Microsoft Word - 025.docx


 CHEMICAL ENGINEERING TRANSACTIONS  
 

VOL. 48, 2016 

A publication of 

 
The Italian Association 

of Chemical Engineering 
Online at www.aidic.it/cet 

Guest Editors: Eddy de Rademaeker, Peter Schmelzer
Copyright © 2016, AIDIC Servizi S.r.l., 
ISBN 978-88-95608-39-6; ISSN 2283-9216 

Safety Issues Related to Stationary Electrochemical Energy 
Storage on Industrial Sites 

Marion Demeestere*, Amandine Lecocq, Guillaume Fayet 

INERIS, Parc Technologique Alata, BP2, 60550 Verneuil-en-Halatte, France 
marion.demeestere@ineris.fr 

The present contribution discusses the safety issues related to stationary applications of electrochemical 
energy storage on hazardous industrial plants. Although only few accidents related to stationary 
electrochemical energy storage were reported, past accidents like the fire which occurred on a sodium-sulphur 
stationary system in September 2011 in Joso City (Japan) remind that the roll-out of such emerging 
technologies on industrial sites should be considered as both hazard sources and targets of potential domino 
effects. 
Different technologies of stationary batteries exist, from low to high energy/power ratio, according to their 
intended use. These various technologies, together with the different configurations that can be expected on 
an industrial site, are addressed. A number of standards specify tests to evaluate the safety and abuse 
tolerance of batteries for mobile and stationary applications. Some of them can be viewed in parallel with the 
French approach for controlling major accident hazards on industrial sites. It appears that some tests may 
enable to exclude some events from the quantification of the annual probability of occurrence of an accident 
scenario, which can provide greater flexibility for the hazardous industrial sites. 

1. Context 

Several guidelines in reducing the production of greenhouse gases have been taken since the Kyoto Protocol 
ratified in 1997. Among others, the diversification of the energy mix through the development of renewable 
energy constitutes a major move. In Europe, wind and solar power are expected to reach respectively 33% 
and 10% of the total power demand in 2030 (IEC, 2012). However, these renewable energies represent 
intermittent production sources and issues such as hourly variations in demand and price have been raised to 
consider them as a sustainable alternative for the production of electricity. 
In this context, electrochemical energy storage techniques enable to integrate higher levels of renewable 
energy sources and to achieve a better balance between demand and supply of electricity. Even if a number 
of electrochemical systems are already installed as stationary storage, this use is under development and its 
deployment is expected to increase in the coming years. 

2. Stationary electrochemical energy storage 

2.1 Technologies 
The selection of the most relevant technology for a stationary application of electrochemical energy depends 
on many parameters such as available power, energy capacity, reaction time, efficiency, life cycle, system 
safety and cost. Secondary (rechargeable) batteries have a short reaction time (< 1 second), a medium 
discharge time from few minutes to several hours and withstand a number of charge/discharge cycles. 
Secondary battery technologies that can be used for stationary applications include lead-acid, nickel-cadmium 
(Ni-Cd), sodium based battery (Na-S, Na-NiCl2) and Li-ion. In 2011, the distribution of the electrochemical 
storage capacity established in the world is as follows: 400 MW of Na-S, 45 MW of Li-ion, 45 MW of lead-acid 
and 40 MW of Ni-Cd (EDF, 2011). The analysis of the current stationary electrochemical energy storage 
market shows that Na-S systems are already widely deployed, especially in Japan, due to their high energy 
density and that Li-ion technology is expanding rapidly for stationary applications due to both its high power 
density and its high energy density. 

