Classification procedure of the explosion risk areas in presence of hydrogenrich syngas: Biomass gasifier and molten carbonate fuel cell integrated plant

Fuel 99 (2012) 245–253

Contents lists available at SciVerse ScienceDirect

Fuel
journal homepage: www.elsevier.com/locate/fuel

Classification procedure of the explosion risk areas in presence
of hydrogen-rich syngas: Biomass gasifier and molten carbonate fuel cell
integrated plant
A. Molino a,⇑, G. Braccio a, G. Fiorenza a, F.A. Marraffa b, S. Lamonaca b, G. Giordano c, G. Rotondo b,
U. Stecchi b, M. La Scala b
a

Italian National Agency for New Technologies, Energy and Sustainable Development, Trisaia Research Centre, S.S. 106 Jonica, km 419+500 – 75026 Rotondella, Matera, Italy
Politecnico of Bari, Electrical and Electronics Dept., Via E. Orabona 4 – 70125 Bari, Italy
c
University of Calabria, Chemical and Materials Engineering Dept., Via Pietro Bucci – 87036 Arcavacata di Rende, Cosenza, Italy
b

a r t i c l e

i n f o

Article history:
Received 24 January 2012
Received in revised form 20 April 2012
Accepted 23 April 2012
Available online 12 May 2012
Keywords:
Explosion risk

Biomass gasifier
Molten carbonate fuel cell

a b s t r a c t
This paper deals with the safety aspects of a 500 kWth (thermal power) biomass gasification plant coupled with a 125 kWe (electric power) molten carbonate fuel cell. In particular, it describes the procedure
for assessing the explosion risk in presence of hydrogen-rich syngas and compares the results given by
the application of technical standards with those obtained by the implementation of a fluid dynamic
model for the potential emission scenarios.
Ó 2012 Elsevier Ltd. All rights reserved.

0. Introduction
Among several hypotheses of energy development, a now well
established prospect look at the hydrogen as an energy carrier.
The reasons for this choice are essentially due to environmental
factors, rather than to the hydrogen use flexibility and, not least,
to the uncertainty on supply costs of the existing conventional
primary energy sources. This rationale has oriented research to develop technologies that allow to directly use hydrogen in energy
conversion systems (fuel cells) with high efficiency and low environmental impact; unlike other devices normally used for energy
production, this technology gives back only water vapor emissions.
Assuming to use hydrogen as the energy carrier of the future, a
crucial aspect is its production. The biomass gasification is of great
interest because of its renewable nature. In this respect, the ENEA
Trisaia Research Centre is involved in developing a 500 kWth
biomass gasifier and a 125 kWe molten carbonate fuel cell integrated plant.
The integration between fuel cell and gasification plant represents a potential path to the electric generation from biomass,
increasing the efficiency and lowering the environmental impact
(50% CO2 reduction). Based on a fluid-dynamic approach, the
explosive atmosphere area has been assessed in order to satisfy a
⇑ Corresponding author. Tel.: +39 (0)835 974736; fax: +39 (0)835 974210.
E-mail address: (A. Molino).

0016-2361/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
/>
fire engineering performance-based approach. In addition, the
explosion risk area has been evaluated with the well-known ATEX
risk assessment. This paper is aimed at comparing both procedures
in order to validate the ATEX-based technical standards for the
pilot plant located at the ENEA – Trisaia Research Center.
At the time of this study, no Integrated Gasification Fuel Cell
(IGFC) system is operating worldwide but ENEA Trisaia pilot plant.
As it can be observed from the literature survey, most of the available studies in this field deal with general aspects of the involved
technologies and perspectives of their combination [1–3].
The biomass gasification pilot plant operating at the ENEA Research Centre of Trisaia exploits a dual fluidised bed (DFB) reactor
having 500 kWth capacity. The reactor uses steam as gasification
agent, so that a fuel gas nearly nitrogen free is produced characterized by a relatively high Lower Heating Value, approaching 13 MJ/
Nm3 on a dry basis.
The gasification concept is the well-known Fast Internally Circulating Fluidised Bed (FICFB) [4], which was developed since the
mid-nineties by the Vienna University of Technology (TUV) and
Austrian Energy & Environment (AE&E) on a laboratory test ring
(100 kWth) [5]. Subsequently, within the scope of the European
project ‘‘Hydrogen-rich gas from biomass steam gasification’’ the
ENEA Trisaia 500 kWth gasifier was designed, constructed and
tested [6].
In the framework of the following European project ‘‘Clean
energy from biomass’’, an innovative hot gas cleaning section


246

A. Molino et al. / Fuel 99 (2012) 245–253

Nomenclature
a
A
c
cp
cv
Co

hazardous distance in the emission direction (m)
surface emission (m2)
gas concentration
specific heat–constant pressure (J/kg K)
specific heat–constant volume (J/kg K)
number of the air exchange, referred to the total volume
(sÀ1)
hazardous distance from a sorgent emission (m)
dz
DHc
combustion heat (kcal/mol)
E
specific internal heat (J/kg)
f
efficiency factor
fe
external force for volume (N/m3)
k
security factor
kdz
safety factor applied to the LFL
Kx, Ky, Kz gas diffusion coefficients

