1 s2.0 S1364032117308250 Main

Renewable and Sustainable Energy Reviews 81 (2018) 1427–1452
Contents lists available at ScienceDirect
Renewable and Sustainable Energy Reviews

journal homepage: www.elsevier.com/locate/rser
A review of international abuse testing standards and regulations MARK

for lithium ion batteries in electric and hybrid electric vehicles
⁎
V. Ruiza, A. Pfranga, A. Kristona, N. Omarb, P. Van den Bosscheb, L. Boon-Bretta,
a
European Commission, Joint Research Centre (JRC), Directorate for Energy, Transport & Climate, Energy Storage Unit, Westerduinweg 3, NL-1755 LE
Petten, The Netherlands
b
Vrije Universiteit, MOBI, Pleinlaan 2, 1050 Brussels, Belgium
A R T I C L E I N F O A BS T RAC T
Keywords: Lithium ion batteries are a proven technology for automotive applications and their continued use in the future
Battery safety electric vehicle fleet is undeniable. In addition to battery performance and durability, battery safety is
Standards paramount to ensure confidence and widespread adoption of electromobility in our society. This comprehensive
Lithium ion batteries review aims at presenting the various international standards and regulations for safety testing of lithium ion
Abuse testing
batteries in automotive applications under various abusive environments. Safety tests are presented and
Mechanical testing
analysed including mechanical, electrical, environmental and hazards of chemical nature. The intention of this
Electromobility
review is compiling the most relevant standards and regulations to identify shortcomings and areas for future
improvement.
1. Introduction notebooks) [5]. Moreover, LIBs are used to power several electric
vehicles available on the market, e.g. BMWi3, Tesla Model S, Nissan
Reducing carbon dioxide emissions is a major driving force for the Leaf, Mitsubishi iMiEV, Chevrolet Volt, Renault Zoe. The widespread
displacement of traditional internal combustion engines (ICE) based deployment of this technology is reinforced by its relatively high
on fossil fuels by "greener" and more efficient alternatives. In this specific energy and power density and its progressive cost reduction,
context, various measures within the policy framework are being with estimations from ∼ 800 $ kW h−1 per pack in 2010 down to
established internationally to accelerate the development and adoption ∼ 248 $ kW h−1 by 2030 (for a 21 kW h BEV) [6] based on the current
of vehicles based on alternative fuels. Based on these efforts, it is chemistries. Predictions assume that by 2020, LIBs will be used in
expected that the electrification of transport will make up a significant 65% of the total EV systems, surpassing other technologies, including
share of the near future automotive fleet [1]. According to the Report NiMH [7].
"Competitiveness of the EU Automotive Industry in Electric Vehicles" Many battery standards and regulations have been specifically
published in 2012 [2] the European Union (EU-27) will reach 14.8 developed to facilitate and regulate battery use in EVs. At this stage
million new light duty vehicle registrations (passenger cars and light it is useful to differentiate between standards and regulations.
commercial vehicles) by 2020, of which 7% will be electric vehicles Standards are in principle voluntary documents, drafted by non-
(including Battery Electric Vehicles, BEVs, plug-in Hybrid Vehicles, governmental organisations such as the International
PHEVs and Fuel Cell Vehicles, FCV). This market share is foreseen to Electrotechnical Commission (IEC), the International Organisation
rise to 31% by 2030 with Europe, Japan and U.S. expected to be for Standardisation (ISO), the Society of Automotive Engineers
leading markets. Other studies considering moderate policy support International (SAE) at international level and the European
and technical advancement present 5–10% of the market share in the Committee for Standardisation (CEN) and European Committee for
2025–2030 time frame [3]. Global registrations of FCVs will still be Electrotechnical Standardisation (CENELEC) at European level.
under 1% in 2030, thus battery driven vehicles will dominate the EV Standards can also be issued by National bodies (e.g. British
market in the near future. Standards Institution (BSI), Japanese Industrial Standards
In 1991 Sony launched the first commercial lithium ion batteries Committee (JISC)) or regional organisations. Regulations, on the other
(LIBs) [4]. Since then it has emerged as the dominant energy storage hand, are issued by governmental authorities and have the force of law.
technology used in most consumer electronics (e.g. cell phones, For road vehicles, the most relevant regulations are type approval
⁎
Corresponding author.
E-mail address: [email protected] (L. Boon-Brett).
http://dx.doi.org/10.1016/j.rser.2017.05.195
Received 15 June 2016; Received in revised form 10 April 2017; Accepted 21 May 2017
Available online 14 July 2017
1364-0321/ © 2017 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/BY/4.0/).
V. Ruiz et al. Renewable and Sustainable Energy Reviews 81 (2018) 1427–1452
regulations issued by the United Nations Economic Commission for Table 1

Europe (UNECE). These regulations define uniform technical prescrip- List of typical components in lithium ion batteries (LIBs).
tions for wheeled vehicles, their parts and equipments, and state
Cathode LMO, LCO, NCA, NMC, LFP, ECPs
conditions for reciprocal recognition of type approvals by several Anode Graphitic carbons, Hard carbons, Synthetic graphite,
countries. In the USA, the National Highway Traffic Safety LTO, Tin-based alloys, Silicon-based alloys
Administration (NHTSA) issues regulations via the Federal Motor Electrolyte salts LiPF6, LiClO4, LiAsF6, LiCF3SO3, LiBF4
Vehicle Safety Standards (FMVSS), setting minimum safety perfor- Electrolyte solvents DMC, EC, DEC, PC, ɣ-GBL, RTIL's
Flame retardants HMPN, TMP, TFP
mance requirements for motor vehicles or items of motor vehicle Gel precursor PEO, PAN, PVDF, PMMA, PTFE
equipment. Binder PVDF, SBR, Glass Fibre, CMC, ACM
Standards may be referred to by laws and regulations. For several Separator Polypropylene, Polyethylene, Cellulosic paper,
technical domains, Europe now follows the so-called New Legislative Nonwoven fabrics, Ceramic
Framework (NLF) adopted in July 2008, which built on the "New
LMO: Lithium Manganese Oxide, LCO: Lithium Cobalt Oxide, NCA: Lithium Nickel
Approach", where directives only mention essential requirements, Cobalt Aluminium Oxide, NMC: Lithium Nickel Manganese Cobalt Oxide, LFP: Lithium
while technical details are specified in harmonized European standards Iron Phosphate, ECPs: Electronic Conducting Polymers, LTO: Lithium Titanate, PVDF:
referred to in these directives. Conformance to these standards Polyvinylidene Fluoride, SBR: Styrene Butadiene Rubber, CMC: Carboxymethyl
subsequently implies conformity to the essential requirements of the Cellulose, ACM: acrylate-type copolymer, RTIL's: Room Temperature Ionic Liquids,
directive. The NLF is used for example in the Low Voltage Directive, DMC: Dimethyl Carbonate, EC: Ethylene Carbonate, DEC: Diethyl Carbonate, PC:
Propylene Carbonate, ɣ-GBL: gamma-Butyrolactone, HMPN: hexamethoxycyclotripho-
but not yet for road vehicles, for which UNECE type approval
sphazene, TMP: trimethyl phosphate, TFP: tris(2,2,2-trifluoroethyl)phosphate, PEO:
regulations are used. The Motor Vehicle Type Approval (EC Polyethylene Oxide, PAN: Polyacrylonitrile, PMMA: Poly Methyl Methacrylate, PTFE:
Directives) allows national type approvals and subsequently ensures Polytetrafluoroethylene.
recognition of this approval in other EU member states, i.e. if one
vehicle is type approved in one member state, it is allowed to be sold in
any other member state. electric vehicles (Section 5). Shortcomings and suggestions for future
In 2012, EU and US standards organisations agreed a Transatlantic development are also identified (Section 5).
Cooperation on Standards for Electric Vehicles to avoid proliferation of
conflicting electric vehicle and battery safety standards. The coopera- 2. Safety issues and challenges related to lithium ion
tion sets the basis towards harmonisation and alignment of standards batteries
in the field of electromobility [8]. The need for such harmonisation of
battery standards for automotive applications has been acknowledged 2.1. Battery materials and components
by others who suggest that performance and safety can hence be
improved [9]. LIBs are rechargeable energy storage devices where Li ions move
Battery safety standards and regulations call for testing in abusive between the anode and cathode, which are electrically separated by a
conditions. In these situations (e.g. overcharging, short circuit, physical membrane. All components are fully soaked in an electrolyte. During
deformation in a vehicle crash) exothermic reactions may be triggered charging, lithium ions move from the cathode towards the anode and in
(e.g. temperature rise of hundreds of degrees within seconds [10]) the discharge cycle the ions travel back. The electrons move via the
leading to thermal runaway. This can lead to the heating up of external electrical circuit and lithium ions and solvent molecules travel
neighbouring cells within a module, which – if sufficient heat is within the electrolyte. When the battery is charged, the Solid
generated – can lead to a chain reaction and propagation [11,12], Electrolyte Interface (SEI) is formed. This passivation layer, Li+
and in a worst case scenario, develop into fire and explosion [13–15]. conducting and electronically insulating, is paramount for optimum
Most of the time LIBs behave as foreseen during their lifetime. battery performance as it allows Li intercalation and prevents further
However a number of highly publicized LIB safety events have led to electrolyte decomposition [26,27].
hazardous situations making the evaluation of battery safety a key As the risks associated with a certain battery technology depend
aspect in battery development. Events such as laptop fires [16], highly on the cell constituents, it is important to consider all relevant
smoking cell phones [17], airplane incidents [18–21], the GM Volt components from a safety perspective. Table 1 summarizes typical
fires [22], ground impacts leading to safety events on Tesla Model S components found in LIBs. A relatively high number of materials have
[23], although scarce, reach the media much easier than events with been used in cathodes, including lithium manganese oxide (LMO),
established technologies (i.e. internal combustion engine vehicle fires). lithium cobalt oxide (LCO), lithium nickel cobalt aluminium oxide
Such events have led to withdrawal of products from the market (e.g. (NCA), lithium nickel manganese cobalt oxide (NMC) or olivine type
Apple removed lithium ion power packs from their PowerBook 5300 materials, such as lithium iron phosphate (LFP). The latter has
line [24], CPSC and EV Global Motors Company announced the recall appeared as one of the safest chemistries due to its thermal stability
of 2000 batteries in their electric bicycles [25]) which may generate an and non-toxicity [28,29]. On the other hand the energy density in LFP
increased concern from the general public towards lithium ion batteries is lower compared to LCO alternatives, which have less
technologies in general. The link between safety related events and desirable behaviour when a thermal event occurs [30,31].
the market uptake of battery driven EVs is of concern to battery Regarding the anode, carbon is commonly used in LIBs. It can
producers, vehicle original equipment manufacturers (OEMs) and reversibly accommodate significant amounts of lithium providing a
transportation policy makers. theoretical capacity of 372 mA h g−1 (LiC6). More recently, lithium
The objective of this review is to compile the most relevant titanate (LTO) has attracted considerable attention due to its long cycle
standards and regulations dealing with abuse testing of LIBs. Safety life without significant structural changes upon cycling [32,33] and its
risks specific to LIBs are summarized (Section 2). Test methods in increased safety in terms of thermal stability and high potential which
these standards and regulations are classified according to the nature of prevents dendrite formation [30] at the cost of a comparatively lower
the misuse conditions applicable (Section 3). Test parameters and voltage [30].
conditions used in test methods are compared – commonalities and Electrolytes used in LIBs are mainly based on aprotic organic
differences are highlighted (Section 3). Relevant forthcoming standards solvents, often highly flammable [34]. The most commonly used
and regulation are listed (Section 4). Stemming from these compar- electrolytes are mixtures of various carbonates (e.g. propylene carbo-
isons, conclusions are drawn identifying areas for improvement with nate) and a dissolved salt (e.g. lithium hexafluorophosphate (LiPF6)).
respect to the relevance and fitness for purpose of existing tests for In the event of thermal runaway, the electrolyte decomposes leading to
1428
the formation of gases. Consequently, significant overpressure is installation considerations [54]. An additional parameter influencing
generated in the cell, which will eventually lead to venting and/or battery safety is cell design [16]). Vehicle manufacturers utilize
rupture. A major hazard is the presence of fluorinated compounds in prismatic (e.g. VW, Audi, Porsche, Citroen, Peugeot, Fiat), pouch
the electrolyte, leading to the release of toxic and corrosive hydrogen (e.g. Mini, Mercedes, Renault) or cylindrical cells (e.g. Tesla).
fluoride (HF). Since some gases generated in such events are toxic Cylindrical cells are cheap to manufacture, have good mechanical
[35,36] and may potentially cause severe harm to individuals in the stability and high energy density. However, they have low packing
surroundings, immediate medical attention is required after exposure efficiency [55]. They do not swell during operation, but when pressure
to vented gases [16]. In order to reduce the flammability of these builds up expulsion of the jelly roll (layers of anode/separator/cathode
electrolytes various flame retardant additives have been explored giving rolled up and inserted into a hollow cylinder casing) can occur [56].
rise to the concept of “non-flammable electrolytes” (e.g. Phosphate Prismatic cells are mechanically robust with high packing efficiency,
solvents [37], phosphazene derivatives [38,39], room temperature however, they have slightly lower energy density and are more
ionic liquids [40,41]). Safety performance of LIBs can be improved expensive [55]. In case of pressure build up, the generated gases are
using alternative electrolytes such as more thermally stable, high released via the safety vent. When the opening of the safety vent is too
flashpoint electrolytes [42] or room temperature ionic liquids small, or when it is clogged it can hinder the escape of gas. This
(RTILs) [40,41], which show promise due to their low volatility, with situation can lead to rupture or explosion of the cell [16]. Soft pouch
virtually no vapour pressure (ca. 100 pPa at 298 K for 1-butyl-3- cells have a higher energy density than the other two designs, their
methylimidazolium hexafluorophosphate [43] compared with 3 kPa fabrication cost is not very high, and they are much lighter. However, at
at 298 K for H2O [44]), high flame resistance, thermal and chemical system level this can be reversed due to the stronger mechanical
stability together with a wide window of electrochemical stability [45]. constructions needed for their protection. They are prone to swelling
Electrolytes in the solid/gel form (solid polymer electrolytes), can also during operation (e.g. ageing, exposure to > 60 °C [56]) and have no
be utilized. On one hand their ionic conductivity is much lower than in designated venting mechanism. In case of venting, gases are not
liquid systems, but on the other hand their safety is improved (e.g. directed towards a safety valve, as all the sealing points in the pouch
lower reactivity versus lithium, absence of risk of electrolyte release) cell impose small resistance to high pressure. Consequently, the release
[46,47]. of gases occurs with smaller energy than for the other assemblies. The
The binder is essential for enabling electrode fabrication. Initially, unconstrained nature of the pouch cells may be more effective
most of the anodes were obtained by utilizing polyvinylidene fluoride preventing a thermal runaway reaction compared to cell designs where
(PVDF), however the current trend is to use styrene butadiene rubber electrodes are forced to maintain close contact [57]. Additionally,
(SBR), which yields more flexible electrodes, higher binding ability pouch cells exhibit smaller internal temperature gradients compared to
with a small amount of binder, larger battery capacity and higher prismatic assemblies [55].
cyclability [48]. SBR is unsuitable for the cathodes, which are prone to Another aspect associated to battery safety relates to the fact that
oxidation and consequently PVDF is still used. Electrode preparation cells within a pack exhibit non-uniform properties upon cycling.
with PVDF requires N-methyl pyrrolidone (NMP) for dissolution with a Consequently, there may be some unbalances (e.g. voltage variations
consequent toxicity concern. Water - soluble binders (e.g. carboxy- between cells) that may trigger a safety hazard.
methyl cellulose, CMC) are preferred from an environmental perspec- Battery ageing also needs to be evaluated. Battery cells degrade
tive. More recently, the highly flexible acrylate-type copolymer (ACM) both by undergoing charge-discharge cycles and by time (calendar
has started to be used in some prismatic batteries [48]. ageing). The application and safety of “second life” automotive
The separator is a key element for preventing the electrical contact batteries should be considered. In this application, decommissioned
between electrodes while allowing ion transport [49]. Currently, thin vehicle traction batteries may be used for stationary storage (e.g.
microporous polyolefin membranes made of polyethylene (PE), poly- electric grid support).
propylene (PP), or laminates of both (e.g. PP/PE/PP [50]) are mostly A final relevant aspect is the design of the battery pack. For
used. In a hazardous situation when a temperature above the melting example, standards such as SAE J2289:2008 [58] describe that
point of PE is reached (135 °C), PE will melt, whereas PP (melting material vented from the battery should not be directed into the
point of 165 °C) will maintain its integrity. As the polymer melts, its passenger compartment where it may pose a hazard to passengers.
pores are blocked resulting in an insulating layer, effectively shutting
down the cell and providing a degree of protection against short circuit 3. Relevant standards and regulations: abuse testing of
and overcharge [49]. Alternatively separators based on ceramic mate- lithium ion batteries for automotive applications
rials have also shown high-temperature stability, good chemical
resistance, and wettability [51]. Lithium ion batteries must pass a series of safety tests to be certified
It is clear that many aspects influence the safety of LIBs and the for use in a particular application (e.g. portable electronics or auto-
evaluation of all battery design parameters (e.g. electrode material, motive). Safety tests are described in international, national and
particle size [52], separator) is needed in order to optimise safety. regional standards, typically developed based upon pre-normative
Furthermore, in order to achieve a safe system for a particular research and experience from industry, academia and regulatory
application a compromise in the selection of cell components with bodies. These tests are performed to understand and identify potential
respect to safety, performance and cost is essential. battery weak points and vulnerabilities when the battery experiences
real-life off-normal conditions and to determine how the battery will
2.2. Battery cell and pack design behave under severe abusive conditions, such as a car crash or thermal
shock. In these situations, thermal runaway can develop. Other causes
Industry experts estimate that between one in 10 million [11] or of a thermal runaway can be the presence of microscopic particles from
one in 40 million [34] cells fail during normal operation, if proper manufacturing or impurities, which can pierce the separator creating
quality control is in place. Despite the low probability, the risk is not an internal short circuit. Therefore, a thermal runaway can be initiated
trivial and the consequences cannot be neglected. For this reason, by both external and internal stimuli. The consequences that thermal
efforts to improve the safety of the batteries are taken along the whole runaway produces vary depending on several factors, including: state of
electric vehicle manufacturing chain [31], from safer components (see charge (SOC), charging/discharging rate, cell-type, cell history, cath-
Section 2.1), smarter energy management [53] and battery manage- ode/anode material, electrolyte composition, etc. [59].
ment systems (BMS), and smarter vehicle designs (e.g. installation of Many tests presented in this review are devoted to the evaluation of
battery pack away from crush zones [31] and other safety related the consequences of a short circuit, which might be followed by thermal
1429
V. Ruiz et al.
Table 2
Overview of tests in standards and regulations applicable to lithium ion batteries in automotive applications. Test level is indicated as C: Cell, M: Module, P: Pack and V: Vehicle.
Region of applicability International EU and USA Korea India China