                               
 
 

 

 
   

                                                  
DOI: 10.3303/CET1648100

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Please cite this article as: Demeestere M., Lecocq A., Fayet G., 2016, Safety issues related to stationary electrochemical energy storage on 
industrial sites, Chemical Engineering Transactions, 48, 595-600  DOI:10.3303/CET1648100  

595



2.2 Configurations 
The battery allows managing fluctuations in demand, delivering frequency regulation, making more efficient 
use of the network (mitigation of congestion), providing emergency power supply for protection and control 
equipment, smoothing the power output provided by renewable energy sources, etc (IEC, 2011). Due to these 
various types of stationary applications, the battery energy storage system can be installed in a solar/wind 
farm or at substations, in a factory, a building, a hospital or a store or even in houses for residential 
application. For this latter, stored energy is about 3 kWh whereas it can reach several MWh in solar/wind 
power plants. 
Stationary electrochemical energy storage can be used on industrial sites for different types of applications 
including telecom, emergency power supply, load leveling, utility switching or peak shaving. Na-S systems can 
be found in various industrial sites such as automotive factories, aircraft manufacturers, non ferrous metal 
manufacturers, tobacco factories, chemical plants, perfume factories, and sewage disposal plants 
(Fukushima, 2013). Na-S system power on these industrial sites is generally between 1 MW and 5 MW. 

3. Safety issues on industrial sites 

3.1 Large scale battery storage accidents 
As they are quite new, only few accidents on large scale battery storages are reported. It’s worth mentioning 
the accident of Joso City in Japan that occurred on September 21, 2011. A 2 MW Na-S storage located in an 
industrial site of carbide parts and tools manufacturing caught fire and burnt during several days (ARIA, 2011). 
The fire was fully extinguished on October 5, 2011. Considering the reactivity of the sodium with water, the 
extinguishing method consisted in covering the fireplace with sand and waiting for the quench of the fire. No 
injuries were reported as a result of this fire. The consequences of this incident were material damages and 
economic loss for NGK battery manufacturer which asked its customers to stop their facilities during 
investigations. NGK published a report on the cause of Na-S battery fire incident, safety enhancement 
measures and resumption of operations (NGK, 2012). The cause of the fire was attributed to the failure of one 
cell in one of the 40 battery modules. This cell had a breach and leaked hot molten material which flowed over 
the sand filler inside the battery module, causing a short circuit between battery cells which heated and in turn 
caught on fire (Figure 1). This fire spread to the whole battery module that released flames and hot molten 
material causing the fire to spread to modules installed above and below. Safety enhancement measures 
taken by NGK manufacturer included the addition of fuses and insulation boards between blocks in battery 
modules to prevent short circuits, the addition of anti-fire boards between modules to stop fire propagation to 
the whole battery (Figure 2) and the development of fire fighting strategy and means to assist fire crew. 

 

Figure 1 : Cause of fire in the 2 MW Na-S storage in Joso (Japan) 

596



 

Figure 2 : Safety enhancement measures 

3.2 Standards for safety of electrochemical energy storage 
There are a number of standardized tests that evaluate the safety and abuse tolerance of batteries for electric 
or hybrid automotive applications, light electric rail applications, cellular phones and consumer electronic 
applications, etc. These standards have been established by numerous standardization bodies or test 
institutes such as UL, NEMA, SAE, IEEE, BATSO, TELCORDIA, JIS, EUCAR, INERIS, UN, IEC and ISO. All 
these standards rely on specific test methodologies that could be different even when considering the same 
application. Tests considered in these standards can be classified in 5 categories, which can be identified to 
accidental circumstances: 

- Electrical tests (overcharge, short circuit, discharge, etc); 
- Mechanical tests (impact, drop, crush, etc); 
- Thermal tests (fire, heating); 
- Environmental tests (altitude test, temperature cycling, etc); 
- Specialized tests (insulation resistance, projectile). 

A standard can be specific to one battery technology (Li-ion or Na-S for example) or can be more generic. 
Figure 3 gives an overview of the different tests of several standards that can be applied to automotive, 
consumer (mobile phone, laptop, etc) and industrial applications. This non-exhaustive list concerns non-
stationary applications. 
 