LFL
lower flammability limit (m3/m3Á100)
UFL
upper flammability limit (m3/m3Á100)
LOC
minimum oxygen concentration
M
molar mass of flammable substance (kg/kmol)
P
absolute pressure in the containment system at the
point of emission (Pa)

was added consisting of an adsorbing reactor for the removal of
acid compounds, a hot gas cyclone and a hot gas filter for the removal of coarse and fine particles, respectively [7].
The DFB steam gasification pilot plant has been coupled with a
molten carbonate fuel cell (MCFC).
Under the maximum load conditions characterized by a current
of 1100 A, this MCFC can generate 125 kW. The stack operating
temperature is around 650 °C, in order to avoid the carbonate salt
mixture solidification, and the operating pressure is 3.5 bar, in order to create the appropriate fluidodynamic conditions for the
system.
The MCFC technology exploits also the carbon monoxide to produce electricity, this is an important fraction of the producer gas of
around 25% of the volume on dry basis [8]. The reaction quickly
balances because of the high temperature of the stack and a greater
amount of H2, flowing into the fuel cell, is available for the following anodic reaction.

1. Safety standards for explosive atmosphere risk analysis
Areas in which there is a risk of explosion that may harm people
or the environment are subject to legal or technical comparable
rules in most countries of the world. While these rules were initially issued at the national level, in Europe they have since been

replaced over the last years by regional European Directives and
Standards, and in the field of standardization they have partially
been replaced by international regulations.
1.1. European directives
In 1976, the Council of the European Community established
the prerequisite for the free trade of explosion-protected electrical
equipment within the European Union by ratifying the ‘‘Directive
on the harmonization of the laws of the member states concerning
electrical equipment for use in potentially explosive atmospheres
(76/117/EEC)’’. This directive has since been adapted to the state
of the art by means of national laws and guidelines on electrical
equipment.

~
q
Qamin
Qg
R
S
t
T
Ta
~
v
Vz
wa
yj

c
u

q
ry, rz

heat flux (W/m2)
minimum ventilation mass flow rate (m3/s)
maximum flow rate emission of gas/vapor (kg/s)
universal gas constant = 8314 J/kmol K
stress tensor (Pa)
time (s)
temperature (K)
environment temperature
vector of gas velocity whose components are respectively u, v, w (m/s)
volume of potentially explosive atmosphere (m3)
reference velocity of the air in the considered ambient
(m/s)
molar fraction
specific heats ratio (cp/cv)
critical ratio of flow rate
density (kg/m3)
standard deviation of the wind speed in the transversal
and orthogonal direction

Complete harmonization and extension to all types of equipment was achieved with the new Directive 94/9/EC in 1994 [9].
The Directive 99/92/EC, which regulates operation in hazardous
areas and defines safety measures for the concerned personnel,
was issued in 1999 [10].
In addition to the 94/9/EC Directive, which regulates how explosion-protected equipment and protective systems are placed on the
market and the design, construction and quality requirements to be
met by them. The 99/92/EC Directive stating ‘‘Minimum requirements for improving the health and safety protection of works
potentially at risk from explosive atmospheres’’ refers to the operation of potentially explosive installations, and is, therefore, intended for the employer. This directive contains only minimum

requirements. When implementing it into national law, the single
states can adopt further regulations. Examples are the implementation of the directive into the British law by ‘‘The Dangerous Substances and Explosive Atmospheres Regulations’’ and into the
German law by ‘‘The Betriebssicherheitsverordnung’’, the German
regulation on Industrial Safety and Health Protection, which takes
into account further European directives on safety on work.
Comparable regulations are found in other European countries.
According to the 99/92/EC Directive, it is duty of the employer to
verify where there is a risk of explosion, classify the hazardous
areas into zones accordingly, and document all measures taken
to protect the personnel in the so-called explosion protection
document.
1.2. Assessment of explosion risks
When assessing the risks of explosion, the following factors are
to be taken into account:
À the likelihood that explosive atmospheres will occur and their
persistence;
À the likelihood of ignition sources, including electrostatic
discharges;
À the installations which can give rise to an explosion, substances
used, processes, and their possible interactions;
À the scale of the anticipated effects.


247

A. Molino et al. / Fuel 99 (2012) 245–253

The employer has to classify the areas in which explosive atmospheres may appear into risk-zones, and ensure that the minimum
organisational and technical requirements of the Directive are observed. Since our assessment refers to gases, we restrict our discussion about ATEX Directives applied to gas explosive atmosphere
[10]. The zone classification refers to gases only.

Zone 0: An area in which an explosive atmosphere consisting in
a mixture of air and flammable substances in the form of gas, vapor
or mist is present continuously or for long periods or frequently.
Zone 1: An area in which an explosive atmosphere consisting of
a mixture with air or flammable substances in the form of gas,
vapor or mist is likely to occur in normal operation occasionally.
Zone 2: An area in which an explosive atmosphere consisting of
a mixture with air of flammable substances in the form of gas,
vapor or mist is unlikely to occur in normal operation but, if it does
occur, will persist for a short period only.
An explosion protection document has to be generated, which
contains at least the following information:
À
À
À
À

assessment of the explosion risk
protective measures taken
zone classification
observance of minimum requirements.