further
countriesa
Test Section SAE SAE ISO 12405- ISO IEC 62660-2 UN/ECE- UL 2580 USABC FreedomCAR KMVSS 18- AIS-048 QC/T 743
J2464 J2929 1 (2) 12405-3 (3) [70,71] R100.02 [62] [63] [72] [65] (2005) 3 [73] [74] [75]d
[61] [66] [67,68]c [69]c (2011 (2013) (2013) (1999) (2009) (2009) (2006)
(2009) (2013) (2012) (2014) (2016))
Mechanical Mechanical 3.1.a CMP CMPV P P C CMPV C Me P MP MP M

shock
Drop 3.1.b P P CP P P P C
Penetration 3.1.c CMP CMP CMP CM CP
Immersion 3.1.d MP P P MP MP MP P
Crush/crash 3.1.e CMP PV PV C CMPV CMP CMP CMP CP
Rollover 3.1.f MP P P MP MP M
Vibration 3.1.g CMP P P C CMP CMP C Me P CMP M P
Electrical External short 3.2.a CMP P P P C CMP CMP CMP CMP P CMP CP
circuit
1430
Internal short 3.2.b C
circuit
Overcharge/ 3.2.c C M Pf P P P C CMPV CMP CMP MP P C M Pg CP
overdischarge
Environmental Thermal stability 3.3.a C C C CMP CMP P CP

Thermal shock 3.3.b CMP CMP P P C CMP CMP CMP CMP
and cycling
Overheat 3.3.c MP P CMPV MP
Extreme cold 3.3.d CMP
temperature
Fire 3.3.e MP P P Vb CMPV CMP CMP CMP P
Chemical Emissions 3.4.a CMP P CMP CMP CMP

Flammability 3.4.b CMP P CMP CMP CMP
a
Norway, Russia, Ukraine, Croatia, Serbia, Belarus, Kazakhstan, Turkey, Azerbaijan, Tunisia, South Africa, Australia, New Zealand, Japan, South Korea, Thailand and Malaysia.
b
Vehicle body may be included,
c
Also possible at battery pack subsystem: representative portion of the battery pack (energy storage device that includes cells or cell assemblies normally connected with cell electronics, voltage class B circuit, and overcurrent shut-off device,
including electrical interconnections and interfaces for external systems,
d
Applicable to the LIB cell and pack whose rated voltage is 3.6 V and nx3.6 V (n: quantity of batteries), respectively.
e
At the module level for those electric energy storage assemblies intended for use in applications larger than passenger vehicles. The module level testing shall be representative of the electric energy storage assembly.
f
Overdischarge not at pack level.
g
Overdischarge not performed.
runaway, as this is one of the scenarios that may create a great risk, charge) in the various standards and regulations, as summarized in
both for the vehicle occupants and first aid responders. In some tests Table 3. To facilitate the comparison between the various parameters
the short circuit is induced externally, such as in the case of crush, please refer to Fig. 1. Standards SAE J2464:2009 and SAE J2929:2013
penetration and drop tests, however other tests aim at inducing the [61,66] follow UN 38.3:2015 transportation regulation [76], and
short circuit internally. The development of tests representative of an require the most stringent conditions of all the standards and regula-
internal short circuit is quite controversial due to the difficulty in tions evaluated, in terms of peak acceleration (150g) for cells of <
emulating a true internal short circuit in a testing environment. For 0.5 kg. For heavier systems the conditions are eased [61,66].
this reason, there is a lack of consensus regarding the "fit for purpose" Interestingly, ISO 12405 part 1:2011 and part 2:2012 [67,68], UL
of internal short circuit tests currently described in existing standards. 2580:2013 [63] and ISO 62660-2:2011 [70] (which follow ISO 16750-
There is little knowledge on how an internal short circuit within a 3:2003 [77]) have the same requirements (500 m s−2 (∼ 51 g) and
battery pack develops. Most of the scientific literature refer to small 6 ms) despite the fact that the test levels are different (P, P, C and C,
batteries or cells [57,60], and analogous data at pack or full vehicle respectively, see Table 3). It is reasonable to assume that the impact
level is scarce due to the high cost of the tests and to the fact that the and outcome of the test is dependent on the DUT size, and that the test
information is in most of the cases proprietary to the testing bodies or conditions should be dimensioned to each level.
the OEM. Under the recently published ISO 12405-3:2014 [69], an optional
Table 2 presents a summary of the most frequently required abuse mechanical shock test is included compared to parts 1 and 2 [67,68],
tests as described in international standards and regulations related to adopting the shock parameters used by UN/ECE-R100.02:2013 [62].
electric vehicles based on lithium ion technologies. Abuse tests are This regulation specifies test parameters for batteries to be installed in
classified according to the nature of the misuse: mechanical, electrical, road vehicles of categories M1-N1, M2-N2 and M3-N31 with varying
environmental and chemical. Tests that appear in only few standards or acceleration profiles depending on orientation and vehicle type. A
regulations will be mentioned but not explained in detail. In some higher shock level and/or longer duration can be applied to the DUT if
circumstances, upon agreement between the manufacturer and the recommended by the manufacturer.
customer, the standard or regulation allows certain flexibility in the test FreedomCAR and USABC standards [65,72] divide the shock test
conditions. The tests can be performed at various system levels: cell into low-level (no damage to the DUT) and mid-level (DUT may be
(C), module (M), pack (P) and vehicle (V). In general we will refer to inoperable after test). While all considered standards and regulations
the device under test (DUT). Definitions for each level follow SAE require a half-sine wave, FreedomCAR and USABC allow also other
J2464:2009 [61] and can be summarized as: pulse shapes which would simulate actual decelerations more accu-
rately. Also deviations from the specified shock parameters may be
• Cell (C): energy storage device composed of at least one cathode and requested by the manufacturer. These two standards [65,72], as well as
one anode, and other necessary electrochemical and structural UN/ECE-R100.02:2013 [62], present significantly higher shock dura-
components. tions (ranging 55–120 ms) compared to the other standards ( <
• Module (M): grouping of interconnected cells in series and/or 20 ms), presumably imposing harder conditions on the DUT.
parallel into a single unit. Mechanical shock testing can also be performed at vehicle level, as
• Pack (P): interconnected modules including all auxiliary subsystems mentioned in UN/ECE-R100.02:2013 [62], SAE J2929:2013 [66] and
for mechanical support, thermal management and electronic con- ISO 12405-3:2014 [69]. For UN/ECE-R100.02:2013 [62], batteries
trol. installed in a vehicle that has already been successfully subjected to
vehicle crash testing in accordance with UN/ECE-R12:2012 – Annex 3
In general, standards and regulations set pass/fail requirements for [78] for protection of the driver against the steering mechanism in the
each test. For example, UN/ECE-R100.02 [62], ISO 12405-3, UL 2580 event of impact, with UN/ECE-R94:2012 – Annex 3 for frontal
[63] set "no fire", "no explosion", "no rupture", and "no leakage" as collision [79] and with UN/ECE-R95:2011 – Annex 4 for lateral
acceptance criteria for tests under reasonable foreseeable misuse (e.g. collision [80] are considered to be compliant. SAE J2929:2013 [66]
vibration, thermal shock, external short circuit), whereas the pass/fail follows requirements described in FMVSS 305:2011 [81] (or equivalent
criterion for fire resistance is "no explosion" only. Specific to auto- regulation depending on the geographical region applicable to vehicle
motive applications, the response of a technology to an abusive front, rear and side crash testing). Similarly ISO 12405-3:2014 [69]
condition can be classified according to the EUCAR hazard levels requires following relevant national or regional regulations on vehicle
[64,65]: from level 0 (no effect, system maintains its functionality) to crash tests.
level 7 (explosion, mechanical disintegration of the system). Battery According to the FP7 project EVERSAFE [82], the majority of real
and car manufacturers often utilize this classification to evaluate the world crashes show acceleration values below 20–30g for frontal and
response of a RESS to an abusive condition. For example, a level 3 or side impacts with durations lower than 100 ms, and accelerations
lower usually represents an acceptable level of performance. significantly lower ( < 12g) in the case of rear impacts. However, when
Direct comparison of the value of each testing parameters should be the aim of the test is to evaluate worst case scenarios, the parameters
performed prudently. Differences in test parameters may be rationa- would need to be more stringent, particularly for standards and
lised by differences in the scope and purpose of the tests. For particular regulations investigating vehicle crash scenarios. For example, full-
tests of interest the reader is advised to consult the reference texts width barrier crash test (56 km h−1) develops shock peaks up to 55g
directly. [83], only ISO 12405 part 1:2011 and part 2:2012 [67,68] approximate
this value at pack level. Based on these examples, comparability of test
3.1. Mechanical tests conditions performed at vehicle level and component level would
require deep evaluation. Another aspect pointed out by the project
3.1.1. Mechanical shock test
The mechanical shock test aims at evaluating the robustness of a
battery in situation of sudden acceleration and/or deceleration of a 1
Vehicles designed for the carriage of passengers: (M1: < 8 seats in addition to the
vehicle. During the test a DUT is exposed to shock forces defined in driver’s seat, M2: > 8 seats in addition to the driver’s seat, mass < 5 tonnes. e.g. small
terms of acceleration and shock duration adapted to different condi- buses and minibuses, M3: > 8 seats in addition to the driver’s seat and a maximum mass
> 5 tonnes. e.g. large buses) and vehicles designed and constructed for the carriage of
tions; from normal in-use driving, driving at high speed over a goods: (N1: having a maximum mass < 3.5 tonnes. e.g. light vans and trucks, N2: having
kerbstone [67], to vehicle crash [62,65,72]. There is a great diversity a mass between 3.5 and 12 tonnes e.g. mid-sized vans and trucks, and N3: exceeding 12
in the test conditions (direction, peak acceleration, duration, state of tonnes. e.g. heavy vans and trucks).
1431
V. Ruiz et al.
Table 3
Test conditions for the mechanical shock test at cell (C), module (M), pack (P) and vehicle (V) level.
Region of International EU and further USA Korea India China

applicability countries#
Shock SAE J2464 [61] SAE J2929 ISO ISO 12405-3 [69]* IEC UN/ECE-R100.02 [62]## UL 2580 [63] USABC Freedom KMVSS AIS-048 QC/T
parameters [66] 12405- 62660- [72]*** CAR [65]*** 18-3 [73] [74] 743
1(2) 2(3) [75]
[67,68]* [70,71]**
Level (C, M, P) CMP CMP P P C CMP C M+ P MP M

### ### ### ###
Direction of Positive and Positive and L and T L and T (C), Positive and Axis that will cause the V and
shock negative negative negative directions. 3 most potential damage H axis
directions. 3 directions. L repeats on 3 axes
repeats on 3 axes and lateral axes (total 18 shocks) (M,
(total 18 shocks) P)
Peak 150 (C), 50 (C > 150 (C M P), 50 51 51 or Vehicles ≤ 3,5 t: 10–28 51 Vehicles ≤ 3,5 t: 10–28 51 (C) 30
acceleration 0.5 kg M,P > (C > 0.5 kg (L)//4.5–15 (T), Medium- (L)//4.5–15 (T), Medium-
(g) 12 kg) or 25 (P) M,P > 12 kg) or duty trucks and midi buses: duty trucks and midi 25 (M, P) 30 (mid-1)
25 (P) 5–17 (L)//2.5–10 (T), buses: 5–17 (L)//2.5–10 20 (mid-2)
Heavy-duty trucks and (T), Heavy-duty trucks and 20 (low)
1432
buses: 4–12 (L)//2.5–10 (T) buses: 4–12 (L)//2.5–10
(T)
Shock duration 6 (C) 11 (C > 6 (C M P) 11 (C 6 6 or 80–120 6 80–120 6 (C) 15 (M, P) ≤ 55 (low) 15
(ms) 0.5 kg M,P > > 0.5 kg M,P > ≤ 65 (mid-1)
12 kg) or 15 (P) 12 kg) or 15 (P) ≤ 110 (mid-2)
SOC (% rated 95-100 95–100% max. 50 50 or > 50% normal 80 (HEV), > 50% normal operating Max. operating SOC 100 100
capacity) normal vehicle operating range 100 (BEV) range (M, P),
operation 80 (HEV), 100
(BEV) (C)
Vehicle level FMVSS 305 Relevant national or regional UN/ECE-R95 [80]: 48.3-
(V)/collision [81]: 48, 54, regulation on vehicle crash 53 km h-1, UN/ECE-R94
speed 80 km h-1 tests [79]: 56 km h-1, UN/ECE-
R12 [78]: 50 km h-1
Longitudinal (L), transversal (T), vertical (V), horizontal (H), * or according to a test profile determined by the customer and verified to the vehicle application. ** If more severe test parameters are requested by any regulation, such test conditions
may be applied. *** It is in the interest of DUT manufacturers to keep the pulse duration as long as possible and still meet the specification. However, if the electrochemical storage system (ECSS) is robust, tests may exceed the peak acceleration,
reduce the duration, reduce the test complexity, and hence, reduce the test cost. #Norway, Russia, Ukraine, Croatia, Serbia, Belarus, Kazakhstan, Turkey, Azerbaijan, Tunisia, South Africa, Australia, New Zealand, Japan, South Korea, Thailand and
Malaysia. ## A higher shock level and/or longer duration can be applied to the tested-device if recommended by the manufacturer. ### Same direction as the acceleration of the shock that occurs in the vehicle. If the direction of the effect is not
known, the cell shall be tested in all six spatial directions. + For those electric energy storage assemblies intended for use in applications larger than passenger vehicles. The module level testing shall be representative of the electric energy storage
assembly.
If only one drop test is performed, it shall not be a flat drop. If the electric energy storage assembly is intended to be installed or removed in a horizontal direction, a drop with the DUT slanted at a 10° angle with pack edge impacted, should be
QC/T 743 [75]
Hardwood floor
20 mm thick
China
100
1.5
C
India
AIS-
048
[74]
range of a vehicle
Max. operating
KMVSS 18-3
Concrete floor
or 80% SOC
Korea
[73]
4.9
P
FreedomCAR [65]
Cylindrical steel object (radius

CMP
≤ 10
150 mm)
Fig. 1. Comparison of peak acceleration and shock duration for various standards and
100
regulations.
USA
EVERSAFE [82] is that the shock parameters defined in the standards
USABC
[72]
and regulations are extracted from conventional car testing, however
10
P
the accelerations experienced by the battery in the vehicle installation
might be different for EVs. Investigations in this respect would be
highly desirable to evaluate whether specific requirements for electric
Flat concrete
UL 2580
operating
surfacea
vehicles need to be imposed.
Max.
SOC
[63]
CP
1
3.1.2. Drop test
This test simulates a situation when a battery is being removed
EU and further
R100.02 [62]
from or installed in an electric vehicle and accidentally drops. Table 4
UN/ECE-
countries#
shows the requirements for surface type, drop height and SOC. During
the test the DUT is let fall onto a rigid flat surface (e.g. concrete floor
[63,73]) or onto a cylindrical object made of steel [65,72]. The shape of
this cylindrical object is supposed to represent a telephone pole or a
IEC 62660-
similar object. Alternatively, it is also possible to perform a horizontal

[70,71]
2 (3)
impact of equivalent velocity as described in SAE J2464:2009 [61] or

USABC:1999 [72]. The fall height varies considerably in the various
standards (from 1 m [63] up to 10 m [65,72]). Consequently, the
outcome of the test can be expected to vary.
ISO 12405-
1 (2)(3)
[67–69]
Systems intended to be removed from the vehicle for charging (or

replacement/swapping) are required to perform this test in UL
2580:2013 [63]. In this case, the test has to be repeated three times
(on the same item) as the likelihood of dropping the battery is higher
according to documented procedures or ≥ 1
maximum distance to ground or maximum
95–100% max. normal vehicle operation

possible drop distance which the battery
than if the battery does not need to be removed from the vehicle. This
system experiences when serviced
International
Test conditions for the drop test at cell (C), module (M) and pack (P) level.
scenario seems very plausible, however, the drop test is not included in
various automotive battery regulations and standards, such as UN/
ECE-R100.02:2013 [62], ISO 12405-1: 2011, ISO 12405-2:2012, ISO
SAE J2929 [66]
Flat surface
12405-3:2014 [67–69] and ISO 62660-2:2011 [70].