U
L

N
EM

A

SA
E

IE
EE

BA
T
SO

Te
lc
or

di
a

JI
S

IN
E
RI

S

U
N

IE
C

IS
OU

L

Te st C ri te ri a S tan dard

U
L

 1
64

2

U
L

20
54

U
L

 S
ub

je
ct

 2
27

1

U
L

 S
ub

je
ct

 2
58

0

U
L

25
75

C
18

.2
M

J2
46

4

IE
E

E
 1

62
5

IE
E

E
 1

72
5

B
A

T
SO

 0
1

G
R

-3
15

0

JI
S 

C
87

14

E
L

L
IC

E
R

T
 D

Pa
rt

 I
I 

S3
8.

3

IE
C

 6
21

33

IE
C

 6
22

81

IE
C

 6
26

60
-2

12
40

5-
1 

&
 1

24
05

- 2

Appl i cati on s* C, A, I C A A C C A C C A C C A C,A,I C C,A,I A A
Ext ernal short  circuit ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●
Abnormal charge ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●
Forced discharge ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●
Crush ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●
Impact ● ● ● ● ● ● ● ● ●
Shock ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●
Vibrat ion ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●
Heat ing ● ● ● ● ● ● ● ● ● ● ● ● ● ●
T emperat ure cycling ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●
Low pressure (alt it ude) ● ● ● ● ● ● ● ● ● ● ● ● ●
P roject ile ● ● ● ● ● ●
Drop ● ● ● ● ● ● ● ●
Cont inuous low rat e charging ● ●
Molded casing heat ing t est ●
Open circuit  volt age ●
Insulat ion resist ance ● ●
Reverse charge ● ●
P enet rat ion ● ● ● ●
Int ernal short  cicuit ● ● ● ●
Immersion ●
Fire ● ●

*Automotive: A; Consumer (Mobile phone; Laptop etc.): C; Industry: I 

Figure 3 : Existing standards for automotive, consumer and industrial applications (non-stationary) 

597



There are also published safety standards that can be used to evaluate the safety of stationary energy storage 
systems. However, many of these standards specifically address more traditional technologies such as lead-
acid or Ni-Cd batteries (IEC 62845-2, Telcordia GR-4228-CORE). There are IEC safety tests standards under 
development that address, among others, stationary Li-ion technology energy storage: IEC CD 62619 (not 
published), which covers industrial applications, and IEC 62897 (new approved work). 
Some standards such as UL 1973 (Table 1) are not technology specific and cover all types of batteries 
technologies as well as electrochemical capacitors, for use in light electric rail (LER) applications and 
stationary applications. This standard also includes requirements for unique technologies such as flow 
batteries and sodium-based batteries. 

Table 1 : Safety tests of UL 1973 standard 

Standard Safety tests 
UL1973 Short circuit, overcharge, over-discharge, imbalanced charge, dielectric voltage withstand,

continuity, temperature, failure of thermal stability system, temperature cycling, vibration*,
shock*, drop, enclosure tests, water exposure, external fire, internal fire 

  
* Only for LER applications 
 
Finally, the standard UL 9540 under development is not battery specific since its covers all types of energy 
storage systems and is not based on a “test only” approach since it includes construction recommendations, 
performance verification, markings and instructions. 
This analysis showed that standards addressing more recent stationary battery technologies (Li-ion) are under 
development and are not necessarily based on a risk assessment methodology that would consider the full 
scope of aggressions generated or imposed to the battery system. This is further addressed in paragraphs 3.4 
and 3.5 below. 