This piece of information can be divided into organisational
measures (e.g. instruction of workers) and technical measures
(e.g. explosion protection measures).
Hazardous areas are classified into zones to facilitate the selection of the appropriate electrical equipment as well as the design of
suitable electrical installations. Information and specifications for
the classification into zones are included in IEC 60079-10 [11]. In
Italy this assessment procedure is introduced through standard
CEI (Comitato Elettrotecnico Italiano) 31-30.

The methodology adopted in this paper is based on two guidelines adopted in Italy: (1) Guide CEI 31-35, ‘‘Electrical apparatus for
explosive atmospheres. Guide for the application of the Norm CEI
EN 60079-10 (CEI 31-30)’’ and (2) Guide CEI 31-35/A, ‘‘Electrical
apparatus for explosive atmospheres – Guide for the application
of the Norm CEI EN 60079-10 (CEI 31-30) Classification of hazardous zones, Examples of application’’ [11,12]. We refer to CEI 31-35
since this standard has no equivalent among the IEC standards.
Furthermore, we have perceived this methodology, as an important tool for the application of IEC 60079-10, since it includes a
lot of specific issues not considered there.
These two guidelines give special features for determination of
the type of the zone and for the evaluation of its extension. The
standard EN 60079-10 does not provide sufficient information
about the decision process for classification. The Italian Guide
CEI 31-35 reports some detailed mathematical formulation on
how to proceed to the application of standard CEI 31-30. When
the type of the zone has been determined, the Italian methodology include a procedure for checking that the likelihood of the
explosive atmosphere in one year and the total duration of the
explosive atmosphere in one year (release duration plus time of
persistence after the release has ends up) are below some critical
values. This verification introduces a probabilistic risk-based
approach.
The method is a stepwise process that gives both the type and
extension of the zone. The guideline contains indications on:
(1) the most suitable leakage hole dependent on the type of
component (i.e. pump/compressor, piping connections,
valve, etc.);
(2) flow rates for structural/continuous grade gas release as a
function of the component type based, on statistical data;

(3) flow rates for primary and secondary grade gas release calculated by specific reference formulas;
(4) evaluation of the extension of the hazardous zone as a function of the release flow rate, ventilation and flammable

substances.

2. Plant description
The plant under study is installed at the ENEA Trisaia Research
Centre. It is an experimental pilot plant for the biomass-derivedhydrogen-rich syngas production and consists of a 500 kWth biomass gasifier coupled with a 125 kWe molten carbonate fuel cell
[13]. Two different areas for storing technical gases used in the
process are utilized. The first area relates to a plant of storage
and vaporization of nitrogen, carbon dioxide and oxygen. The second one is devoted to the implementation of a bunker for storing
hydrogen in a cylinder, with the relative unit of decompression.
Fig. 2.1 shows an advanced technological platform on biomass
gasification is available at ENEA Research Centre of Trisaia,
including:
(1) two air-blown fixed bed downdraft gasifiers having a fuel
capacity of 120 and 300 kW, respectively, with conventional
gas cleaning, consisting in filtration units and water scrubber, and combined to an internal combustion engine (ICE);
(2) a dual fluidised bed steam gasification pilot plant, having a
fuel capacity of 500 kW, with both hot gas cleaning, via an
adsorbing reactor and a filtration unit (cyclone plus ceramic
filter), and conventional cold gas cleaning;
(3) air/steam-blown fixed bed updraft gasifier, having a fuel
capacity of 150 kW, with advanced gas cleaning: coalescent
filters and bio-diesel scrubber;
(4) interconnected fluidised bed steam/oxygen gasification pilot
plant, having a fuel capacity of 1 MWth, with catalytic ceramic candles located inside the gasifier.
The evaluation of the risk areas in presence of hydrogen-rich
syngas was carried out with the dual fluidised bed steam gasifier
(2) coupled with a molten carbonate fuel cell.
The dual fluidised bed steam gasifier is based upon the FICFB
(Fast Internally Circulating Fluidised Bed) gasification process,
which was formerly developed by the Vienna University of

Technology [14]. This process is considered as commercial,

5

1

2

4
3

(1) fixed bed downdraft gasifiers
(2) dual fluidised bed steam gasifier
(3) fixed bed updraft gasifier
(4) interconnected fluidised bed gasifier
(5) molten carbonate fuel cell
Fig. 2.1. 3D view of the plant.


248

A. Molino et al. / Fuel 99 (2012) 245–253

provided that it has been tested since 2002 on the 8 MW thermal
capacity plant in Güssing and has by now reached a level of use
of 80%.
Furthermore, being rich in hydrogen, the producer gas is especially suitable to be used as the fuel for a MCFC. For these reasons,
amongst the different gasification systems available at the Research Centre of Trisaia, the dual fluidised bed steam gasifier has
been selected in order to be directly coupled to a MCFC.
The MCFC system is provided by Ansaldo Fuel Cells (AFCo), an