P
3.1.3. Penetration test

In this test, both mechanical and electrical damage is induced in the
battery. A sharp steel rod – the 'nail' – is forced through the battery at a
certain constant speed, generally 8 cm s−1 [65,72,74]. Although the
consequence of the test is a short circuit, this short circuit is
mechanically induced. For this reason, the penetration test is usually
classified within the mechanical tests and not within electrical tests. As
95–100
J2464
SAE
[61]
the nail penetrates through the cells and the integrity of the separator
P
and electrodes is compromised, short circuits are created and conse-

quently heat is released. Multiple electrode layers are in electrical
Drop height (m)
contact, together with the shorting occurring on the nail, so relatively

Level (C, M, P)
SOC (% rated
important damage occurs in a short period of time. Additionally, due to

Surface type
applicability
capacity)
parameters
Region of
the fact that the deformation is localized in a relatively small area, the
considered.
heat dissipation is quite limited.

Table 4
Drop
Depending on the test level (cell, module or pack), the depth of

a
penetration and the dimension of the nail vary as described in many of

#
1433
the standards (see Table 5). In most cases, a 3 mm diameter rod is
Not specified (C) 3

QC/T 743 [75]
required at cell level and a 20 mm diameter rod is required at module
cells (M, P)
or pack level. The depth of penetration is at least through the entire cell
1–4
3–8
100
CP
for cell level testing and through cells or 100 mm (whichever is greater)
China
for module or pack testing [61,65,72,74,75]. In all cases the rod

remains in place during the post-test observation period (e.g. 1 h).
For FreedomCAR:2005 [65] this is not explicitly mentioned.
Cell (C) 3 cells or
AIS-048 [74]
The usefulness of this test is questioned by many in the research

3 (C) 20 (M)
100 mm (M) community [57,84], for three main reasons: first the test is not fully
CM
8
representative of an event that would likely occur in a real - situation

India
(e.g. a sharp object penetrating inside the battery compartment within

a vehicle), second it has been proven that the test does not represent a
spontaneous internal short circuit [57,85,86] and finally there are
KMVSS 18-3
many parameters that can strongly affect the outcome of the test, for
[73]
example: nail speed, nail dimension and SOC of the battery [60,87,88].
Korea
Furthermore, it is uncertain as to the influence the quality and

composition of the nail material may have on the outcome of the test
Cell (C) 3 cells or 100 mm (M, P)
and standards and regulations do not provide guidance on this. For all
Freedom CAR
the reasons mentioned above it seems comprehensible that this test is

[65]
not included in many of the automotive standards and regulations as

3 (C) 20 (M, P)
displayed in Table 5.
CMP
100
8
3.1.4. Immersion test

USABC [72]
USA
The immersion test has been developed to simulate a situation

where a battery is submerged or where battery assemblies installed in
the underbody of the vehicle are partially flooded. In order to perform
this test, the (fully charged) battery is completely submerged in salt
UL 2580
water with a composition similar to seawater (e.g. 5 wt% NaCl (aq.)) at

[63]
25 °C for a period of at least 1–2 h or until any visible effects (e.g.

bubbling) have stopped [61,63,65,66,72,73] (see Table 6). The immer-
sion into other liquids, such as engine coolant or fuel, is also
UN/ECE-R100.02
recommended in FreedomCAR:2005 [65].

EU and further
In the latest version of ISO 12405, part 3 published in 2014 [69],

[62]
countries#
the water immersion test is newly introduced (not included in parts 1

and 2 [67,68]). Unfortunately, the test is not described in much detail
and it is merely pointing out that the consequence of the test is a short
circuit with hazardous gases possibly being released.
It is important to evaluate the frequency or likelihood of an electric
IEC 62660-2
(3) [70,71]
vehicle exposed to a flood situation. For example in the Netherlands

around 700–800 car accidents per year result in vehicles flooded in a
ditch or canal [89]. Around 1200 to 1500 vehicles end up submerged in
water in the United States every year [90]. Additionally, there are quite
frequent situations where hurricanes or storms cause numerous
(2)(3) [67–69]
ISO 12405-1
Test conditions for the penetration test at cell (C), module (M) and pack (P) level.
vehicles to be submerged. Just to mention some examples, up to

250,000 cars were destroyed in Hurricane Sandy in 2012 [91] and
International
several incidents involving EVs occurred (e.g. 16 submerged Fisker

Karma’s in Hurricane Sandy leading to fire [92]). Another example
where moisture led to a fire event was the BAE Systems HybriDrive
SAE J2929
incident [93]. Based on these incidents, it seems that the performance

[66]
of the test would be of relevance, however many of the standards and

regulations do not include this test (see Table 6).
3.1.5. Crush/crash test

SAE J2464 [61]
Cell (C) 3 cells or

3 (C) 20 (M, P)
100 mm (M, P)
In this test, the applied crush force emulates a vehicle accident or any
95–100
CMP
external load force that may damage the battery enclosure and cause its
≥8
deformation. In the crush test, also referred to as battery enclosure

integrity test [66], an electrically insulated plate usually textured or ribbed
[62,65,72] is pressed down onto the battery until a certain compression is
Diameter of Rod (mm)
SOC (% rated capacity)

Region of applicability
reached (e.g. crush to 85% of initial dimension and after 5 min continue
Minimum Depth of
crushing up to 50% of initial dimension [61,65,72,75]) or until an abrupt

voltage drop is observed (e.g. reduction by 1/3 of original cell voltage
Level (C, M, P)
Speed (cm s−1)
Penetration
[70]). Two standardized crush surfaces are normally used, type A and type
Penetration
parameters
B whose characteristics and dimensions are displayed in Fig. 2. Typically,

for cell level testing (cylindrical or prismatic), type B crushing bar (as
Table 5
described in IEC 62660-2:2011 [70]) is used. For module or pack

assemblies crushing plate type A is generally recommended.
#
1434
743 [75]
QC/T
China
AIS-
048
[74]
India
KMVSS 18-3 [73]
range of a vehicle
Max. operating
or 80% SOC
0.6 M NaCl
25 ± 5
P
1h
Korea
≥ 2 h or until any visible reactions have

25 ± 2 ± 5% of reading
FreedomCAR [65]
Nominal composition of sea water

CMP
stopped
100
USA
USABC
[72]
MP
25
> 1 h or until any
Fig. 2. Crushing plate a) type A and b) type B. Reprint from IEC 62660-2 ed.1.0
visible reactions
Max. operating
UL 2580 [63]
have stopped
5 wt% NaCl
Copyright © 2010 IEC Geneva, Switzerland. www.iec.ch.3 3 “The author thanks the
25 ± 5
SOC
MP
International Electrotechnical Commission (IEC) for permission to reproduce

Information from its International Standard IEC 62660-2 ed.1.0 (2010). All such
extracts are copyright of IEC, Geneva, Switzerland. All rights reserved. Further
information on the IEC is available from www.iec.ch. IEC has no responsibility for the
placement and context in which the extracts and contents are reproduced by the author,
R100.02 [62]
nor is IEC in any way responsible for the other content or accuracy therein”.
countries#
UN/ECE-
EU and
further
Guidance on crush plate position is missing in most of the cases.

SAE J2464:2009 [61] requires the test to be performed at the most
vulnerable location to include the main cell area, whereas UN/ECE-
R100.02:2013 [62] allows the manufacturer together with the technical
(3) [70,71]
62660-2
service (e.g. certified testing body) to decide the plate position taking
IEC
into consideration the direction of the travel of the DUT relative to its
installation in the vehicle.
Some standards indicate that the force to be applied in the crush test
operating range (HP)
Max. SOC at normal
Clear or salty water
ISO 12405-3 [69]
> 50% normal
operation (HE)
has to be limited to 1000 times the weight of the battery [61,65,70,72]

Not specified
whereas others have a fixed force (e.g. 100 kN [62,63,66]) ( Table 7),
25 ± 2
Test conditions for the immersion test at cell (C), module (M) and pack (P) level.
independent of the size of the battery to be tested. The implementation

of the first option might lead to some issues testing traction batteries. As
International
pointed out by Wech et. al. [94] maximum forces of less than 1000 times
the battery weight might not be sufficient to achieve the required
(2) [67,68]
compression (e.g. 50% of battery dimension [61,65,72,75]). This would

12405-1
ISO
be the case for pack level testing following USABC:1999 [72],

HP: High power applications. HE: high energy applications.
FreedomCAR:2005 [65] and SAE J2464:2009 [61]. For example, in

the case of the small battery (24 kg, 0.8 kW h) of the Mercedes-Benz
J2929 [66]
S400 HYBRID , only 11% deformation would be achievable [94].

≥ 2 h or until any visible
reactions have stopped
SAE
Applying this requirement to a full HEV battery pack with 1.5–

P
5 wt% NaCl
3 kW h, or to an EV battery pack with 15–35 kW h, having weights

95–100
25 ± 5
ranging from 50 to 200 kg, would require a minimal load of 500–

J2464 [61]
2000 kN. This is an unrealistic scenario, as maximum loads rarely

SAE
MP
exceed 200 kN based on crash test simulations [94].

Another aspect that can raise some concern is the comparability of
results between tests performed at component and vehicle level.
Temperature (°C)
Immersion Fluid
Investigations of real world accident scenarios on occurrence of defor-

Immersion time
Level (C, M, P)
mations in selected positions of the vehicle, together with simulations on

SOC (% rated
applicability
capacity)
parameters
Immersion
fuel cell vehicles equipped with a compressed hydrogen storage system

Region of
lead to the conclusion that maximum contact loads are usually <
Table 6
100 kN [95]. Applied crush force on the DUT at component and vehicle
level might not be comparable as in the latter case the battery has extra
#
1435
V. Ruiz et al.
Table 7
Test conditions for the crush test at cell (C), module (M) pack (P) and for the crash test at vehicle (V) level.
Region of International EU and USA Korea India China

applicability further
countries#
Crush SAE J2464 SAE J2929 ISO ISO 12405 (3) IEC UN/ECE- UL 2580 [63] USABC [72] FreedomCAR KMVSS18- AIS- QC/T 743
parameters [61] [66] 12405- [69] 62660-2 R100.02 [62] [65] 3 [73] 048 [75]
1 (2) (3) [70,71] [74]
[67,68]
Level (C, M, CMP P P C CMP CMP CMP CMP CP

P)
Crush speed 0.5–1 (C) 5–10 5–10 5 – 10 (M, P)
(mm mi- (M, P)
n−1)
Crush force ≤ 1000 DUT 100 kN or 100–105 kN** ≤ 1000 100– ≤ 1000 DUT weight (C) 100 ± 6 kN (M, P) ≤ 1000 DUT weight
weight expected DUT 105 kN***
intrusion as weight
per FMVSS
305 [81]*
Crush plate B (cylindrical C) A A B A (≤ 600 mm × B (C) A (M, P) A Solid cylindrical Crush area
type Parallel to 600 mm) impactor half the > 20 m2
1436
crush surface cell average (C)
(prismatic and diameter (C) A (M, A (30 mm
pouch C) P) × 150 mm)
A (M, P) (P)
SOC (% rated 95–100 95–100% > 50% normal. 80 (HEV) > 50% normal 80 (HEV) 100 100
capacity) max. normal operating range operating 100 (BEV) (C)
vehicle (HP) range
operation Max. SOC at 100 Max. operating SOC (M, P)
normal (BEV)
operation
(HE)
Vehicle level FMVSS 305 Relevant UN/ECE-R12
(V) [81] ## national or [78], UN/ECE-
regional R 94 [79], UN/
regulation on ECE-R 95 [80]
vehicle crash
test
* If due to battery packaging location, no battery enclosure deformation is expected, this requirement is presumed to be met. The responsible organisation shall be responsible to make and document this conclusion. ** or a value determined by the
customer depending on expected forces in vehicle crash tests. These values shall be based on appropriate analysis, e.g. vehicle crash tests or vehicle crash simulations. ***A higher crush force, a longer onset time, a longer hold time, or a
combination of these, may be applied at the request of the manufacturer. ## or equivalent regionally applicable for vehicle front, rear and side crash conditions. #Norway, Russia, Ukraine, Croatia, Serbia, Belarus, Kazakhstan, Turkey, Azerbaijan,
Tunisia, South Africa, Australia, New Zealand, Japan, South Korea, Thailand and Malaysia.
protection provided by the chassis and battery enclosure. Additionally,
743 [75]
China
QC/T
there could be some disagreement centred around the fact that real
world accidents have a dynamic nature, that is the battery is moving
towards the impact zone, which is different from component level tests
where the crush plate moves towards a static battery. Various published
AIS-048 [74]b
360º min−1
investigations have shown discrepancies between current standards and
India
90º h−1
M
regulations and dynamic crash tests, and hence the authors of these
investigations recommend appropriate modifications to the tests in-
cluded in the regulatory framework [94,96].
The crush test can also be performed at vehicle level, the so-called
KMVSS 18-
Crash test. Electric vehicles shall comply with the crash safety
Korea
3 [73]
requirements as for conventional vehicles. In Europe, vehicles have
to pass the tests defined by the UNECE: steering mechanism, front
impact, and side impact tests from UN/ECE-R12:2012, UN/ECE-
R94:2012 and UN/ECE-R95:2011 [78–80] as described in UN/ECE-
Freedom
CAR [65]
R100.02:2013 [62]. In the USA, vehicles need to comply with the test
defined in the Federal Motor Vehicle Safety Standards, such as frontal
360º min−1
90º h−1
rigid barrier, a side moving deformable barrier, a rear rigid barrier and
MP
100
a rear deformable barrier according to FMVSS 305:2011 [81].
USABC [72]
Following the crush test, a roll over test, is performed in some cases.
This is the case for SAE J2929:2013 [66], described in Section 4.1.f
(Table 7).
USA
360º min−1 in 3 mutually

3.1.6. Rollover test
perpendicular directions
Max. operating SOC

This test, also referred to as rotation test, simulates overturn of a
UL 2580 [63]a
vehicle that might occur in an accident. Comparison of the different test
parameters is presented in Table 8. In order to perform the test, the
P
battery pack or module is slowly rotated (e.g. 6° s−1) for one complete
revolution (360°) in order to evaluate the presence of any leak (e.g.
battery electrolyte, coolant liquid) or venting. Then, the DUT is rotated
in 90° increments for another full rotation staying at each position for
R100.02 [62]
one hour [61,63,65,72,74].