3.3 European regulation regarding major hazard accidents 
As stated in paragraph 2.2, stationary electrochemical energy storage can be used on industrial sites for 
different types of applications. Some of those sites may be hazardous to the surrounding people and 
environment, because of their activity (e.g. petrochemical plants) or because of the electrochemical energy 
storage located on the site, whose constituents have flammable and toxic properties. If need be, the operation 
of the site is regulated under the Directive 2012/18/EU of July 4, 2012, on the control of major-accident 
hazards involving dangerous substances (also called Seveso III Directive). This Directive introduces a 
common European policy for the prevention of major industrial accidents, based on risk analysis, safety 
management systems, emergency plans, land use planning policies, and information to the public. 
If the industrial site where stationary electrochemical energy is stored is covered by the regulation regarding 
hazardous facilities, it has to comply with several regulatory requirements. One of them can be the production 
of a safety report, which aims at demonstrating that major accident hazards and possible major accident 
scenarios have been identified and that the necessary measures have been taken to prevent such accidents 
and to limit their consequences for human health and environment. 
Acceptance criteria have been developed based on the combination of the two following parameters: the 
annual probability of occurrence of the accident scenario and its severity. The quantification of those two 
parameters for each accident scenario identified is one of the tasks of the safety report. The basis of a safety 
report is the risk analysis, where all the accident scenarios that could potentially happen on the site are 
identified, together with their prevention and mitigation measures. This exercise is based on the mapping of 
the hazard potentials present on the site. 
In the light of this, the electrochemical energy storage should be considered as a hazard potential when the 
risk analysis of the site is conducted. This activity should be assessed by itself, answering the question “what 
can happen to the storage?” It can also be a potential aggressor of other equipment, leading to a major 
accident scenario. For example, a liquefied petroleum gas sphere can be taken in a fire caused by the 
electrochemical energy storage. 

3.4 Stationary electrochemical energy storage and major accident scenarios 
Based on the past accident described in paragraph 3.1, the loss of containment of a cell is an undesirable 
event that can happen on stationary electrochemical energy storage. This event can lead to various 
dangerous phenomena, generating thermal or toxic effects according to the technology of the battery (e.g. Li-
ion batteries can produce hydrogen fluoride in case of combustion). Those effects threaten both human health 
and environment. When conducting the risk analysis, questions regarding the causes of the loss of 
containment are raised. These causes can be external or internal to the site (Figure 4). 

598



 

 

Figure 4 : External / Internal causes of loss of containment of electrochemical energy storage 

Several methods exist to quantify the annual probability of occurrence of the dangerous phenomenon. The 
most precise one consists of quantifying all the causes (also called initiating events) that can lead to the loss 
of containment of a cell and a resulting dangerous phenomenon in terms of frequency of occurrence. The 
aggregation of those frequencies throughout the accident scenario (e.g. combination with safety barriers) 
leads to the determination of the probability of occurrence of the dangerous phenomenon. 
Nevertheless, in the French regulation (Circular of May 10, 2010), it is clearly specified that some initiating 
events can be disregarded when the major accident scenario is to be quantified in terms of probability of 
occurrence. Regarding natural causes, the binding idea is the following: if the equipment has been conceived 
with regard to the corresponding regulation, the initiating event is not taken into account to assess the 
probability of occurrence of the accident scenario. For example, the initiating event “earthquake” can be 
excluded from the quantification process if the equipment has been conceived based on the seismic 
regulation. For the other types of initiating events, exclusion from the quantification process can happen if it 
has been demonstrated that the accident scenario is “physically impossible”. This approach consists in 
demonstrating that the conception of the equipment is ruled by a standard and that the conditions of use of the 
equipment cannot lead to aggressions higher than the ones described in the standardized tests procedures. 
The example developed in the next paragraph illustrates this approach. 

3.5 Example 
The standard UL 1973 evaluates the electric energy storage system's ability to safely withstand simulated 
abuse conditions (see Table 1). The defined requirements cover electric energy storage systems for use as 
energy storage for stationary applications and are not technology-specific. Table 2 describes, for a few 
examples, the nature of the tests performed on the energy storage system. 

Table 2 : Examples of safety tests of UL 1973 standard 

Safety tests of 
standard UL 1973 

Short description Results 

Short circuit (§ 14) The sample is to be short-circuited by 
connecting the positive and negative
terminals of the sample with a resistive
circuit load having a maximum total
resistance of 20 mΩ. 