Italian manufacturing company, which is part of the of the Ansaldo
Group.
The electrical capacity is 125 kW, while operating temperature
and pressure are 650 °C and 3.5 bar, respectively. The high working
temperature is required in order to keep the electrolyte, which is
an alkali carbonate salt mixture, in the liquid state.
An horizontal ‘‘hot vessel’’ configuration is adopted, which requires the anodic stream at 200 °C, while the cathodic stream
can be utilized at ambient temperature [15]. In effect, the cathodic
stream is heated up to the operating temperature via an internal
catalytic burner, where exhaust gases from the anode are conveyed. The high temperature exhaust gases leaving the burner
are additionally used to heat up to the working temperature the
incoming anodic stream, via an heat exchanger located inside the
vessel also. Finally an internal blower allows the recirculation of
the cathode stream, thus minimizing the correlated energy
requirements.
As it can be deduced, the operation of the MCFC requires several
auxiliary devices. Besides the equipments which are located inside
the vessel and the anode heater, a cathode pre-heater is necessary
in order to increase the cathodic stream temperature up to 300 °C
during the plant heat up stage. However, the most onerous auxiliary device is represented by the system for storage, vaporization
and mixing of technical gases, such like hydrogen, nitrogen and
carbon dioxide. These gases are necessary during MCFC heat up
and cool down, which must occur under controlled feed conditions.
Furthermore, they can be used to create apposite fuel gas mixtures simulating, during the tests, a gas from a given process (biomass gasification, anaerobic digestion, etc.).

3. Flammable substances
Flammable gases are in various sections of the plant. The syngas
produced by gasification contains H2, CO, CH4, C2H2, C2H4, N2. In
some tests, the fractions of the combustible products in the mixture could reach values such as to achieve a high calorific value.
Therefore, in order to guarantee the absolute safety of the plant,

the syngas has been characterized from a flammability point of
view, so as to explore any possible flammable scenarios that could
involve the reactor. The flammability limits of each components of
the typical mixture in the gasifier are reported hereafter: (see also
Table 3.1)
When the fuel is a mixture of several substances, the lower and
the upper limits are computed on the basis of additivity criteria. A

Table 3.1
Lower flammability limit (LFL) and upper flammability limit (UFL) of each syngas
component.

CO
CO2
CH4
H2
N2
C2H6
C3H8

%Vol.

LFLair25°C

UFLair25°C

25.00
19.00
10.00
33.00

10.00
0.50
2.50

12.50
0.00
5.30
4.00
0.00
3.00
2.37

74.00
0.00
15.00
75.00
0.00
12.40
9.50

widely used rule is the so-called Le Chatelier rule, also known as
the law of mixtures, as follows:

100
LFL ¼ P yj

j LFLj

100
UFL ¼ P yj

j UFLj

where LFLj and UFLj denote lower/upper flammability limit in air of
the j-th substance, respectively.
Once knowing the flammability limits of each component of the
mixture, it is possible to compute the lower and the upper limits in
the case of a mixture with a greater amount of hydrogen [16,17]
obtaining the following results:

LFLmix25 C ¼ 7:11½m3 =m3 Á 100Š
UFLmix25 C ¼ 63:45½m3 =m3 Á 100Š
In order to evaluate the risk related to the explosive atmospheres, one have to consider that the mixture, for incidental
causes, may be in contact with the combustive agent (air, oxygen
and nitrogen) at the process temperature (around 900 °C); consequently, it is necessary to take into account these change of LFL,
UFL with the temperature. Usually this effect is taken into account
through safety margins, in our approach we propose to actually
evaluate these variables at the process temperature.

4. Safety analysis according the CEI 31-35 norm
4.1. Flammability limits
The LFL, UFL have significance if computed at 25 °C. Therefore,
because the mixture may come into contact with the oxidizing at
the process temperature (900 °C), they have to be assessed as a
function of the temperature. To address this need, the norm CEI
31-35 adopts wide margins using the safety coefficient k that is
the safety factor applied to the LFL for the definition of the minimum ventilation mass flow rate Qamin and hypothetical volume
of potentially explosive atmosphere Vz [12]. So, applying the CEI
31-35 Italian standards, this correction coefficient is assumed
equal to 0.5.

4.2. Sources of emissions
The Emission Source (ES) is a point or a part of a plant from
which a flammable gas, vapor or liquid can be emitted to generate
an explosive atmosphere. The Norm CEI 31-35 provides some estimates, useful for making an assessment of the emission flow.
For that it regards the plant under study, the sources of emissions are essentially represented by flanges and valves. The size
of the emission hole of a flange is defined by taking into consideration the seal failure.
For a ring type joint (RTJ) flange (metal-to-metal), a serious failure may lead to a hole with a thickness of 0.05 mm and a length of
10 mm, or an area of 0.5 mm2.
The size of the emission hole of a valve is instead defined by
taking into account the emission of the stem. In industrial practice,
the area of the emission hole is assumed equal to 0.25 mm2 for
general use valves on pipes having a diameter smaller or equal to
150 mm; 2.5 mm2 for general use valves on pipes having a diameter higher than 150 mm and for severe service valves on pipes of
any diameter.


249

A. Molino et al. / Fuel 99 (2012) 245–253

Safety valves that do not discharge into a torch or blow down
provide both second-grade and first-grade emissions depending
on their behavior during the ordinary operation of the plant [12].
In CEI 31-35 the Emission Sources are categorized according to
level of hazard:
1. Continuous Grade Emission Sources, when the flow is continuous or at least the case for a long time
2. First Grade Emission Sources, when the emission is in regular
form, but not prolonged, or occasional, but nevertheless
expected in normal operation
3. Second Grade Emission Sources, when the emission is not provided for short periods and in normal operation.

