs are different due to difference of thermal energy, but it doesn’t seem to really affect whether they explode or not. 3. Review of current research results

11/22 Test Result Internal temperature comparison test (PRIUS REESS Case) MeasurementLPGGasoline Avg. Temp.150 sec130 sec - REESS Inside557 ℃678 ℃ - REESS Bottom838 ℃690 ℃ Max. Temp.939 ℃887 ℃ LPG burner fire Gasoline pool fire MeasurementLPGGasoline Avg. Temp.150 sec130 sec - REESS Inside109 ℃37 ℃ - REESS Bottom122 ℃38 ℃ - Body inside141 ℃35 ℃ - Body Bottom(High)693 ℃570 ℃ - Body Bottom(Low)872 ℃537 ℃ Max. Temp.976 ℃887 ℃ Vehicle Compare Component test vs Vehicle test Component 3. Review of current research results

12/22 Test Result Comment  At the component based test, when we compared the LPG and Gasoline, the temperature of Gasoline is 100~200 ℃ higher than LPG difference of thermal energy.  It is possible to perform vehicle based test with LPG burner, but some area at the bottom couldn’t make the same condition as component level since the shape of underneath vehicle body. If we apply LPG burner to the vehicle based test, we have to consider to change the parameter of temperature lower than 800 ℃  Vehicle based test may be more realistic than component based test for simulating exposure to fire from outside of the vehicle due to a fuel spill from a vehicle which is written in purpose of current GTR draft. But vehicle based test could not represent all real fire situation perfectly. Because the test procedure requires to remove a combustible material in a vehicle and even apply the flame just to the size of the REESS not the whole size of vehicle.  According to the test result, there is a significant difference in temperature between component and vehicle based test. If performing vehicle based test with REESS, it may not be damaged as low temperature. 3. Review of current research results

Canada Research Result No vehicle showed a catastrophic explosion, although numerous pops, flares and fire jets were expelled during the test indicative of a reactive fire, which includes tires, shocks and air bags. Transport Canada proposal: No test at the vehicle level If a fire testing is to be performed at the component level; SOC should be at highest level that the vehicle can charge the REESS 2 minutes is not sufficient to initiate reactions within the REESS Suggestion is to follow/perform a test similar to the UL 2580(min 590°C; 20 minutes; pass criteria: no explosion) or a test similar to the one specified in GTR 13 to assure that the REESS case can contain cell venting and explosion without affecting its structure. 13/22 3. Review of current research results

4. Review of test purpose 14/22 Purpose of Fire resistance test The purpose of this test is to verify the resistance of the REESS, against exposure to fire from outside of the vehicle due to e.g. a fuel spill from a vehicle (either the vehicle itself or a nearby vehicle). This situation should leave the driver and passengers with enough time to evacuate. Review of the purpose The situation described in current purpose represent too limited scope. Need to consider the purpose in terms of causes and status of real fire accident on vehicle. Analyzed the regular pattern of real vehicle fire based on a data from National Fire Data System, Ministry of Public Safety and Security, Korea. A limitation of the data ∙ The cases include fire reported from the fire station and insurance company, there might be no missing and overlap. ∙ xEV account for 0.75%(150,000 cars) among all vehicles(20 million cars) currently in Korea. The data doesn’t show up xEV separately. ∙ The scale of fire and an event about a diffusion from other vehicles are not stated in the data.

15/22 Annual Status of the entire fire The fire in Korea occurs annually 45,000 cases on average, building(residential) is 25%, building(non- residential) is 36%, the vehicle is about 5,700 cases and 13%, hazardous materials and the gasworks 0.1%, railways, aircraft and ships 0.3%, forest land 7%, and others 19%. Personal injury due to fire is about 2,000 people annually on average, the injury due to vehicle fire accounts for approximately 160 people and 7%, that is relatively lower than percentage of fire cases occurred due to vehicle. Building(resident) Building(non-resident) Vehicle Hazardous materials, Railway, ship, Forest land Others Gasworks etc. Aircraft etc. 4. Review of test purpose