UN/ECE-
The rollover test is usually performed after crash tests such as

EU and further
described in FMVSS 305:2011 [81] or after a crush test as described in

SAE J2929:2013 [66]. In relation to the Korean standards, rollover
countries#
testing is not part of Article 18-3, discussed in this review, but part of
IEC 62660-2
Article 91 (fuel system). The need for this testing is supported by the
(3) [70,71]
fact that around 220,000 light motor vehicles sustain rollover crashes
in the US annually, which accounts for almost a third of all highway
vehicle occupant fatalities [97]. Despite this fact, the rollover test is not
included in various relevant standards and regulations, such as UN/
ISO 12405-1
ECE-R100.02:2013 [62].
(2)(3) [67–
It is interesting to point out the discrepancy between standards with

69]
respect to the applicability of the rollover test for certain types of

battery chemistry: while UL 2580:2013 [63] mentions specifically that
International
flooded lead acid batteries are not subjected to this test, AIS-048:2009
SAE J2929 [66]
[74], on the contrary, states that the test is applicable only for flooded
90º (within 60–
95–100% max.
normal vehicle
Test conditions for the rollover test at module (M) and pack (P) level.
180 s)/5 min
operation
lead acid batteries.

P
3.1.7. Vibration test

Although vibration occurs in any driving environment under normal
operating conditions, because it may be considered abusive to the
Applicable only for flooded lead acid batteries.
SAE J2464
360º min−1
battery, almost all of the standards and regulations evaluated in this

95–100
90º h−1
[61]
MP
review include a vibration test. The purpose of this test is to evaluate the
effect of long-term vibration profiles – representative of driving – on the
Except for flooded lead acid batteries.
battery, both in terms of the durability and in terms of identification of

design flaws. The vibration profiles vary quite considerably over a wide
Rotation speed (continuous
revolution/hold time per

Rotation steps incremental
range of frequencies and amplitudes (see Table 9). In order to facilitate

SOC (% rated capacity)
Region of applicability
the comparison of the various vibration parameters, Fig. 3a and b

Rollover parameters
display the sine wave and random profiles used in the evaluated
standards and regulations. Sine swept testing is commonly used to
revolution)
increment
identify product resonances, while random vibration simulates everyday

Level (M, P)
life scenarios that a DUT would experience [98,99].

The vibration profiles in standards and regulations are derived
Table 8
from generic measurements from conventional vehicles at locations

b
a
appropriate for mounting traction batteries in EVs. In fact, there are

#
1437
V. Ruiz et al.
Table 9
Test conditions for the vibration test at cell (C), module (M), pack (P) and vehicle (V) level.

countries#
Vibration SAE SAE J2929 [66]c ISO 12405-1 IEC UN/ECE- UL 2580 [63] USABC Freedom KMVSS AIS-048 QC/T
parameters J2464 (2)(3) [67–69]c 62660-2 R100.02 [62] [72] CAR [65] 18-3 [73] [74] 743 [75]
[61] (3) [70,71]
Level (C, M, P) CMP P C CMP C Mf P CMP M P

g
Type of profile Random Sine wave Random Random Sine wave Random or Sine wave or Random Sine wave Linear
logarithmic sweep
sweep
Frequency range 10–190a 7–200d 5–200e 10–2000e 7–50b 10–2000e 10–190a 10–190a 30–150 10–55
(Hz)
PSD wave 0.4–11a 0.02–5.8e 0.14–20e 0.14–20e 0.4–11a 0.4–11a
random
(m s−2)2/Hz
Loading range 9.81–19.6 ( > 12 kg)d 2–10b 7–49 30 30
sine wave 9.81–78.5 ( < 12 kg)d
1438
(m s−2)
Axis Vertical, Three mutually Vertical, Vertical Vertical, longitudinal, lateral Vertical axis Vertical
longitudinal, perpendicular longitudinal, and
lateral mounting positions of transversal horizontal
the cell
SOC (% rated 95–100% max. normal vehicle operation 50 80 (HEV) > 50% normal 80 (HEV) 100 & 20 (Z- 100 & 20 (Z-sine, random) 100
capacity) 100 (BEV) operating range 100 sine, random) 60 (X & Y-sine, random)
(BEV) 60 (X & Y-sine,
random)
Vehicle level (V) SOC at normal vehicle operation
PSD: Power spectral density

#
a
Based on SAE J2380:2009 [102].
b
Higher can be requested by manufacturer.
c
A profile 'which reflects the application' may be used as alternative.
d
Based on UN 38.3:2015 [76].
e
Based on IEC 60068-2-64:1993 [101].
f
At the module level for those electric energy storage assemblies intended for use in applications larger than passenger vehicles. The module level testing shall be representative of the electric energy storage assembly.
g
Vibration endurance test in accordance with the anticipated end application vehicle vibration profile.
(cell level testing, up to 2000 Hz) due to the difference in mass of the DUT.
Lastly, SAE J2380:2009 [102] is also widely used to define random
vibration profiles. Actually this standard is the basis for SAE
J2929:2013 [66], UL 2580:2013 [63] (module and pack level only),
USABC:1999 [72] and related FreedomCAR:2005 [65] standards and it
reflects rough-road measurements at locations where traction batteries
are likely to be installed in EVs/HEVs, equivalent to 100,000 miles
usage.
Interestingly, ISO 12405 part 1 and 2 [67,68] are to our knowledge
the only standards, that require vibration testing at different ambient
temperature conditions, namely at + 25 °C, + 75 °C and − 40 °C. The
combined effect of vibration and temperature could certainly be
relevant during in-use situations. However, a malfunction of the
cooling and/or heating unit of the vehicle is required to observe such
extreme temperatures.
Fig. 3a highlights how regulation UN/ECE-R100.02:2013 [62]
requires significantly milder requirements compared to the other
standards. Additionally, this regulation is one of the few documents
that require performing the test in only one axis (vertical axis), whereas
other standards require testing in two or three axes.
Vehicle level testing can also be performed as mentioned in SAE
J2929:2013 [66] under conditions defined by the testing body.
3.2. Electrical tests
In this section the series of abusive tests to evaluate the electrical

safety of the devices will be presented: external short circuit or short
circuit protection test, internal short circuit test and overcharge/over
discharge protection tests. Some other standards, which do not
describe abusive tests, but still of relevance to electrical safety are also
mentioned: general electrical functional safety requirements of elec-
trically propelled road vehicles are specified in the international
standard ISO 6469 series. Part 1 of this standard particularly covers
the Rechargeable Energy Storage System (RESS) safety specification
[103], while part 3 deals with general protection against electric shock
[104] and part 4 deals with post-crash safety requirements, aimed at
the protection of persons inside and outside the vehicle [105]. ISO
17409:2015 [106], deals with the safety of the electric vehicle during
conductive charging. This document, which was initially part of the IEC
61851:2010 [107] series, focuses on electrical risks. Another document
for wireless charging, ISO 19363 [108], is in an earlier stage of
development (preparatory stage).
Fig. 3. Vibration profiles for the various standards and regulations requiring: a) sine
wave profile and b) random profile. 3.2.1. External short circuit test
The purpose of this test is to evaluate the safety performance of a
DUT when applying an external short circuit. The test can evaluate the
very few works published on vibration profiles designed specifically to activation of the overcurrent protection device or the ability of cells to
EVs and HEVs. Work published by Hooper et. al. pointed out that withstand the current without reaching a hazardous situation (e.g.
many of the vibration profiles described in the standards represent a thermal runaway, explosion, fire). The main risk factors are heat
short term abuse rather than a mechanical durability test to represent a generation at cell level (thermal runaway [109]) and arcing which
battery life. Additionally it is suggested that battery packs may be may damage circuitry or may lead to reduced isolation resistance.
exposed to vibration loads outside the range evaluated in existing The most relevant test parameters are presented in Table 10.
standards [100]. During the test a low resistance element (e.g. 5 mΩ [62,70,72],
Vibration test profiles in standards are adapted in most cases from UN 20 mΩ [63,68] or 100 mΩ [67]) is connected externally across the
38.3:2015 [76], IEC 60068-2-64:1993 [101] or SAE J2380:2009 [102]. battery terminals in less than one second and maintained for a defined
The vibration profile described in UN 38.3:2015 replicates vibrations period of time (e.g. 10 min). As a consequence, current flows through
during transport. This test is not considered relevant for the evaluation of the system until an overcurrent protection device – if present – limits
battery resilience to vibration during driving conditions, since test the current [72]. Typically fuses, circuit breakers (passive elements)
conditions are not representative of the position of the battery and its and contactors (active elements) are used to protect against over-
fixture in the vehicle. For this reason the transport regulation is not currents at module or pack level. At cell level, built in current
discussed in detail in this review. Despite this, UN 38.3:2015 testing is interruption devices (CID) or positive thermal coefficient (PTC) devices
provided as an alternative within SAE J2929:2013. On the other hand, can be used, which disconnect the internal current collector from its
standard IEC 60068-2-64:1993 [101] has been taken as the basis for IEC terminal or limit the passage of current if the inside pressure and/or
62660-2(3):2011(2013): [70,71], ISO 12405-1(2):2011(2012) [67,68] and temperature reach a certain limit. All these protection devices have a
UL 2580:2013 [63]. It defines different test conditions for battery pack time characteristic (e.g. for circuit breaker IEC 60898-1:2015 [110]) of
testing (up to 200 Hz) and for the electronic devices of the battery pack how quickly they limit or interrupt the current. The higher the current,
1439
V. Ruiz et al.
Table 10
Test conditions for the short circuit test at cell (C), module (M) and pack (P) level.

Short circuit SAE J2464 [61] SAE J2929 [66] ISO 12405-1 (2) IEC UN/ECE-R100.02 UL 2580 [63] USABC [72] Freedom CAR KMVSS 18-3 AIS-048 [74] QC/T
parameters (3) [67–69] 62660-2 [62] [65] [73] 743 [75]
(3) [70,71]
Level (C, M, P) CMP P P C CMP CMP CMP P CMP CP

Cooling medium Operational Operational if necessary Operational if Operational
for operation necessary for
operation
Passive short- Disabled or bypassed Operational Operational Operational if Operational Operational
circuit (C M P) or relevant to the
Protection Operational (C M) outcome of the test
device
Non-passive Disabled Disabled
protective
device
1440
Resistance (mΩ) Hard short: ≤ 5 and < < DUT DC impedance 100 (20)(c) ≤5 ≤5 ≤ 20 (UL 1642 ≤5a 50 ≤ 5b < 5
Soft short: ≥ 10 and resistance comparable to [114]) or 5 (C) ≤
DUT DC resistance 20 (M, P)
100 (20) (ISO 12405–
1(2) [67,68]) or 100
(UN 38.3 [76])d
SOC (% rated 95–100 95–100% max. normal 100 100 > 50% max. Max. operating 100 Max. operating 100 100
capacity) vehicle operation operating SOC SOC range of a vehicle
or 80% SOC
#
a
For systems with ≤ 5 mΩ internal resistance, a conductor of 1/10 of the minimum resistance of the cell/module shall be used.
b
For systems with ≤ 0.9 mΩ V−1 system voltage ± 0.1 mΩ internal resistance, a conductor of 1/10 of the minimum resistance of the cell/module shall be used.
c
ISO 12405-3: the test can be conducted at a lower resistance or higher temperature than specified in ISO 12405-1 (2), as appropriate for the DUT, according to agreement between the customer and the supplier.
d
Test temperature 55 ± 2 °C.
the faster they are typically able to interrupt it. If the current is not high works suggest that nail penetration is not representative of real field
enough (e.g. at low SOC) or if it drops quickly [111] the current may be situations [57,118]. For this reason, various alternative tests have been
not interrupted, potentially creating a hazardous situations. Therefore, developed in order to represent a more realistic scenario; however
standards require a hard short circuit when the external resistance is these tests have not been widely implemented in the legislative land-
minimal [62,70,72] or a soft short circuit when the external resistance scape. We will summarize here three of the most relevant tests:
is comparable with of the internal resistance of the DUT. In this case,
the soft short circuit will assure that the response of the cell is 3.2.2.1. Separator shutdown integrity test. The purpose of this test is
evaluated, rather than that of the protection device [61]. to evaluate the efficiency of the shutdown separator at high
As mentioned, standards require a fixed external resistance irre- temperatures and the possible failure propagation within cells
spective of the size of the DUT. However, the initial short circuit connected in series (in a module) as described in SAE J2464:2009
current is influenced by the size of the DUT [112,113] as well as by its [61]. In this test the cell shall be heated to a temperature slightly above
type of connection (i.e. parallel, serial or a combination thereof). the shutdown temperature (i.e. ≥ 5 °C). For detailed explanation on
Consequently, applying the same external resistance to DUTs having how to measure the shutdown temperature, please refer to the
different sizes and types of connection, may result in not necessarily standard. Once the temperature is stable for 10 min, a voltage above
comparable initial short circuit currents per cell. Therefore some (or equal to) 20 V is applied at a maximum current of 1 C and
standards indicate for hard short conditions, that the external resis- maintained for 30 min (or until separator failure).
tance needs to be much smaller than the DUT DC impedance [61] or 1/
10 of the minimum resistance of the cell/module for systems with less
than 0.9 mΩ V−1 system voltage ± 0.1 mΩ internal resistance [74], as 3.2.2.2. Forced internal short circuit or nickel particle test. The
the initial short circuit current depends on the internal resistance of the international standards IEC 62133-2:2017 [119] and IEC 62660-
DUT. For soft short conditions, when the external short circuit 3:2016 [71] provide detailed instructions for the internal short
resistance is higher than that of the DUT, the initial short circuit circuit test for cylindrical and prismatic type cells. The test, which is
current is governed primarily by the external resistance, therefore performed at two temperatures, +10±2 °C and + 45±2 °C, requires the
resulting in initial short circuit currents independent of the size of the disassembly, insertion of an L-shaped nickel particle (e.g. between
RESS. positive coated area and negative coated area, between positive
Temperature affects the internal resistance of a battery, i.e. the rate active material and separator) and reassembly of the cell. A short
of electrochemical reactions and transport; therefore a higher initial circuit is subsequently induced with a pressing machine at a speed of
current can be generated at elevated temperature, which creates more 0.1 mm s−1.
heat. Moreover the higher the temperature, the closer the DUT
temperature is to the onset temperature of thermal runaway. At low This test has obvious drawbacks due to the need to manipulate the
temperature, the activation of the protecting device (e.g. fuse, circuit cell. As an alternative, the particle could be introduced during the
breaker) can be inhibited, or the time to interruption may increase. manufacturing process.
Only UN 38.3:2015 and UL 1642:2007 require a short circuit test to be
performed at a temperature higher than room temperature (55 ± 5 °C) 3.2.2.3. Blunt rod indentation test. Another ISC variation, also
[76,114]. Standards and regulations specific to electric vehicle applica- referred to as Indentation-Induced ISC (IIISC) was developed by
tions (Table 10) do not require increased temperature testing. Underwriters Laboratories and NASA [120]. It entails the application
However, it may be considered reasonable that the short circuit test of a mechanical force to the cell/battery, using a blunt rod instead of a
needs to be performed at temperatures higher than room temperature, sharp one, in order to deform the most outer electrode layers and
which are likely to be reached during driving or when the cooling eventually create a short circuit. The rod speed applied is several orders
system is malfunctioning. In addition, none of the standards and of magnitude lower than that used for the penetration tests
regulations considers low temperature as a safety problem, where (0.01 cm s−1 vs. 8 cm s−1).
dendrite formation is prone to occur. Overall, it can be concluded that these alternative tests exhibit
Another parameter that influences greatly the outcome of the uncertainties and difficulties, mostly from a practical point of view.
test is the SOC. The worst case is achieved at high SOC, as the Researchers are still actively looking for ways to evaluate the ISC
initial short circuit current created is maximum [112] and the onset hazard in a more realistic and practical way, allowing successful
temperature of thermal runaway is lowest [115]. Consequently, implementation of these tests in future automotive safety tests.
most of the standards require testing at 100% of the rated capacity Alternative approach taken by some battery manufacturers consists
(Table 10), however in the case of UN/ECE-R100.02:2013 [62], the on designing systems where cell to cell propagation is hindered or
test can be performed at 50% SOC (or above) of the maximum designing packs able to contain a potential thermal runway within.
operating SOC value.
3.2.2. Internal short circuit test 3.2.3. Overcharge/overdischarge test

Standardisation of the internal short circuit (ISC) test is under In order to evaluate the functionality of the overcharge/over-
development, and no regulation dealing with batteries for automotive discharge protection system, the battery is charged or discharged
applications requires this test. The occurrence of internal short circuits, beyond the limits recommended by the manufacturer, situations that
one of the main concerns for battery manufacturers, potentially leads to could occur due to a charger failure, for example. The relevance of the
venting, thermal runaway, along with sparking which can ignite the test is underlined by the fact that almost all evaluated standards and
electrolyte vapours escaping from the cell [57,116]. The generation of regulations (with the exception of overdischarge in AIS-048:2009 [74])
these internal shorts can be triggered by manufacturing imperfections, require its application. A summary of test parameters is presented in
presence of impurities in the cells, dendritic growth of lithium etc. [57] Table 11.
and leads to most of in-field safety incidents [117]. Multiple internal The main safety risks during overcharge are the decomposition of
short circuits scenarios are possible (e.g. electrical contact of cathode/ the electrolyte [109,121], cathode and anode breakdown, exothermic
anode, aluminium current collector/copper current collector, aluminium decomposition of the SEI layer, separator degradation, and the Li
current collector/anode) each with a different contact resistance [117]. plating [122], which can lead to self-heating of the battery and thermal
As mentioned previously, mechanical nail penetration tests aim at runaway. Also fluorinated binders, such as polyvinilidene fluoride
investigating the effects of an internal short circuit, however some (PVDF), have been found to react exothermically with lithiated carbon
1441
Table 11
Test conditions for the overcharge/overdischarge test at cell (C), module (M), pack (P) and vehicle (V) level.