4.3. Emission discharge
For each SE, the emission discharge under cautionary conditions
can be computed. In case of continuous or first-grade emissions, it
should be evaluated according to the features of the containment
system and the effective size of the openings; in case of secondgrade emissions the above mentioned evaluation criteria should
be applied. In order to assess the emission discharge in case of a
gas leakage from a containment system in which the pressure does
not substantially drop for the effect of the considered emission, the
following formula has to be applied:

"
Qg ¼ u Á c Á A Á c Á

b #0:5
2
P
ÁÀ
Á0:5
cþ1
RÁ T



M

where b ¼ ccþ1
, Qg is the maximum flow rate emission of gas/vapor,
À1
u is critical ratio of flow rate, c is the concentration of gas, A is the
surface emission, c is the specific heats ratio, P is the absolute pressure in the containment system at the point of emission, T is the reference temperature and M is the molar mass of flammable

substance.
4.4. Degree and availability of ventilation
With regard to ventilation in areas with flammable gases or vapors, the CEI EN 60079-10 guide considers the ventilation in a
quantitative manner (degree of ventilation) and according to the
reliability which air is available with [11]. The degree of ventilation
represents the ratio between the amount of air which affects the
emission source and the amount of flammable substances emitted
in the environment. The assessment of the degree of ventilation requires first information on the minimum mass flow rate of ventilation air (Qamin), defined as the mass flow rate of air (m3/s) needed
for diluting the mass flow rate (Qg) of the dangerous substance
associated to the emission, below the LFL, with a safety margin
varying with the emission degree. Either for indoors or outdoors,
the minimum mass flow rate of ventilation air (Qamin) can be computed using the following formula:

Q a min ¼

Qg
Ta
Á
k Á LFLm 293

where Ta is environment temperature and k is a safety factor applied to the LFL.
Then, it is necessary to determine the hypothetical volume of
potentially explosive atmosphere (Vz) around the source of emission. This can be done using the following formula:

Vz ¼

f Á Q a min
Co

For outdoors, ventilation per unit time Co is assumed equal to

0.03 sÀ1. The efficiency factor f, in the case of outdoors with the
presence of some free air circulation impediments not able to

reduce the effective dilution capacity of the air in the volume
affected by the flammable emissions, was assumed to be two
[12]. Looking at the obtained results, for the plant under study
the hypothetical volume of explosive atmosphere Vz can be considered negligible, as a result the degree of ventilation is high.
In order to define the effectiveness of ventilation, another
important parameter should be considered: its availability. The
availability of ventilation has an influence on the presence or formation of an explosive atmosphere and expresses the availability
level of the degree of ventilation. It can be: good, adequate or poor.
The availability is good when ventilation (mass flow rate and
factor of efficiency) is continuous. Very brief interruptions can
sometimes be admitted. With natural outdoors ventilation, the
availability is generally good if a wind velocity of 0.5 m/s is considered, conventionally representative of ‘‘calm wind’’ which is always present in practice. For what concerns the plant under
study, the wind measures have shown that a wind velocity of
2 m/s can be assumed, therefore a good availability is present.
4.5. Zone type
Once known the degree of emission, the degree of ventilation
and the availability of ventilation the zone can be classified. Basically, there are three zone types: 0, 1 and 2. The type of emission
is closely related to the degree of emission. Generally, a continuous-grade emission produces a 0-type zone, a first-grade emission
a 1-type zone and a second-grade emission a 2-type zone. The
element which can affect this biunique correspondence is the ventilation [12].
In this case, since the sources of emission in the plant are second-grade, the degree of ventilation medium and the availability
of ventilation good, the risk areas can be classified as 2-type areas.
4.6. The risk distance
Once determined the zone type, the risk distance have to be
computed. The risk distance (dz) is the distance from the source
of emission (SE) starting from which the flammable gas or vapor
concentration in the air is less than LFL. Again, technical literature

provides adequate formulas for computing this value. Hereafter,
some of them, taken from the CEI 31-35 guide, provide precautionary values for the classification of risk areas. The formulas are
applicable to outdoors problems [12]. For emissions as a free jet
of gas or vapor with high velocity the formula reads:

dz ¼

1650
Á ðP Â 10À5 Þ Á M À0:4 A0:5
kdz Á LFLV

where dz is the dangerous distance from the emission source, M is
the molar mass of flammable substance (kg/kmol), A is the area
(section) of the hole of emission (m2) and P is absolute pressure
in the containment system at the point of emission (Pa).
For emissions as a free jet of gas or vapor with slow velocity, instead, the formula reads:

dz ¼



42300 Á Q g Á f
M Á kdz Á LFLV Á wa

0:55

where wa is the reference velocity of the air in the considered ambient (m/s), kdz is the safety coefficient applied to the LFL for the
definition of the distance dz.
The risk distance, computed with the above mentioned formulas, can be used to approximately evaluate the extension of the risk
zone but not its real size, which instead, have to be defined considering the specific situation by an expert technician, through experimental works and/or specific guides or recommendations.