The following events cannot be excluded:
explosion, fire, combustible vapour 
concentrations, toxic vapour release, 
electric shock hazard, leakage, rupture, 
loss of protection controls 

Impact (§ 27) The sample is to be subjected to a
minimum of three impacts of 6.8 J on any
surface that can be exposed to a blow
during intended use. 

The following events cannot be excluded: 
explosion, fire, combustible vapour 
concentrations, toxic vapour release, 
electric shock hazard, leakage, rupture, 
loss of protection controls 

External fire (§ 36) The sample is subjected to a hydrocarbon
pool fire for 20 minutes. The fuel used shall
be heptane similar hydrocarbon fuel. 

There shall be no explosion of the sample 
that results in projectiles falling outside a 
defined circular inner perimeter. 

Internal fire (§ 37) The sample is to be subjected to the
internal fire test which consists of heating
one internal cell that is centrally located
within the system until thermal runaway. 

There shall be no fire propagating from 
the system or explosion of the system. 

599



These results, combined with the conditions of use of the system on the industrial site, could lead to exclude 
the initiating events “external fire” and “internal fire” respectively from the quantification process of the major 
accident scenarios “explosion” and “fire / explosion”. In other words, the annual probability of occurrence of 
the potential fire or explosion can be assessed without taking into account the causes “external / internal fire” 
since the electric energy storage system is designed to prevent an external fire and a single cell failure within 
the system from cascading into a fire and explosion. In this specific example, the other initiating events cannot 
be disregarded because the results of the safety tests run in the standard are non-compliant for the events 
explosion and fire. 

4. Conclusions and perspectives 

Stationary electrochemical energy storage on industrial sites can be the source of accident scenarios 
potentially leading to major consequences. If the industrial plant is to be regulated by the European policy for 
the prevention of major industrial accidents, it may have to produce a safety report where all the physically 
possible accident scenarios are identified through a risk analysis. The acceptance of the risk is based on the 
combination of the probability of occurrence of those accident scenarios and their severity. 
Several methods exist for the quantification of the probability of occurrence of the identified accident 
scenarios. One of them consists of attributing a frequency of occurrence to each event initiating the scenario 
and propagating these data to get the final result. The present contribution shows that some initiating events 
could be excluded from this quantification process if the electrochemical energy storage system complies with 
some tests provided in acknowledged standards. 
This work could be completed by a more in-depth study of all the existing standards for stationary applications 
of electrochemical energy storage on industrial sites and a thorough risk analysis of the use of those systems 
considering specific configurations. For example, it would be interesting to apply this approach based on the 
standard IEC CD 62619, which addresses, among others, stationary Li-ion technology energy storage for 
industrial applications. 

References 

ARIA database, 2011, Incident N°41473: Joso fire 
Circular of May 10, 2010 summarizing the methodological rules applicable to safety reports, to the 

assessment of the risk reduction approach at the source and to the technological risks prevention plans 
(PPRT) in the installations classified under the Law of July 30, 2003 

EDF R&D, 2011, Le stockage sur territoire insulaire 
Fukushima Eishi, 2013, Inside cooperative innovation: development and commercialization of sodium-sulfur 

batteries for power storage, Hitotsubashi University Institute of Innovation Research. 
IEC White Paper, 2011, Electrical energy storage 
IEC White Paper, 2012, Grid integration of large-capacity Renewable Energy Sources and use of large-

capacity Electrical Energy Storage 
NGK press releases October 28, 2011 (www.ngk.co.jp/english/news/2011/1028_01.html) and June 7, 2012 

(www.ngk.co.jp/english/news/2012/0607.html) 
NGK, 2012, Cause of NAS Battery Fire Incident, Safety Enhancement Measures and Resumption of 

Operations 
UL 1973, 2013, UL Standard for Safety for Batteries for Use in Light Electric Rail (LER) Applications and 

Stationary Applications 
 

600