V. Ruiz et al.
countries#
Overch./disch. SAE J2464 [61] SAE J2929 [66] ISO 12405- IEC UN/ECE- UL 2580 [63] USABC FreedomCAR [65] KMVSS18-3 [73] AIS- QC/T 743 [75]
parameters 1 (2)(3) [67– 62660- R100.02 [62] [72] 048
69] ### 2 (3) [74]
[70,71]
Overcharge
Level (C, M, CMP P P C CMPV CMP CMP MP P CMP CP
P,V)
Passive Operational Operational Operational Operational if Operational Operational Electronic
overcharge relevant to the protection
Protection outcome of the circuit removed'
device test
Non-passive Disabled Disabled Disabled Disabled
protective
device
Charge rate a) 1C CC and b) at max. current supplied Max. possible 5C (HP) 2C 5It ≥ C/3 and < 5It (HEV) (C) 1It 32 A CC* and < 450 Vdc According to C/10 3I3 (≈ 9I3(≈
by regenerative braking or charging rate for the (HE) (HEV) max. current (BEV) (C) Max. manufacturer's 1C) 3C)
system (or 3C) (C) 1C CC (M, P) application 1It (BEV) within the specified charging recommendation or
normal rate (M, P) 32 A CC until 1.5 x
operating range rated voltage
specified by the followed by constant
manufacturer voltage
+ ++
End of charge > 200% SOC or destructive factor (e.g. > 130% 2 Vmax or 2 x rated 2 Vmax reached or 200% SOC 200% SOC or 4 h or 150% SOC or 2.5 h 10 h 5 V or 10 V
thermal runway) SOC, > 55 °C reached capacity 200% SOC (C) +++ or 4 h DUT fail after full charging 90 min
1442
or ++ or 200% or 110% rated
SOC capacity, or a
manufacturer
specified limit or
DUT failure
(explosion, fire)
(M, P)
Overdischarge
Level (C, M, CM P P C CMPV CMP CMP MP P CP
P,V)
Passive Operational Operational Operational Operational if Operational Operational Electronic
overdi- relevant to the protection
scharge outcome of the circuit removed'
Protection test
device
Non-passive Disabled Disabled Disabled Disabled
protection
device
Discharge rate Max. recommended current 1C (HEV/PHEV) 1C (HP) C/3 1It ≥ C/3 and < 1It (C) Max. 1C 1C I3(≈ C/3)
C/3 (BEV) (HE) max. current specified
within the discharge rate (M,
normal P)
operating range
specified by the
manufacturer
++ ##
End of − 100% SOC (C) or 0.0 ± 0.2 V (M) ** or 0 ± 0.2 V 25% of 90 min or 25% of (M, P) 90 min 1.5 h or 1.5 h or until 50% of 30 min 0 V (C) A cell is 0
discharge nominal nominal voltage (C) until every subassemblies have V (P)
(continued on next page)
Table 11 (continued)

V. Ruiz et al.
countries#
Overch./disch. SAE J2464 [61] SAE J2929 [66] ISO 12405- IEC UN/ECE- UL 2580 [63] USABC FreedomCAR [65] KMVSS18-3 [73] AIS- QC/T 743 [75]
parameters 1 (2)(3) [67– 62660- R100.02 [62] [72] 048
69] ### 2 (3) [74]
[70,71]
voltage, cell is achieved voltage

30 min after reversed reversal for 15 min
passing the for 15 min
normal
discharge
limits or ++
CC: constant current. It : current in amperes which is expressed as It (A) = Cn (Ah)/1 (h) where Cn is the rated capacity of the cell ; n is the time base (hours) for which the rated capacity is declared. Vmax maximum voltage specified by the
manufacturer. + Until the charge device voltage is reached or the connection interface disconnects battery from charge device. ++ Until the DUT interrupts the charging (discharging) by an automatic disconnect of the main contactors. +++
Termination by protective circuitry whether it is due to voltage or temperature controls. #Norway, Russia, Ukraine, Croatia, Serbia, Belarus, Kazakhstan, Turkey, Azerbaijan, Tunisia, South Africa, Australia, New Zealand, Japan, South Korea,
Thailand and Malaysia. ## Until the passive protection device(s) are activated, the minimum cell voltage/maximum temperature protection is activated, or the DUT has been discharged for an additional 30 min after it has reached its specified
normal discharge limits, whichever comes first. ### Higher temperature can be used according to agreement between the customer and the supplier. *When performing this test at less than the pack level, the voltage (series pack configuration) or
the voltage/current (series/parallel pack configuration) shall be scaled down appropriately. ** Until the connection interface disconnects battery from discharge load.
1443
Table 12
Test conditions for the thermal stability test at cell (C), module (M) and pack (P) level.

Thermal stability SAE J2464 [61] SAE ISO 12405- IEC 62660-2 UN/ECE- UL 2580 [63] USABC [72] FreedomCAR [65] KMVSS18-3 [73] AIS- QC/T 743
parameters J2929 1 (2)(3) (3) [70,71] R100.02 [62] 048 [75]
[66] [67–69] [74]
Level (C, M, P) C C C CMP P CP

Heating rate ≥5 5 5 5–10
(C min−1)
Heating steps 5 °C 5 °C (C) 10 or 5 °C (C) 10 °C (M, P)
20 °C (M, P)
Termination 300 °C above maximum 130 ± 2 °C 130 ± 2 °C Additional self-heating is detected or 200 °C 80 ± 2 °C 85 ± 2 °C
operating temperature or above maximum operating temperature or
catastrophic event catastrophic event
Holding time 30 or until self-heatinga 30 30 30 (C) 120 (M, P) or until self-heating 240 120
(min)
Repetition in case 2 °C heating steps - hold 2 °C heating steps - 2 °C heating steps - hold
of self-heatinga for > 1 h hold for > 1 h (C) for > 1 h
SOC (% rated 95–100 or lower SOC at 80 (HEV) 100 80 (HEV) 100 100 Max. operating 100
capacity) which thermal stability is (BEV) (BEV) range of a vehicle
degraded or 80% SOC
#
a
Self-heating is defined when temperature increases at a rate higher than 1.0 °C min−1.
V. Ruiz et al.
Table 13
Test conditions for the thermal shock test at cell (C), module (M) and pack (P) level.

countries#
Thermal SAE J2464 [61] SAE J2929 [66] ISO IEC 62660-2 (3)a UN/ECE- UL 2580 USABC [72] Freedom KMVSS AIS- QC/T
Shock 12405-1 [70,71] R100.02 [63] CAR [65] 18-3 [73] 048 743
parameters (2)(3)a [62] [74] [75]
[67–69]
Level (C,M, CMP CMP P C CMP CMP CMP

P)
Protection Active thermal controls (primary and secondary) disabled Thermal Without With Protection Active Thermal controls (primary and
device Not defined in UN 38.3 alternative controls electrical electrical devices shall thermal secondary) disabled
disabled operation operation be controls
operational (primary and
secondary)
disabled
1444
Tmax (°C) + 70 ± 2 or ± 5% of reading + 70 ± 2 or ± 5% of reading or + 72 ± 2 + 8 5 ± 1b + 85 ± 2 + 65 ± 2 + 60 ± 2 + 85 ± 2 + 80
Tmin (°C) − 40 ± 2 or ± 5% of reading − 40 ± 1 − 40 ± 2 − 20 ± 2 − 40 ± 2 − 40 ± 2 − 40
Hold time ≥ 1 (C) ≥6 (M)c (P) ≥ 1 (C) ≥ 6 (M)c (P) or ≥ 12 (C > 0.5 kg M, P > ≥ 1 6 (3) 1.5 (Tmin) 1.83 (Tmax) 6 1.5 (Tmax) (C) Appropriately ≥ 1 (C) ≥ 6
(h) 12 kg) 1.83 (Tmin) adjusted (C) ≥ 6 (M)c (P)
(C) ≥ 6 (M)c (M)
(P) appropriately
adjusted (P)
Repetitions 5 5 or 10 5 30 5 30 5
SOC (% rated 95–100 95–100% max. normal vehicle operation 50 (HP) 80 80 (HEV) 60 (HEV) ≥ 50% max. 80/60 (HEV) 50
capacity) (HE) 100 (BEV) 80 (BEV) operating 100/80
SOC (BEV) (C)
Max.
operating
SOC (M, P)
#
a
Or Tmin Tmax as specified by manufacturer.
b
Part 3 of the standard allows: + 60 ± 2 °C with a hold time of 6 h.
c
As required to reach uniform temperature ( ± 5 °C).
if sufficient temperature is reached (e.g. 200 °C [60]). Factors affecting
743 [75]
China
QC/T
the outcome of the test are amongst others, the charging rate and the
finally reached SOC.
For the overcharge test, a controlled current is applied to the
battery (e.g. 1/3 It2 -rate) up to a set charge limit (e.g. 200% SOC
India
AIS-
048
[74]
[61,62,70,72], 110% SOC [63], 130% [67]) or until the tested-device
(automatically) interrupts or limits the charging. Although most of the
18-3 [73]
standards provide a general description for all types of energy storage

KMVSS
Korea
devices, others describe specific tests for EVs, HEVs and PHEV
applications (e.g. charging rate at 5 It [123] for HEV and 1 It for
BEV [70,71]). Some other standards recommend much lower C-rate
(e.g. C/10 in AIS-048:2009 [74], C/3 rate in UN/ECE-R100.02:2013
Rate comparable to a 3 kW
constant power rate for
[62]). Tobishima et al. showed that cells overcharged at low rates did
FreedomCAR [65]
Active thermal controls (primary and
recommended charge
not show any venting whereas those cells overcharged at a 2C rate did
Manufacturer's
entire DUT [124]. Golubkov et al. showed that NCA cells with SOC ≤ 100% had a
algorithm
secondary) disabled
thermal runaway onset temperature in the range 136–160 °C, whereas

overcharged cells (SOC > 100%) showed much lower onset tempera-
MP
100
20
tures (ranging 65–80 °C). Although serious events occur at cell level
USA
with significant overcharge (e.g. 2 times the rated capacity), repeated

charge/discharge cycling at moderate overcharge (110% SOC) can also
USABC [72]
lead to internal short and failure of the cell in only 10 cycles [125].
To address another scenario of great importance, an over-
1C
1C
discharge (or forced discharge) test is generally required. Safety

risks during overdischarge are polarity reversal leading to oxidation
of the anode current collector (Copper) and to plating on the cathode
2580
[63]
UL
side. Even minor over-discharge may cause dendrite formation and

finally short circuit [126]. During the overdischarge test, a fully
charged battery is discharged (e.g. 1C rate for 1.5 h [65], C/3 rate)
operation defined by the manufacturer
Steady current that will increase the
possible within the range of normal
until the tested-device interrupts or limits the discharging [62]. The

temperature of cells as rapidly as
EU and further countries#
Cooling system deactivated
great variability in test parameters found in the various evaluated

UN/ECE-R100.02 [62]
documents (Table 11) can lead to the conclusion that the outcome of
the tests might be dependent on the standard or regulation followed.
CMPV
For this reason harmonisation of testing parameters is required to

allow comparable testing.
3.3. Environmental tests
Environmental testing aims at evaluating the safety performance of

a system under conditions of temperature change, such as an accident
IEC 62660-
scenario involving fire, or extreme weather exposure in certain

[70,71]
2 (3)
geographic areas. In this section, the most common environmental

Test conditions for the overheat test at cell (C), module (M), pack (P) and vehicle (V) level.
tests, thermal stability, thermal shock, overheat and extreme cold

temperature and fire tests are described.
ISO 12405-
1 (2)(3)
[67–69]
3.3.1. Thermal stability test

This test evaluates the stability of a battery at an elevated
International
temperature to identify the temperature where thermal runway begins.

Cooling system
operating SOC
For this test, the temperature of the cell is increased sequentially in

Max. normal
Max. normal
SAE J2929
deactivated
Rate comparable to the intended
5 °C steps with a holding time of 30 min at each incremental step, until

[66]
rate
P
the temperature reaches 200 °C [65,72] above the maximum operating

temperature of the battery (or until a catastrophic event occurs such as
application
venting or major damage to the DUT). For modules and packs, the
Active thermal controls
increments of temperature are set to 10 °C with a longer holding time

20
secondary) disabled
As defined by the
SAE J2464 [61]
of 120 min [65,72]. Standard SAE J2464:2009 [61] has a higher

manufacturer
(primary and
95–100
threshold temperature of 300 °C above the maximum operating

MP
temperature. These tests require a second execution in order to refine

the exact start temperature of the thermal event. During the second
execution, the temperature is increased in 2 °C increments and held for
a minimum of one hour at each incremental step [61,65,72].
Some other standards evaluate the performance of the system at
Level (C, M, P,
Discharge rate
SOC (% rated
elevated temperature, not aiming at reaching thermal runaway, but at the

applicability
capacity)
Charge rate
parameters
Number of
Protection
Over heat
Region of
device
cycles
Table 14
2
V)
The current It represents the discharge current in amperes during one hour
discharge and C is the measured capacity of a battery pack (or a cell): It (A) = Cn
(Ah)/1 (h); n is the time base (hours) for which the rated capacity is declared.
#
1445
V. Ruiz et al.
Table 15
Test conditions for the fire test at cell (C), module (M), pack (P) and vehicle (V) level.

countries#
Fire parameters SAE J2464 SAE J2929 [66] ISO ISO 12405 (3) [69] IEC UN/ECE- UL 2580 [63] USABC [72] Freedom KMVSS 18-3 [73] AIS- QC/T
[61] 12405-1 62660-2 R100.02 [62] CAR [65] 048 743 [75]
(2) [67,68] (3) [70,71] [74]
Level (C,M, P, MP P P Va CMPV CMP CMP P