Therefore, it is needed to evaluate the extension in the emission


250

A. Molino et al. / Fuel 99 (2012) 245–253

direction (quota ‘‘a’’) which has to be at least equal to the risk
distance dz. Usually, this distance is assumed for safety scopes. In
the absence of exact data, it is reasonable to assume a safety margin of 20% for defining the quota ‘‘a’’.

Table 5.2
LFL, UFL, LOC at 900 °C.
LFLmix900°C

UFLmix900°C

LOCmix900°C

6.37

76.45

3.5

5. Explosion risk analysis with a fluid dynamic model
5.1. Flammability limits

LOC ¼ LFL Á

The flammability limits computed in Section 2 are valid at the
temperature of 25 °C. Being the process temperature around
900 °C, in our approach we introduce their evaluation as a function
of temperature. At this aim, the following empiric relationships valid for alkanes are used [16]:

LFLt ¼ LFL25 À

0:75 Á ðT À 25Þ
DHC

UFLt ¼ UFL25 À

0:75 Á ðT À 25Þ
DHC

where DHc is the combustion heat and t is the time.
The applicability of these empiric equations requires, because of
the different species involved in the syngas, data on the flammability limits at the same temperature and pressure conditions. These
equations requires the following assumptions. During the evolution of the reaction, the thermal capacity and the molar composition of the mixture are considered constant. The kinetics of the
combustion of the pure species is not significantly influenced by
the presence of other fuels. In Table 5.1 the values of the upper
and lower flammability limits, LFL/UFL (expressed in volume percent of the substance, %vol.) are shown when the temperature varies for each species in the syngas:
While these assumptions may be considered valid for calculating the lower flammability limit, but they introduce not negligible
errors for upper flammability limits. Fortunately most of the procedure illustrated before about the application of Guidelines CEI
31-35 is based on LFL which plays a major role in safety. The low
flammability limit influence directly the LOC number while the
upper flammability limit has lower influence for the analysis because in this work is considered the diameter risk necessary to exit
outside the flammability area started from syngas concentration
composed by a fixed concentration of combustible gases.
Moreover, the flammability limits, as well as the reaction velocity and flame propagation velocity, are influenced by the pressure.

The effect of the pressure on the flammability limits is not always
easily predictable, since it is specific for each mixture.
The flammability range of fuel-oxidizer mixtures is more easily
computable by using the triangular diagram (mixture-oxidizerair). For the construction of this diagram, it is necessary to know
the flammability limits in air, in pure oxygen and the lower oxygen
concentration (LOC) under which the reaction does not provide the
energy needed to heat the entire mixture. The minimum oxygen
concentration is defined by the following relationship:

molO2 stoichiometric
moltotal

The table below (Table 5.2) shows the lower and the upper
flammability limits (% of Volume, express in m3/m3Á100) in oxygen
and the minimum oxygen concentration at the temperature of
900 °C:
The flammability diagram (Fig. 5.1) at the temperature of 900 °C
represents the worst condition in which the syngas is utilized
namely the temperature of the process. Flammability diagrams
show the regimes of flammability in mixtures of fuel, oxygen and
an inert gas, typically nitrogen. Mixtures of the three gasses are
usually depicted in a triangular diagram, also known as a Ternary
plot.
All the black lines represent the different composition of the
syngas mixture. The air line represents all the possible combinations of air/fuel. The UFL and the LFL spots are defined just above
this line. The stoichiometric mixture line describes all the possible
combinations between fuel and oxygen.
5.2. Environmental conditions
From a meteorological point of view, both the wind and the
atmospheric stability widely affect the gas dispersion. Wind is described and quantified by the following attributes: velocity, direction and turbulence. In meteorology, atmospheric conditions can

be: stable, unstable and neutral. Dispersion is greatest for unstable
conditions and lowest for stable conditions. The gas dispersion
mainly depends on meteorology (wind, atmospheric stability,
humidity, solar radiation, ambient temperature, cloudiness). Other
important aspects to take into consideration are: latitude, month of
year, time of day, roughness, topography of the area. In meteorology, the atmospheric turbulence seriously affect the dispersion of
dangerous substances. In order to study the atmospheric turbulence, the atmospheric boundary layer needs to be assessed
focused.
For the proposed analysis the considered environmental conditions are:
À wind speed: 2 m/s;
À wind direction: x;
À solar radiation: 1520 kWh/mq/year
These environmental conditions influence the standard deviation rx and ry contained in the Pasquill–Giffors law’s that will be
used for the evaluation of the risk distances.