16/22 Analysis on causes of vehicle fire in 2014 In 2014, there was 4,462 cases of fire due to vehicle, result in 24 people dead, 94 people injured, and $ 18 million of property loss. Causes by vehicle defects such as electric and mechanical factor etc. are about 2,500 cases account for 57%, causes by accident are about 500 cases account for 11%, and causes by arson and suspected arson are 230 cases account for 5%. 4. Review of test purpose

17/22 In case of vehicle fire, the ignition part of automobile is mainly electric wiring(25%), engine(24%), brake(8%), battery(6%). The causes of ignition are mainly from vehicle defect(78%) such as electric, mechanical, chemical, gas leak, and then accident (11%), inattention(3%), others(17%), arson(1%), unknown(6%) 4. Review of test purpose

18/22 Source : National Fire Data System, Ministry of Public Safety and Security The fire accident due to fuel leaking from fuel tank and line is 15 cases and can be accounted for about only 0.7% among all fire accident(2,045 cases). ☞ The fire potential occurred by a diffusion of nearby vehicle’s fuel leakage gets even fewer. 4. Review of test purpose

19/22 Comment  The purpose of this test is to verify the resistance of the REESS, against exposure to fire from outside of the vehicle due to e.g. a fuel spill from a vehicle (either the vehicle itself or a nearby vehicle). This situation should leave the driver and passengers with enough time to evacuate.  The purpose of this test is to verify the resistance of the REESS in case of the vehicle fire. This situation should leave the driver and passengers with enough time to evacuate. In case of Korea, fire caused by vehicle happened about 5,700 times yearly which account for 13% among all fire. As xEV market is getting bigger, verification of the safety for REESS fire resistance is essential. In vehicle fires, main causes were vehicle defects(57%) and secondly, car accidents(11%). These figures is expected to be similar in terms of xEV. Some of parts caused fire would be changed in xEV.(e.g Engine ☞ Electric motor, inverter etc.) A fire caused by gasoline spill from a nearby vehicle rarely happen. So, it is necessary to redefine the appropriate purpose of fire resistance test and make a test procedure corresponds to a new purpose. ☞ Need to analysis on status and causes of vehicle fire in other countries additionally. 4. Review of test purpose

 GTR 13. (Hydrogen and fuel cell vehicles)  UN R110 (CNG vehicles) LPG is a suitable resource to get high response and controllability. Any fuel can be used but, should consider air pollution cencerns and reproducibility. 5. Review of other regulations 20/22 ( 600℃ +800 ℃ ) ( 590℃ )

 FMVSS 304 (CNG Fuel Container Integrity) TP Dec. 8, 2003 Diesel fuel or any other fuel can be used as long as that can maintain required temperature. ( 430℃ )  In other regulations, there are examples that shows LPG, diesel, or unspecified fuel is used in fire test for gas container.  Most of regulations specify the terms of temperature.  The temperature is more confident controllable factor to reproduce same condition regardless of fuel. Comment 21/22 5. Review of other regulations

6. Korea opinion for ‘Fire resistance test’ 22/22 Reproducibility is one of the important factor for compliance test in self-certification system. It is possible to make thermal energy equivalency of gasoline and LPG burner if direct exposure time of LPG burner fire test is extended 30 seconds more. According to the test result, there is a significant difference in temperature condition between component and vehicle based test. Vehicle based test could not represent all real fire situation perfectly because of a removal of combustible materials and a small size of fire flame. According to the analysis of real vehicle fire based on a data from National Fire Data System in Korea, a fire caused by gasoline spill from a nearby vehicle rarely happen. So it is necessary to redefine the appropriate purpose of fire resistance test and make a test procedure corresponds to a changed purpose of fire resistance test. In other regulations, there are examples that shows LPG, diesel, or unspecified fuel is used in fire test for gas container. Most of regulations specify the terms of temperature.  The purpose of this test is to verify the resistance of the REESS in case of the vehicle fire. This situation should leave the driver and passengers with enough time to evacuate.

Thank you!