V)
Heat source Radiant heat Flame Flame Flame Flame Radiant heat Flame
Set-up Cylindrical Wire mesh screen Grating table (Fig. 4b) Grating table Wire mesh screen or Cylindrical Cylindrical Not specified
metallic fixture (Fig. 4a) (Fig. 4b) floor perimeter (C)b metallic fixture fixture
1446
(Fig. 4a) Floor
perimeter (M, P)
Tmax (ºC) 890 °C To be defined by Not specified Not specified 590 °C 890 °C 890–900 °C
Holding time at 10 min responsible testing 70 s 70 s 20 min 10 min 2 min
Tmax organisation
SOC (% rated 100 > 50% normal > 50% max. Fully charged (C) ≥ 80 100 Max. operating
capacity) operating range (HP) operating SOC Max. operating SOC range of a vehicle
Max. SOC at normal (M, P) or 80% SOC
operation (HE)
#
a
The vehicle body may be included.
b
The overall dimensions of the projectile test aluminium test screen may be increased to accommodate cells with dimensions larger than 127 mm (5 in.), but shall not exceed a distance of 305 mm (12 in.) from the cell in any direction.
Fig. 4. Examples of fuel fire test set ups: a) wire mesh screen (copied from UL 1642 [112]) and b) grating table (copied from UN/ECE-R34 [127]). Copyright © Underwriters
Laboratories Inc. UL 1642, 4th edition, 2007.
assessment of the thermal stability of the DUT at this temperature. 3.3.3. Overheat test
Examples include ISO 62660-2:2011 (High temperature endurance test) The overheat test, also referred to as rapid charge/discharge,
[70], QC/T 743:2006 (heat test) [75] and KMVSS18-3:2009 (Heat cycling without thermal management, single point thermal control
exposure test) [73] as they require placing the battery in a chamber and system failure, over-temperature protection test, aims at evaluating
increasing the temperature to only 130 °C, 85 °C and 80 °C, respectively. the effect of temperature control failure or failure of other protection
Although it seems that both variants of the test provide useful features against internal overheating during operation. Test para-
insight into the safety of the energy storage system, they are not widely meters required in this test are displayed in Table 14.
required as can be seen in Table 12. For this test, a fully charged DUT, whose active thermal control
system (e.g. cooling system) is disabled, is cycled (e.g. 20 cycles with no
resting period between charge and discharge [61,66]). As a conse-
3.3.2. Thermal shock test quence, the temperature of the DUT will increase. According to some
This test is designed to evaluate changes in the integrity of the DUT standards, the test must be performed in a closed volume in order to
arising from expansion and contraction of cell components upon evaluate the flammability of any materials being released from the
exposure to extreme and sudden changes in temperature (e.g. the battery during the test [61,65,66,72]. In this case, a spark source has to
vehicle is entering or exiting a heated garage, during transport [63]) be present to ignite any potentially flammable gases or vapours from
and potential consequences of such changes. During a thermal shock the the DUT or, alternatively, a gas concentration measuring device can be
DUT is exposed to two temperature limits and held at each temperature utilized as suggested by SAE J2929:2013 [66].
limit for a specified period of time. The thermal shock tests described in In the case of UN/ECE-R100.02:2013 [62], the test is stopped when
standards have different maximum temperature limits (see Table 13). either: (a) the DUT interrupts the charging/discharging to prevent
ISO 12405-1:2011 [67], ISO 12405-2:2012 [68], IEC 62660-2:2011 [70] temperature increase, (b) the temperature of the DUT is stable (i.e.
and UL 2580:2013 [63] have set the highest upper limit at +85 °C, while variation < 4 °C in 2 h) or (c) there is evidence of DUT damage (e.g.
the lowest upper limit is set at +60 °C in UN/ECE-R100.02:2013 [62]. electrolyte leakage, rupture, fire or explosion).
The lower temperature limit is − 40 °C in all the cases.
Noteworthy is that amongst all the documents evaluated, only UN/ 3.3.4. Extreme cold temperature test
ECE-R100.02:2013 [62] permits operation of the protection devices The rationale behind this test is the effect of possible exposure of
during this test. In the other standards the protection device is the DUT to low temperatures (e.g. vehicle parked in a cold environ-
disabled, which imposes harder testing conditions. ment). At low temperatures, the electrolyte has poor ionic conductivity
1447
and the anode experiences high over-potentials [127] with can lead to 3.4. Chemical hazards evaluation tests
dendrite formation. Metallic plating can be a safety concern because
growing dendrites could short circuit the cell. Despite these issues, only Lithium ion batteries contain, as mentioned in the introduction,
one standard deals with this topic. USABC:1999 [72] describes a significant amounts of potentially hazardous materials (e.g. highly
matrix for charging at the normal primary charge rate for the specific flammable electrolytes, corrosive and toxic components [16,134,135]).
system and discharging at 1C down to various DOD's (depths of If exposed to certain conditions, it is expected that the integrity of the
discharge): 20, 50, 60, 80, 100% at the following temperatures: − 40, battery is compromised which may lead to electrolyte leakage, venting,
− 20, 0 and 25 °C [72]. The liquid coolant is present, but not circulating rupture or even fire and explosion. Amounts of gas released from various
during the test. The test shall be stopped if abnormal conditions (e.g. 18650 cells during a thermal runaway event have been measured to be
voltage, temperature) or physical damage to the DUT becomes evident. around 1.2 l (A h)−1 [87,136] for various cathode materials. Golubkov
et al. found higher amounts of vented gas on LCO/NMC batteries (e.g.
3.3.5. Fire test 2.3 l (A h)−1 as calculated from 0.27 mol of gas released) [10]. The gases
The objective of the fire test is to expose a battery or a vehicle to a being released are composed of a mixture of species: carbon monoxide,
fire and assess the risk of explosion. The source of the fire can be spilled carbon dioxide, methane, hydrogen, oxygen, ethane, ethylene, hydro-
fuel either from the vehicle itself or a nearby vehicle. This test it is often fluoric acid as measured in various studies [10,35,131,137,138]. The
termed Fuel fire test but can be also called: Radiant heat, Projectile exposure of persons in the vicinity of such compromised batteries can
fire, External fire simulation, Exposure to simulated vehicle fire, High- lead to serious injuries (e.g. eye irritation, chemical burns, poisoning,
temperature hazard or Fire resistance test. Table 15 displays test abrasion, skin injuries). Thus, it is of importance to identify and quantify
parameters. Three types of the test are described: substances being released from the battery during tests representing
misuse and abuse events and to ensure that the amounts released are not
i) Radiant-heat test: the battery (e.g. ≥ 80% SOC [72], 100% SOC hazardous to vehicle occupants and first aid responders. Within this
[61,65]) is placed inside a cylindrical metallic fixture, which is context, the development of warning sensors for passengers, first aid
externally heated by means of radiant heat (e.g. quartz lamps, tube responders and rescue workers has been advised [82]. For example, fire
furnace and conveyor mechanism). A temperature of 890 °C shall brigades include in their guidelines advice related to the chemical risks
be reached in less than 90 s and held for 10 min. Hazardous of batteries for EVs and HEVs (i.e. gas and liquid releases) such as: use
substance monitoring (e.g. EPA Methods TO-15 [128] and TO-17 of full PPE (personal protective equipment), avoid standing close to hot
[129]) is performed by sampling of combustion products to battery remnants and avoid inhaling the fumes under any circumstances
determine the possible presence of hazardous gas species released [139–141].
during the test [61,65].
ii) Projectile test: in this case the DUT, exposed to a uniform fire, is 3.4.1. Emissions related tests
surrounded by a steel wire mesh screen in a way that no part of an Some standards require hazardous substances measurements (e.g. gas,
exploding cell or battery can penetrate through the mesh (e.g. smoke, flames, and particulates) and for this analytical techniques or gas
0.25 mm diameter wire and grid density of 6–7 wires cm−1) sensors are recommended. Moreover, many standards require that the
(Fig. 4a) [63,66]. amounts measured need to be below certain concentrations
[61,63,65,66,72], such as those defined by the Emergency Response
UL 2580:2013 requires testing at least at 590 °C for a duration of Planning Guidelines ERPG-2 [142], from the American Industrial
20 min [63]. In this case, the use of a mesh screen is not mandatory Hygiene Association [143] or other industry practice documents or
and as alternative the DUT can be placed within a circular inner standards such as from the Occupational Safety and Health
perimeter area (e.g. < 1 m marked on the floor). No explosion of the Administration (OSHA) [144], Acute Exposure Guidelines Levels from
DUT that results in projectiles falling outside of this perimeter is the Environmental Protection Agency (EPA) [145], Short-Term Exposure
allowed. A second outer perimeter (around 1.5 m from inner peri- Limits (STEL) [146].
meter) made of a non-combustible material surrounds the inner SAE J2464:2009 [61] points out that the concentration of the
perimeter. released hazardous substances shall be scaled to the full pack for
quantitative comparison and scaled to a volume appropriate to human
i) Grating table configuration test: this test as described in UN/ECE- exposure in the vehicle (e.g. below ERPG-2 level: maximum airborne
R100.02:2013 [62] is an adaptation from UN/ECE-R34:2012 - concentration levels below which most individuals could be exposed for
Annex 5 [130], where a flame is created by burning fuel in a pan. up to one hour without experiencing or developing serious or irrever-
The DUT shall be placed on a grating table positioned above the pan sible health effects or symptoms which could impair an individual's
(Fig. 4b). The different steps of the test require first the preheating ability to take protective action).
of the DUT during 60 s by placing the burning fuel pan at a distance When manufacturers indicate the possibility that toxic gases can be
of 3 m. Then, the DUT is directly exposed to the flame for another released during abusive conditions, gas monitoring is needed during the
70 s. Immediately after, a screen of refractory material is placed in tests by utilizing one of the following techniques (or equivalent) as
between the pan and the DUT in order to reduce the flame for described in UL 2580:2013 [63] and SAE J2464:2009 [61]:
further 60 s as depicted in Fig. 4b. This test is passed if there is no
evidence of explosion during the test. • ASTM (the American Society for Testing and Materials) D4490:
standard practice for measuring concentrations of toxic gases of
Only two standards, SAE J2464:2009 [61] and UL 2580:2013 [63] vapours using detector tubes [147].
highlight the importance of quantifying toxic and determining flam- • ASTM D4599: standard practice for measuring concentrations of toxic
mable emissions providing suitable testing procedures (see Section 3.4 gases of vapours using length-of-stain dosimeters [148].
for further details). Although it has been proven by various authors that • OSHA: Evaluation guidelines for air sampling methods utilizing
significantly higher amounts of, for example HF, are generated in EV spectroscopic analysis [149].
fires compared to ICE vehicle fires (e.g. 1500 g compared to 600– • NIOSH (The National Institute for Occupational Safety and Health):
800 g, respectively [131–133]), the implementation of analysis of Manual for analytical methods [150].
emissions is not widely adopted. Moreover, with such a variation of • EPA Methods TO-15 [128] for the determination of VOC's (volatile
conditions and requirements for the fire test, it seems clear that the organic compounds) in air analysed by Gas Chromatography and Mass
comparability of test results is not ensured. spectrometry.
1448
• EPA Methods TO-17 [129] for the determination of VOC's in air published in 2015 under GB/T 31467.1, GB/T 31467.2 and GB/T
using active sampling onto sorbent tubes. 31467.3. Within IEC, IEC62660-2 [70] was published in 2011,
describing safety tests for propulsion cells. IEC 62660-3:2016 [71],
More sophisticated devices for gas detection of evaporated com- defining cell safety specifications was published in 2016. Also a new
pounds can be Fourier Transformed Infrared Spectroscopy (FTIR) and standard, IEC 62485-6 [151], on safety requirements for lithium-ion
mobile detection systems (e.g. detection of O2, CO, H2, C2H4O, HF and batteries and battery installations is proposed.
of toxic VOC's as used by German fire brigades [82]). On vehicle safety, ISO 6469-4 [105] on post-crash safety has been
Standard SAE J6469-1:2009 [103] requires that potentially dan- published in 2015.
gerous concentration of hazardous gases or other hazardous substances Once published, standards go into a maintenance cycle with
shall not be allowed anywhere in the driver, passenger and load periodic revision at least every five years. Topics for revision may
compartments. The maximum allowable quantity accumulated during include consideration of upcoming battery technologies such as lithium
testing of hazardous gases and other substances (for normal operating sulfur, lithium air as well as lithium ion capacitors for which specific
and environmental conditions) shall refer to the latest version of test procedures may be required. Furthermore, the use of batteries in a
applicable National/International standards or regulations. “second life” application will require specific test regimes to determine
UN/ECE-R100.02:2013 [62] regulates emissions from open-type their state of health and their cycling in stationary applications, taking
traction batteries, which may produce hydrogen gas during normal into account the specific safety requirements of the operating environ-
operation. The quantification of hydrogen during normal charging ment.
follows the protocol indicated in the regulation and must remain below A Global Technical Regulation on Electric Vehicle Safety (GTR-
certain limits (i.e. below 25 x h (g)). Other gases are not considered. EVS) has been submitted for a vote to the UNECE World Forum for
Systems with a closed chemical process, such as LIBs, are considered Harmonization of Vehicle Regulations (WP.29). A decision on the
'emission free' (i.e. do not emit gases under normal operation). In the adoption of this regulation is expected in November 2017.
case of abusive conditions, this regulation does not enforce any
requirements or limitations for emissions of hazardous gases (e.g. 5. Conclusions
venting) from any type of rechargeable energy storage systems. An
improvement of the regulation in this regard could be of high This work presents a comprehensive review of the various stan-
importance to ensure the safety of users and first aid responders. dards and regulations dealing with the safety performance of lithium
ion batteries to be used in electrified transport. Test parameters and
3.4.2. Flammability tests conditions adopted in the test methods which are described in these
In abusive conditions, it is possible that LIBs emit flammable gases standards and regulations are compared. From the analysis performed
(e.g. methane, ethane, hydrogen, carbon monoxide). SAE J2929:2013 the following conclusions are drawn:
[66], for example, highlights this hazard and recommends that con-
sideration should be given to preventing the build-up of flammable • Most of the existing standards and regulations impose test require-
gases that could get in contact with vehicle ignition sources (e.g. sparks ments derived from regulatory documents originally intended for
from a short circuit, fire in the vicinity). Determination of the conventional vehicles. It is clear that more analysis and data
flammability of any substance (e.g. liquid, solid materials) emitted evaluation specific to EVs and HEVs is highly desirable to cover
from the battery is mandatory in many standards [61,63,65,66,72]. the specificities of electrified technologies. For example, recent
One method is to incorporate one or several spark ignition source(s) in research has indicated that battery pack installations may be
the testing area, located close to the DUT. Alternatively, gas monitors exposed to vibration loads outside the range evaluated by existing
can also be used, as mentioned in UL 2580:2013 [63]. On the contrary, standards [100].
some other standards do not give indications on how to assess this • Another concern is whether the tests performed at component level
property, e.g. UN/ECE-R100.02:2013 [62]. are comparable to those carried out at vehicle level. For example, the
Overall it can be concluded that the evaluation of chemical hazards force experienced by a DUT in a crush test (component level) or
is tackled very differently in the various standards and regulations. In crash test (vehicle level) is expected to depend on the presence and -
some cases, such as in SAE J2464:2009 [61], SAE J2929:2013 [66] and if present, on the properties of – mechanical protection (such as the
UL 2580:2013 [63], detailed information on quantifying and determin- chassis or battery enclosure).
ing toxicity and flammability of LIB emissions is provided, while in • Comparability of component testing at cell, module and pack level
other cases this issue is only slightly mentioned, such as in UN/ECE- should also be examined. For example, it has been proven that the
R100.02:2013 [62]. In some instances chemical hazards are not even initial current created in the short circuit test is influenced by the
considered, such as in ISO 12405-1(2,3):2001(2012,2014) [67–69], size of the DUT [112,113] as well as by its type of connection (i.e.
IEC 62660-2(3):2001(2013) [70,71], KMVSS 18-3:2009 [73], AIS- parallel or serial). Similar influence on test outcome may be
048:2009 [74] and QC/T 743:2006 [75]. Taking into consideration the expected when applying a single crush force and crushing plate to
importance of the issue, it would be advisable that future standardisa- DUTs of different sizes.
tion/regulation developments consider a harmonized testing guidance • Dispersion in test conditions (e.g. SOC, temperature) is rather
or protocol to ensure that chemical hazards of automotive batteries are wide for most tests (e.g. overcharge, thermal shock, short circuit).
appropriately assessed. This has an important impact in the comparability of data
obtained utilizing various standards, while in some cases differ-
4. Current evolutions and future perspectives ences in parameters might be due to different considered scenar-
ios. Alignment of parameters is advisable in order to perform fair
International standards on lithium traction battery safety are being and equivalent tests. As the worst case typically corresponds to
developed by ISO and IEC, focusing respectively on system and cell maximum SOC, it is logical that abuse testing is performed in
level. Documents already published by ISO include ISO 12405-1:2011 such condition. For example, in the short circuit test, the higher
and 12405-2:2012 [67,68], defining test specifications for high-power the SOC value of the DUT, the higher the short circuit current
(for hybrids) and high-energy batteries (for battery electric vehicles), generated [112] and the lower the onset temperature of thermal
respectively. Both these documents were complemented with the runaway [115]. Most standards already require 100% SOC,
recently published ISO 12405-3:2014 [69], which sets pass/fail however regulation UN/ECE-R100.02:2013 [62] allows testing
requirements to the precedent documents. Chinese counterparts were at ≥ 50% SOC.
1449
• Real world accidents are dynamic events i.e. the battery moves [8] American and European standards organisations agree to strengthen transatlantic
cooperation on standards for electric vehicles. The European Committee for
towards the impact zone. However, testing at component level is Standardization (CEN) and the European Committee for Electrotechnical
carried out using static assemblies where the impactor moves Standardization (CENELEC); 2012. Available Online: 〈http://www.cencenelec.eu/
towards the battery. Investigations have shown discrepancies in news/press_releases/pages/PR-2012-18.aspx〉.
[9] Opitz A, Badami P, Shen L, Vignarooban K, Kannan AM. Can Li-ion batteries be
mechanical loads between current standards and regulations and the panacea for automotive applications?. Renew Sustain Energy Rev
dynamic crash tests [94,96], and for this, appropriate modifications 2017;68(Part1):685–92.
within the regulatory framework are advisable. [10] Golubkov AW, Fuchs D, Wagner J, Wiltsche H, Stangl C, Fauler G, et al. Thermal-
• Systems intended to be removed from the vehicle for charging, runaway experiments on consumer Li-ion batteries with metal-oxide and olivin-
type cathodes. RSC Adv 2014;4:3633–42.
swapping or replacement may be accidentally dropped during [11] Jacoby M. Burning batteries, chemical & engineering news. C & EN Chem Eng
handling. Although this scenario seems plausible, the drop test is News 2007;85:26–8.
[12] Lamb J, Orendorff CJ, Steele LAM, Spangler SW. Failure propagation in multi-cell
not included in various automotive battery regulations and stan-
lithium ion batteries. J Power Sources 2015;283:517–23.
dards, such as UN/ECE-R100.02:2013 [62], ISO 12405- [13] von Sacken U, Nodwell E, Sundler A, Dahn JR. Comparative thermal stability of
1(2,3):2011(2012,2014) [67–69] and ISO 62660-2:2011 [70]. carbon intercalation anodes and lithium metal anodes for rechargeable lithium
• The occurrence of internal short circuits is one of the main concerns batteries. Solid State Ion 1994;69:284–90.
[14] Chen Y, Evans JW. Thermal analysis of lithium-ion batteries. J Electrochem Soc
for battery manufacturers; however these tests have not been widely 1996;143:2708–12.
implemented in the legislative landscape. It is recognized, though, [15] Bandhauer TM, Garimella S, Fuller TF. Temperature-dependent electrochemical
that the practicalities of this test are complex and implementation of heat generation in a commercial lithium-ion battery. J Power Sources
2014;247:618–28.
such testing would require significant research for test method [16] Levy SC, Bro P. Battery hazards and accident prevention. New York and London:
development. Plenum Press; 1994.
• Only SAE J2464:2009 [61], SAE J2929:2013 [66] and UL [17] Mikolajczak MK C, White K, Long RT. Lithium-ion batteries hazard and use
assessment. Menlo Park, CA: Exponent Failure Analysis Associates, Inc, Springer;
2580:2013 [63] highlight the importance of determining toxic and 2011 .
flammable emissions and provide suitable testing procedures. The [18] Boeing 787 battery fire. National Transportation Safety Board (NTSB) Accident
implementation of specific analysis is not widely adopted by other Investigations; 2013. 〈http://www.ntsb.gov/investigations/pages/boeing_787.
aspx〉.
bodies. Taking into consideration the importance of this issue, it is
[19] Hazardous materials accident reports. National Transportation Safety Board.
advisable that future standardisation/regulation developments con- 〈http://www.ntsb.gov/investigations/AccidentReports/Pages/hazardous.aspx〉.
sider a harmonized testing guidance and protocols to ensure that [20] Lithium battery fire risk linked to Dubai plane crash. The National; 2010. 〈http://
www.thenational.ae/news/uae-news/lithium-battery-fire-risk-linked-to-dubai-
chemical hazards of automotive batteries are appropriately assessed
plane-crash〉.
in order to ensure the safety of vehicle occupants and surrounding [21] Air accident investigation report: uncontained cargo fire leading to loss of control