Table 5.1
LFL e UFL (%Vol.) for different temperatures and different species in the syngas [18].
T (°C)

LFL CO

UFL CO

LFL CH4

UFL CH4

LFL H2

UFL H2

LFL C2H6

UFL C2H6

LFL C3H8

UFL C3H8

100
200
300
400
500
600
700
800
900

12.3
12.0
11.7
11.5
11.2
10.9
10.7
10.4
10.1

73.8

73.5
73.2
73.0
72.7
72.4
72.2
71.9
71.6

5.23
5.15
5.06
4.98
4.89
4.81
4.73
4.64
4.56

14.9
14.8
14.7
14.6
14.5
14.5
14.4
14.3
14.2

3.87

3.87
3.87
3.87
3.87
3.87
3.87
3.87
3.87

74.8
74.8
74.8
74.8
74.8
74.8
74.8
74.8
74.8

3.00
3.00
3.00
3.00
3.00
3.00
3.00
3.00
3.00

12.4

12.4
12.4
12.4
12.4
12.4
12.4
12.4
12.4

2.35
2.35
2.35
2.35
2.35
2.35
2.35
2.35
2.35

9.48
9.48
9.48
9.48
9.48
9.48
9.48
9.48
9.48

251

A. Molino et al. / Fuel 99 (2012) 245–253

Fig. 5.1. Triangular diagram for the fuel mixture leaving the Joule plant at 900 °C.

5.3. Emission sources modeling
Once defined the flammability range, the accidental release of
dangerous chemical substances from different sources (pipes,
flanges, valves) have to be modeled. In this case, the more cumbersome (and less probable) situation has been modeled: the release
of syngas from a pipe after a cumbersome breakdown. The dispersion close to the gasifier area has been modeled. The input data for
doing this evaluation are: the mixture features, the source, the
modality and the duration of the release.
The dispersion of syngas represents an emission of a gaseous
substance in the environment followed by a dispersion of the gas
(vapor) cloud. The dispersion is therefore an effect of the emission.

where c is the gas concentration and Kx, Ky, Kz are gas diffusion coefficients (assuming anisotropy)
Depending on simplifying hypothesis (isotropy of diffusion, absence or constancy of the velocity component/s), the convection/
diffusion equation may have an analytical solution, otherwise it
needs to be integrated.
The dispersion of alight gas having a neutral buoyancy is defined a passive dispersion. In general, the neutral buoyancy is given
either by the high emitted gas dilution (low concentration) or by
its molecular weight similar to that of the surrounding air (in this
case the emitted gas temperature is similar to the atmospheric
one).
This representation has been longly adopted to describe the
emission from chimney [19,20]. These models also describe the
dispersion of substances for instantaneous or continuous emissions from land. Experimentally it was detected that either for
instantaneous or continuous releases from a punctiform source

posed on the ground, the concentration profiles are Gaussian
(Pasquil–Gifford) [20,21].
At the same time, for both kind of releases the variability of concentration increases with sampling time. The plume generated
from a continuous release tends to spread. It follows that the dispersion due to the turbulence is increased.
The concentration downstream of the point of emission depends on the intensity of the source except in the case it is significantly responsible of the convective motion transferred to the
emitted fluid.
For that it concerns the continuous and punctiform source, the
concentration is inversely proportional to the average of wind
velocity. In order to assess the concentration of pollutants and
the effect of the relative dilution, the model of Pasquill–Gifford
has been chosen as a starting point [19,20]. As an example, for a
continuous punctual dispersion from land the following functional
dependence is valid:


cðx; y; zÞ ¼

Qg

pry rz u

e

À12

y2 z2
þ
r2y r2z

!

5.4. Dispersion modeling
The laws
equations:

modeling

dispersions

are

the

Navier–Stokes

@q
mÞ ¼ 0
þ rðq~
@t
@ðq~
vÞ
þ rðq~
m~
v À ~SÞ ¼ ~f e
@t

where ry, rz are standard deviations of the wind velocity in the
transverse and vertical directions, Qg is the mass flow rate of the
dangerous substance associated to the emission, u is the wind
speed, y is the transversal distance from the emission hole and z

is the orthogonal distances from the emission plane.
The essential feature of the dispersion coefficients is that they
depend on the downwind distance and the class of meteorological
stability.

@ðqEÞ
v À ~S Á ~
v þ ~qÞ ¼ ~f e Á ~
v
þ rðqE~
@t
where ~
v is vector of gas velocity whose components are respectively u, v, w (m/s), q is the density (kg/m3) and E is the specific
internal heat (J/kg).
The frequently modeled and described scenarios are the instantaneous release (snort) and the continuous release (plume) from a
punctiform source [22].
There are two approaches for modeling the turbulent dispersion: the Eulerian and the Lagrangian approach, respectively.
Focusing attention only on the Eulerian approach, a possible dispersion model is that based on the convection/diffusion equation
(k model) [23].
The convection/diffusion for a gas in rectangular coordinates is:

@c
@c
@c
@c
@2c
@2c
@2c
þu þv
þ w ¼ Kx 2 þ Ky 2 þ Kz 2

@t
@x
@y
@z
@x
@y
@z

Fig. 6.1. Concentration profiles in a fuel mixture versus transversal and horizontal
distance from the emission hole (A = 0.5 mm2).


252

A. Molino et al. / Fuel 99 (2012) 245–253
Table 6.1
Results obtained by applying the CEI 31–35 guide and the fluid dynamic model.
Method

LFL (%)
A (mm2)
Vz (m3)
dz (cm)
a (cm)

Fig. 6.2. Concentration profiles in a fuel mixture as a function of transversal and
longitudinal distance from the emission hole (A = 0.5 mm2) for the syngas mixture
percentage lower than the lower flammability limit.