persons. inflight and uncontrolled descent into terrain. General Civil Aviation Authority of
• In relation to safety testing, the evaluation of realistic scenarios is the United Arab Emirates; 2010. 〈https://www.gcaa.gov.ae/en/ePublication/
admin/iradmin/Lists/Incidents%20Investigation%20Reports/Attachments/40/
greatly recommended in order to ensure a safe future for the use of 2010-2010%20-%20Final%20Report%20-%20Boeing%20747-44AF%20-
lithium ion battery technologies. To ensure this, the addition of %20N571UP%20-%20Report%2013%202010.pdf〉.
some tests, such as roll over, drop, immersion, low temperature [22] Smith B. Chevrolet volt battery incident overview report, 2012th ed.. Washington,
DC: National Highway Traffic Safety Administration (NHTSA); 2012.〈http://
hazards, toxicity, flammability, etc. into future standards and www.nhtsa.gov/staticfiles/nvs/pdf/Final_Reports.pdf〉.
regulations should be considered. [23] On-site electric vehicle fire investigation. US Department of Transportation –
• Finally, clear and unambiguous testing guidelines should be pro- National Highway Traffic Safety Administration; 2013.
[24] Apple says product shortage will hurt quarterly earnings: computers: company
vided as part of the test method and rationale description. Examples unable to keep up with consumer demand. It also is recalling new PC because of
include descriptions of the method for setting the SOC, the location fire concern. Times, Los Angeles; 1995. 〈http://articles.latimes.com/1995-09-15/
of temperature sensors, the exact position of the DUT in the various business/fi-46242_1_many-apple-computers〉.
[25] CPSC and EV Global Motors Co . Announce recall of lithium batteries in electric
tests, in addition to the minimum tolerance required for the testing bikes. United States: Consumer Product Safety Commission; 2002. 〈https://www.
equipment. Such guidelines facilitate the correct and harmonized cpsc.gov/Recalls/2002/CPSC-and-EV-Global-Motors-Co-Announce-Recall-of-
interpretation of the standard or regulation by the testing bodies and Lithium-Batteries-in-Electric-Bikes-/〉
[26] Ogumi Z, Wang H. Chapter 3: carbon anode materials. In: Yoshio Masaki, Brodd
comparability of results would be improved.
Ralph J, Kozawa A, editors. Lithium-ion batteries science and technologies. New
York, NY: Springer; 2009. p. 51.
[27] Zhang SS. A review on electrolyte additives for lithium-ion batteries. J Power
Acknowledgements
Sources 2006;162:1379–94.
[28] Illig J, Chrobak T, Ender M, Schmidt JP, Klotz D, Ivers-Tiffée E. Studies on
Authors thank Marc Steen for his excellent support reviewing this LiFePO4 as cathode material in Li-ion batteries. In: Proceedings of the batteries
manuscript. and energy technology joint general session – 217th ECS meeting, 30 ed.
Vancouver, BC; 2010, p. 3-17.
[29] Saw LH, Ye Y, Tay AAO. Electrochemical–thermal analysis of 18650 lithium iron
References phosphate cell. Energy Convers Manag 2013;75:162–74.
[30] Nitta N, Wu F, Lee JT, Yushin G. Li-ion battery materials: present and future.
Mater Today 2015;18:252–64.
[1] Electric Vehicle Market Forecasts . Global forecasts for light duty hybrid, plug-in [31] Balakrishnan PG, Ramesh R, Prem Kumar T. Safety mechanisms in lithium-ion
hybrid, and battery electric vehicle sales and vehicles in use: 2014–2023. Nav Res batteries. J Power Sources 2006;155:401–14.
2014. [32] Zhu G-N, Du Y-J, Wang Y-G, Yu A-S, Xia Y-Y. Electrochemical profile of lithium
[2] Proff H, Kilian D. Competitiveness of the EU automotive industry in electric titanate/hard carbon composite as anode material for Li-ion batteries. J
vehicles. Duisburg: University of Duisburg-Essen; 2012. Electroanal Chem 2013;688:86–92.
[3] Lutsey N. Transition to a global zero-emission vehicle fleet: a collaborative agenda [33] Nowack LV, Waser O, Yarema O, Wood V. Rapid, microwave-assisted synthesis of
for goverments, international council on clean transportation transition to a global battery-grade lithium titanate (LTO). RSC Adv 2013;3:15618–21.
zero-emission vehicle fleet: a collaborative agenda for goverments. Int Counc [34] Reddy T. Linden's handbook of batteries, 4th edition. New York, NY: McGraw-
Clean Transp 2015. Hill, Inc.; 2011.
[4] Lithium ion rechargeable batteries technical handbook. Sony Corporation. [35] Campion CL, Li W, Lucht BL. Thermal decomposition of LiPF 6-based electrolytes
〈https://www.4project.co.il/documents/doc_286_2661.pdf〉. for lithium-ion batteries. J Electrochem Soc 2005;152:A2327–A2334.
[5] Lowe M, Tokuoka S, Trigg T, Gereffi G. Lithium-ion batteries for electric vehicles: [36] Lebedeva N, Boon-Brett L. Considerations on the chemical toxicity of contem-
the US value chain. Centre on Globalization, Governance & Competitiveness Duke porary Li-ion battery electrolytes and their components. J Electrochem Soc
University; 2010. 2015;163:A821–A830.
[6] Cluzel C, Douglas C. Cost and performance of EV batteries. final report. Elem [37] Ue M. Fire-resistant electrolyte for lithium battery JP1992–184870A;
Energy Ltd Comm Clim Change 2012. JP3274102B [in Japanese]. Japan: Mitsubishi Petrochem; 1992.
[7] Lache R, Galves D, Nolan P, et al. Autos & auto parts electric cars: plugged in 2. A [38] Ue M, Sasaki Y, Tanaka Y, Morita M. Chapter 2: nonaqueous electrolytes with
mega-theme gains momentum. FITT Res 2009. advances in solvents. Electrolytes for lithium and lithium-ion batteries. In: Jow
1450
TR, Xu K, Borodin O, Ue M, editors. Modern aspects of electrochemistry. [73] KMVSS Article18-3. Traction battery; 2009.
Springer; 2014. [74] AIS-048. Battery operated vehicles – safety requirements of traction batteries;
[39] Lee CW, Venkatachalapathy R, Prakash J. A novel flame retardant additive for 2009.
lithium batteries electrochemical and solid-state letters, Vol. 3; 2000, p. 63–5. [75] QC/T 743. Lithium-ion batteries for electric vehicles Chinese voluntary standards
[40] Damen L, Lazzari M, Mastragostino M. Safe lithium-ion battery with ionic liquid- for automobiles; 2006.
based electrolyte for hybrid electric vehicles. J Power Sources 2011;196:8692–5. [76] UN 38.3; Recommendations on the transport of dangerous goods manual of test
[41] Balducci A, Jeong SS, Kim GT, Passerini S, Winter M, Schmuck M, et al. and criteria, 6th revised edition; 2015.
Development of safe, green and high performance ionic liquids-based batteries [77] ISO 16750-3. Road vehicles – environmental conditions and testing for electrical
(ILLIBATT project). J Power Sources 2011;196:9719–30. and electronic equipment Part 3: mechanical loads; 2003.
[42] Kerner M, Lim D-H, Jeschke S, Rydholm T, Ahn J-H, Scheers J. Towards more [78] UN/ECE Regulation No. 12. Uniform provisions concerning the approval of
thermally stable Li-ion battery electrolytes with salts and solvents sharing nitrile vehicles with regard to the protection of the driver against the steering mechanism
functionality. J Power Sources 2016;332:204–12. in the event of impact; 2012.
[43] Paulechka YU, Kabo GJ, Blokhin AV, Vydrov OA, Magee JW, Frenkel M. [79] UN/ECE Regulation No. 94. Uniform provisions concerning the approval of
Thermodynamic properties of 1-butyl-3-methylimidazolium hexafluorophosphate vehicles with regard to the protection of the occupants in the event of a frontal
in the ideal gas state†. J Chem Eng Data 2003;48:457–62. collision; 2012.
[44] Lemmon EW, McLinden MO, Friend DG . Thermophysical properties of fluid [80] UN/ECE Regulation No. 95. Uniform provisions concerning the approval of
systems. In: Linstrom PJ, Mallard WG, editors. NIST chemistry WebBook, NIST vehicles with regard to the protection of the occupants in the event of a lateral
standard reference database number 69. 〈http://webbooknistgov〉. collision; 2011.
[45] Ohno H. Chapter 1: importance and possibility of ionic liquids. Electrochemical [81] FMVSS 305. Electric powered vehicles: electrolyte spillage and electrical shock
aspects of ionic liquids. New Jersey: John Wiley & Sons, Inc., Publication; 2005. protection; 2011.
p. 1–4. [82] Wisch M, Ott J, Thomson R, Léost Y, Abert M, Yao J. EVERSAFE.
[46] Daigle JC, Vijh A, Hovington P, Gagnon C, Hamel-Pâquet J, Verreault S, et al. Recommendations and guidelines for battery crash safety and post-crash safe
Lithium battery with solid polymer electrolyte based on comb-like copolymers. J handling. In: Proceedings of the 7th framework programme era-net transport
Power Sources 2015;279:372–83. electromobility+; 2014.
[47] Janek J, Zeier WG. A solid future for battery development. Nat Energy [83] Adolph T, Edward M, Thomson R, Stein M, Lemmen P, V N, et al. FIMCAR full
2016;1:16141. width test procedures: updated protocol in FIMCAR-frontal impact and compat-
[48] Yamamoto H, Mori H. Chapter 7: SBR binder (for negative electrode) and ACM ibility assessment research , Bd VIII. Berlin: Universitatverslag der TU Berlin;
binder (for positive electrode). In: Yoshio Masaki, Brodd Ralph J, Kozawa A, 2015. p. 1–85.
editors. Lithium-ion batteries science and technologies. New York: Springer; 2009. [84] Dubaniewicz TH, DuCarme JP. Are lithium ion cells intrinsically safe? In:
[49] Zhang Z, Ramadass P. Chapter 20: lithium-ion battery separators. In: Yoshio Proceedings of the industry applications society annual meeting (IAS). IEEE;
Masaki, Brodd Ralph J, Kozawa A, editors. Lithium-ion batteries science and 2012, p. 1–10.
technologies. New York: Springer; 2009. [85] Maleki H, Howard JN. Internal short circuit in Li-ion cells. J Power Sources
[50] Yu WC, Hux SE. Method of making a trilayer battery separator. US Patent, 5,952, 2009;191:568–74.
120; 1999. [86] Cai W, Wang H, Maleki H, Howard J, Lara-Curzio E. Experimental simulation of
[51] Hying C. Separation separators for lithium batteries – safety & performance. internal short circuit in Li-ion and Li-ion-polymer cells. J Power Sources
Batteries. Paris; 2004. 2011;196:7779–83.
[52] Pfrang A, Kriston A, Ruiz V, Lebedeva N, di Persio F. Chapter eight – safety of [87] Dubaniewicz TH, Jr., DuCarme JP. Further study of the intrinsic safety of
rechargeable energy storage systems with a focus on Li-ion Technology A2 – internally shorted lithium and lithium-ion cells within methane-air. J Loss Prev
Rodriguez-Martinez, Lide M. In: Omar N, editor. Emerging nanotechnologies in Process Ind 2014;32:165–73.
rechargeable energy storage systems. Boston: Elsevier; 2017. p. 253–90. [88] Spek EJ. Lithium ion abuse test methods improvement IEEE symposium on
[53] Kumar MS, Revankar ST. Development scheme and key technology of an electric product compliance engineering (IEEE PSES). Portland, Oregon; 2012.
vehicle: an overview. Renew Sustain Energy Rev 2017;70:1266–85. [89] Bierens JJLM. Handbook on drowning: prevention, rescue, treatment; 2006.
[54] Arora S, Shen W, Kapoor A. Review of mechanical design and strategic placement [90] Florida leads the nation in car crash drownings; 2014. 〈http://touch.
technique of a robust battery pack for electric vehicles. Renew Sustain Energy Rev orlandosentinel.com/#section/-1/article/p2p-81915801〉
2016;60:1319–31. [91] Hurricane sandy destroyed up to 250,000 cars; 2012. 〈http://www.huffingtonpost.
[55] Yeow K, Teng H, Thelliez M, Tan E. 3D thermal analysis of Li-ion battery cells with com/2012/11/07/hurricane-sandy-cars_n_2090916.html〉
various geometries and cooling conditions using abaqus. In: Proceedings of the [92] Fisker loses over 300 karmas in sandy floods at Port Newark; 2012. 〈http://www.
SIMULIA community conference; 2012, p. 1–16. autoblog.com/2012/11/07/fisker-loses-over-300-karmas-in-sandy-floods-at-
[56] Mikolajczak C, Kahn M, White K, Long RT. Lithium-ion batteries hazard and use port-newark/〉
assessment. In: Exponent failure analysis associates IFPRF, editor. Exponent [93] Brecher A. Chapter 9: transit bus applications of lithium-ion batteries. In: Pistoia
Failure Analysis Associates, Inc., Fire Protection Research Foundation; 2011. G, editor. Lithium-ion batteries: advances and applications; 2014, p. 194.
[57] Lamb J, Orendorff CJ. Evaluation of mechanical abuse techniques in lithium ion [94] Wech L, Richter R, Justen R, Schöneburg R. Crash safety aspects of HV batteries
batteries. J Power Sources 2014;247:189–96. for vehicles. In: Proceedings of the international technical conference on the
[58] SAE J2289. Electric-drive battery pack system: functional guidelines; 2008. enhanced safety of vehicles ESV-22. Washington, USA; 2011.
[59] Lisbona D, Snee T. A review of hazards associated with primary lithium and [95] Bakker J, Sachs J, Otte D, Justen R, Hannawald L, Friesen F. Analysis of fuel cell
lithium-ion batteries. Process Saf Environ Prot 2011;89:434–42. vehicles equipped with compressed hydrogen storage systems from a road
[60] Spotnitz R, Franklin J. Abuse behavior of high-power, lithium-ion cells. J Power accident safety. Perspect SAE Int J Passeng Cars – Mech Syst 2011;4:332–42.
Sources 2003;113:81–100. [96] Justen R, Schöneburg R. Crash SAfety of Hybrid and Battery Electric Vehicles. In:
[61] SAE J2464. Electric and hybrid electric vehicle rechargeable energy storage Proceedings of the international technical conference on the enhanced safety of
system (RESS) safety and abuse testing; 2009. vehicles ESV-22. Washington, USA; 2011.
[62] UN/ECE Regulation No. 100.02. Uniform provisions concerning the approval of [97] El-Hennawy HM, El-Menyar A, Al-Thani H, Tuma M, Parchani A, Abdulrahman
vehicles with regard to specific requirements for the electric power train; 2013. H, et al. Epidemiology, causes and prevention of car rollover crashes with ejection.
[63] UL 2580. Batteries for use in electric vehicles; 2013. Ann Med Health Sci Res 2014;4:495–502.
[64] Ashtiani CN. Analysis of battery safety and hazards' risk mitigation. 19th ed. [98] Baren JV. What is random vibration testing?. Sound Vib Instrum Ref Issue
Washington, DC; 2008, p. 1–11. 2012:9–12.
[65] Doughty DH, Crafts CC. SAND 2005–3123: freeDomCAR electrical energy storage [99] Jayahari L. Correlation of sinusoidal sweep test to field random
systems abuse test manual for electric and hybrid electric vehicle applications; vibrations [Master's degree thesis]. Karlskrona, Sweden: Blekinge Institute of
2005. Technology; 2005.
[66] SAE J2929. Safety standards for electric and hybrid vehicle propulsion battery [100] Hooper JM, Marco J. Characterising the in-vehicle vibration inputs to the high
systems utilizing lithium-based rechargeable cells; 2013. voltage battery of an electric vehicle. J Power Sources 2014;245:510–9.
[67] ISO 12405-1. Electrically propelled road vehicles – test specification for lithium- [101] IEC 60068-2-64. Environmental testing Part 2: test methods – test Fh: vibration,
ion traction battery packs and systems Part 1: high-power applications; 2011. broad-band random (digital control) and guidance; 1993.
[68] ISO 12405-2. Electrically propelled road vehicles – test specification for lithium- [102] SAE J2380. Vibration testing of electric vehicle batteries; 2009.
ion traction battery packs and systems Part 2: high-energy applications; 2012. [103] ISO 6469-1. Electrically propelled road vehicles – safety specifications Part 1: on-
[69] ISO 12405-3. Electrically propelled road vehicles – test specification for lithium- board rechargeable energy storage system (RESS); 2009.
ion traction battery packs and systems Part 3: Safety performance requirements; [104] ISO 6469-3. Electrically propelled road vehicles – safety specifications – Part 3:
2014. protection of persons against electric shock; 2011.
[70] IEC 62660-2. Rechargeable cells standards publication secondary lithium-ion cells [105] ISO 6469-4. Electrically propelled road vehicles – safety specifications Part 4: post
for the propulsion of electric road vehicles Part 2: reliability and abuse testing; crash electrical safety requirements; 2015.
2011. [106] ISO 17409. Electrically propelled road vehicles – connection to an external
[71] IEC 62660-3. Rechargeable cells standards publication secondary lithiumion cells electric power supply – safety requirements; 2015.
for the propulsion of electric road vehicles Part 3: safety requirements of cells and [107] IEC 61851. Electric vehicle conductive charging system; 2010.
modules; 2016. [108] ISO PAS 19363. Electrically propelled road vehicles – magnetic field wireless
[72] Unkelhaeuser T, Smallwood D. SAND99-0497-USABC: United States advanced power transfer – safety and interoperability requirements.
battery consortium electrochemical storage system abuse test procedure manual; [109] Wang Q, Ping P, Zhao X, Chu G, Sun J, Chen C. Thermal runaway caused fire and
1999. explosion of lithium ion battery. J Power Sources 2012;208:210–24.
1451
[110] IEC 60898-1. Electrical accessories – circuit-breakers for overcurrent protection for Renewable Energy Laboratory (NREL); 1995.
household and similar installations Part 1: circuit-breakers for ac operation; 2015. [135] Lebedeva N, Ruiz V, di Persio F, Kriston A, Pfrang A, Kosmidou T. et al.
[111] Okazaki S, Higuchi S, Kubota N, Takahashi S. Measurement of short circuit Evaluation of volume of free electrolyte in various cell types – method and
current for low internal resistance batteries. J Appl Electrochem 1986;16:513–6. preliminary results. EVS-07-24e – (JRC) free electrolyte. Working Party on
[112] Conte FV, Gollob P, Lacher H. Safety in the battery design: the short circuit. World Passive Safety (GRSP) – Electric Vehicle Safety (EVS) 7th session in 〈https://
Electr Veh J 2009:3. www2.unece.org/wiki/display/trans/EVS+7th+session〉; 2015.
[113] Okazaki S, Higuchi S, Kubota N, Takahashi S. Predicted and observed initial short [136] Roth EP, Orendorff CJ. How electrolytes influence battery safety. Electrochem Soc
circuit current for lead-acid batteries. J Appl Electrochem 1986;16:631–5. Interface 2012;21:45–9.
[114] UL 1642. UL standard for safety for lithium batteries; 2007. [137] Roth EP, Crafts CC, Doughty DH, McBreen J. Thermal abuse performance of
[115] Mendoza-Hernandez OS, Ishikawa H, Nishikawa Y, Maruyama Y, Umeda M. 18650 Li-ion cells; 2004.
Cathode material comparison of thermal runaway behavior of Li-ion cells at [138] Ohsaki T, Kishi T, Kuboki T, Takami N, Shimura N, Sato Y, et al. Overcharge
different state of charges including over charge. J Power Sources reaction of lithium-ion batteries. J Power Sources 2005;146:97–100.
2015;280:499–504. [139] Wargclou D. Extrication from cars during road traffic accidents. Swedish Civil
[116] Wu A, Tabaddor M, Wang C, Jeevarajan J. Simulation of internal short circuits in Contingencies Agency (MSB); 2011.
lithium-ion cells. Corp Res UL LLC 2014. [140] Casey C, Grant PE. Fire fighter safety and emergency response for electric drive
[117] Zhao W, Luo G, Wang C-Y. Modeling internal shorting process in large-format Li- and hybrid electric vehicles. Fire Prot Res Found 2010.
Ion cells. J Electrochem Soc 2015;162:A1352–A1364. [141] Accident assistance and recovery of vehicles with high-voltage systems. German
[118] Orendorff CJ, Roth EP, Nagasubramanian G. Experimental triggers for internal Association of the Automotive Industry (VDA); 2013.
short circuits in lithium-ion cells. J Power Sources 2011;196:6554–8. [142] Emergency Response Planning Guidelines: ERPG level 2. 〈https://www.aiha.org/
[119] IEC 62133-2. Secondary cells and batteries containing alkaline or other non-acid get-involved/AIHAGuidelineFoundation/
electrolytes – safety requirements for portable sealed secondary cells, and for EmergencyResponsePlanningGuidelines/Documents/2014%20ERPG
batteries made from them, for use in portable applications - Part 2: Lithium %20Introduction.pdf〉.
systems; 2017. [143] American industrial hygiene association. 〈https://www.aiha.org/Pages/default.
[120] Jones HP, Chapin JT, Tabaddor M. Critical review of commercial secondary aspx〉.
lithium-ion battery safety standards. European Space Agency, (Special [144] Occupational safety and health administration (OSHA). 〈https://www.osha.gov/〉.
Publication). In: Proceedings of the ESA SP 4th IAASS conference 'making safety [145] Acute exposure guidelines levels from the environmental protection agency (EPA).
matter'; 2010. 〈http://www.epa.gov/oppt/aegl/pubs/chemlist.htm〉.
[121] Feng X, Fang M, He X, Ouyang M, Lu L, Wang H, et al. Thermal runaway features [146] Short-term exposure limits (STEL). 〈http://www.acgih.org/store/〉.
of large format prismatic lithium ion battery using extended volume accelerating [147] ASTM D4490. Standard practice for measuring concentrations of toxic gases of
rate calorimetry. J Power Sources 2014;255:294–301. vapours using detector tubes; 2011.
[122] Lu W, López CM, Liu N, Vaughey JT, Jansen A, DW D. Overcharge effect on [148] ASTM D4599. Standard practice for measuring concentrations of toxic gases of
morphology and structure of carbon electrodes for lithium-ion batteries. J vapours using length-of-stain dosimeters; 2014.
Electrochem Soc 2012;159:A566–A570. [149] OSHA. Evaluation guidelines for air sampling methods utilizing spectroscopic
[123] IEC 61434. Secondary cells and batteries containing alkaline or other non-acid analysis. 〈https://www.osha.gov/dts/sltc/methods/spectroguide/spectroguide.pdf〉.
electrolytes – guide to designation of current in alkaline secondary cell and battery [150] NIOSH. Manual for analytical methods. 〈http://www.cdc.gov/niosh/docs/2003-
standards; 1996. 154/〉.
[124] Tobishima S, Sakurai Y, Yamaki J. Safety characteristics of rechargeable lithium [151] IEC 62485-6. Safety requirements for secondary batteries and battery installa-
metal cells. J Power Sources 1997;68:455–8. tions Part 6: lithium-ion batteries for traction applications PNW.
[125] Xu F, He H, Liu Y, Dun C, Ren Y, Liu Q, et al. Failure investigation of LiFePO4
cells under overcharge conditions. J Electrochem Soc 2012;159:A678–A687.
[126] He H, Liu Y, Liu Q, Li Z, Xu F, Dun C, et al. Failure investigation of LiFePO4 cells Glossary
in over-discharge conditions. J Electrochem Soc 2013;160:A793–A804.
[127] Zhang SS, Xu K, Jow TR. Study of the charging process of a LiCoO2-based Li-ion ASTM: The American Society for Testing and Materials
battery. J Power Sources 2006;160:1349–54. CC: Constant Current
[128] U.S. Environmental Protection Agency (EPA) method TO-15. Determination of CV: Constant Voltage
volatile organic compounds (VOCs) in air collected in specially-prepared canisters ERPG: Emergency Response Planning Guidelines
and analyzed by gas chromatography/mass spectrometry (GC/MS); 1999. http:// EPA: Environmental Protection Agency
www.epa.gov/ttnamti1/files/ambient/airtox/to-15r.pdf. EV: Electric Vehicle
[129] U.S. Environmental Protection Agency (EPA) method TO-17. Determination of DUT: Device Under Test
volatile organic compounds in ambient air using active sampling onto sorbent BEV: Battery Electric Vehicle
tubes; 1999. https://www3.epa.gov/ttnamti1/files/ambient/airtox/to-17r.pdf. FMVSS: Federal Motor Vehicle Safety Standards
[130] UN/ECE Regulation No. 34. Uniform provisions concerning the approval of HP: High Power
vehicles with regard to the protection of fire risks; 2012. HE: High Energy
[131] Lecocq A, Bertana M, Truchot B, Marlair G. Comparison of the fire consequences HEV: Hybrid Electric Vehicle
of an electric vehicle and an internal combustion engine vehicle. In: NIOSH: The National Institute for Occupational Safety and Health
Proceedings of the international conference on fires in vehicles – FIVE. Chicago, NLF: New Legislative Framework
United States; 2012. p. 183–94. LIB: Lithium Ion Battery
[132] Andersson P, Blomqvist P, Lorén A, Larsson F. Investigation of fire emissions OSHA: Occupational Safety and Health Administration
from Li-ion batteries. SP technical research institute of Sweden; 2013. PSD: Power Spectral Density
[133] Sturk D, Hoffmann L, AT A. Fire tests on e-vehicle battery cells and packs. Traffic STEL: Short-Term Exposure Limits
Inj Prev 2015;16:159–64. SOC: State of Charge
[134] Vimmerstedt LJ, Ring S, Hammel CJ. Current status of environmental, health and VOC's: Volatile Organic Compounds
safety issues of lithium ion electric vehicle batteries. Golden, CO: National
1452

1 s2.0 S1364032117308250 Main

Uploaded by

Copyright:

Available Formats

1 s2.0 S1364032117308250 Main

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

1 s2.0 S1364032117308250 Main

Uploaded by

Copyright:

Available Formats

Renewable and Sustainable Energy Reviews 81 (2018) 1427–1452

Contents lists available at ScienceDirect

Renewable and Sustainable Energy Reviews

A review of international abuse testing standards and regulations MARK

regulations issued by the United Nations Economic Commission for Table 1

Region of applicability International EU and USA Korea India China

Mechanical Mechanical 3.1.a CMP CMPV P P C CMPV C Me P MP MP M

Environmental Thermal stability 3.3.a C C C CMP CMP P CP

Chemical Emissions 3.4.a CMP P CMP CMP CMP

Region of International EU and further USA Korea India China

Level (C, M, P) CMP CMP P P C CMP C M+ P MP M

Cylindrical steel object (radius

similar object. Alternatively, it is also possible to perform a horizontal

impact of equivalent velocity as described in SAE J2464:2009 [61] or

Systems intended to be removed from the vehicle for charging (or

95–100% max. normal vehicle operation

12405-3:2014 [67–69] and ISO 62660-2:2011 [70].

3.1.3. Penetration test

and electrodes is compromised, short circuits are created and conse-

contact, together with the shorting occurring on the nail, so relatively

important damage occurs in a short period of time. Additionally, due to

heat dissipation is quite limited.

Depending on the test level (cell, module or pack), the depth of

penetration and the dimension of the nail vary as described in many of

the standards (see Table 5). In most cases, a 3 mm diameter rod is

Not specified (C) 3

for module or pack testing [61,65,72,74,75]. In all cases the rod

The usefulness of this test is questioned by many in the research

representative of an event that would likely occur in a real - situation

(e.g. a sharp object penetrating inside the battery compartment within

Furthermore, it is uncertain as to the inﬂuence the quality and

the reasons mentioned above it seems comprehensible that this test is

not included in many of the automotive standards and regulations as

3.1.4. Immersion test

The immersion test has been developed to simulate a situation

water with a composition similar to seawater (e.g. 5 wt% NaCl (aq.)) at

25 °C for a period of at least 1–2 h or until any visible eﬀects (e.g.

recommended in FreedomCAR:2005 [65].

In the latest version of ISO 12405, part 3 published in 2014 [69],

the water immersion test is newly introduced (not included in parts 1

vehicle exposed to a ﬂood situation. For example in the Netherlands

vehicles to be submerged. Just to mention some examples, up to

several incidents involving EVs occurred (e.g. 16 submerged Fisker

incident [93]. Based on these incidents, it seems that the performance

of the test would be of relevance, however many of the standards and

3.1.5. Crush/crash test

Cell (C) 3 cells or

deformation. In the crush test, also referred to as battery enclosure

SOC (% rated capacity)

crushing up to 50% of initial dimension [61,65,72,75]) or until an abrupt

B whose characteristics and dimensions are displayed in Fig. 2. Typically,

described in IEC 62660-2:2011 [70]) is used. For module or pack

KMVSS 18-3 [73]

≥ 2 h or until any visible reactions have

Nominal composition of sea water

International Electrotechnical Commission (IEC) for permission to reproduce

Guidance on crush plate position is missing in most of the cases.

> 50% normal

has to be limited to 1000 times the weight of the battery [61,65,70,72]