6. Test results

On the basis of the assumptions made for the two approaches,
namely the CEI 31-35 standard method and the fluid dynamic simulation, it was possible to compare them in order to validate the
standard approach when applied to a new technology such the
use of syngas obtained from biomass.
The fluid dynamic simulation is characterized by a more detailed representation of the problem and introduced LFL evaluation
at the process temperature (900 °C) instead of applying safety coefficients. In this case, assuming an hypothetical failure of a flange
with an emission hole of 0.5 mm2, a continuous spill of the overall
gaseous mass flow rate and a wind velocity of 2 m/s, the trends of
the syngas concentration at 900 °C as a function of distance can be
obtained.
Fig. 6.1 shows the effect of the dilution as a function of the longitudinal and transversal distances from the emission hole. In

Flange

Valve

CEI 31-35

Fluid dynamic

CEI 31-35

Fluid dynamic

7.11
0.5
0.02
10
12

6.37
0.5
//
5
6

7.11
0.25
0.01
6
8

6.37
0.25
//
3
4

order to appreciate the lower flammability limit of the syngas mixture at the emission temperature (LFLmix900°C = 6.73%), it is necessary to show a zoom for mixture composition lower that LFL, as
shown in Fig. 6.2:
Fig. 6.2 shows the volumetric percentage of syngas as a function
of the transversal and longitudinal distance from the emission hole
for volumetric percentage of syngas mixture around the LFL. For a
distance about of 1 cm from the emission hole, a syngas mixture is
lower than 2%Vol. This value vanishes almost completely for a distance of 2.5 cm from the emission hole (dz = 5 cm) at an orthogonal
distance from the emission plane of z = 0 cm.
At the other hand, assuming an hypothetical failure of a valve
with an emission hole of 0.25 mm2, a continuous spill of the overall
gaseous mass flow rate and a wind velocity of 2 m/s, a total dilution of the syngas mixture is observed at almost 3 cm from the
emission hole can be obtained.

Because the emission direction of the Emission Sources in the
plant is unknown, a spherical shape of the dangerous area has been
assumed. The study showed that all the emission sources give rise
to 2-type areas.
Figs. 6.1 and 6.2 show that the dangerous distance dz is equal to
10 cm for the flanges and 5 cm for valves.
Table 6.1 shows the results obtained applying the CEI 31-35
guide compared with the fluid dynamic model.
Where a is the effective extension of the dangerous area in the
direction of emission (m), Vz is the hypothetical volume of potentially explosive atmosphere. Table 6.1 shows that the approach
proposed by CEI 31-35 guidelines results very conservative with
regard to the fluid dynamic approach.
The results of the risk analysis are shown in the picture Fig. 6.3:
Fig. 6.3 shows that the biggest volumes of risk are located both
near the compression zone and in the exhaust gases zone in outlet
of the molten carbonate fuel cells.
7. Conclusions
This article discusses the safety aspects related to a 500 kWth
biomass gasifier and a 125 kWe molten carbonate fuel cell integrated plant, actually under construction at the ENEA Trisaia Research Centre. In particular, it describes the procedure to assess
the explosion risk due to the electricity in presence of hydrogenrich syngas. The results obtained by following the CEI 31-35 guide
were compared with a fluid dynamic model.
The most interesting result is that, either by applying the CEI
31-35 guide or by performing a fluid dynamic analysis, the dangerous distance from the emission sources has the same order of magnitude however the guidelines provide conservative results.
Therefore, the validity of the Italian guide is confirmed for this specific plant although it results very conservative.
References

Fig. 6.3. 3D plant view with risk volumes identification.

[1] Amos WA, Analysis of two biomass gasification/fuel cell scenarios for smallscale power generation. National Renewable Energy Laboratory. NREL/TP-57025886; 1998.

A. Molino et al. / Fuel 99 (2012) 245–253
[2] Lobachyov KV, Richter HJ. An advanced integrated biomass gasification and
molten fuel cell power system. Energy Convers Manage 1998;39:1931–43.
[3] McIlveen-Wright DR, Williams BC, McMullan JT. Wood gasification integrated
with fuel cells. Renew Energy 2000;19:223–8.
[4] <www.ficfb.at>.
[5] Hofbauer H, Veronik G, Fleck T, Rauch R. The FICFB gasification process. Dev
Thermochem Biomass Convers 1997;2:1016–25.
[6] Hofbauer H, Rauch R. Hydrogen-rich gas from biomass steam gasification.
Publishable final report of project JOR3-CT97-0196; 2001.
[7] Hofbauer H, Rauch R, Loeffler G, Kaiser S, Fercher E, Tremmel H. Six years
experience with the FICFB-gasification process. In: Proceedings of 12th
European conference and technology exhibition on biomass for energy.
Industry and climate protection, Amsterdam; 2002.
[8] Fiorenza G, Canonaco J, Blasi A, Braccio G. Biomass steam gasification at the
Trisaia dual fluidized bed pilot plant: experimental results and optimization
analysis of process variables. In: proceedings of zero emission power
generation workshop, Gebze, Turkey; 2007.
[9] Directive 94/9/EC of the European parliament and the council of 23 March
1994 on the approximation of the laws of the Member States concerning
equipment and protective systems intended for use in potentially explosive
atmospheres.
[10] Directive 1999/92/EC of the European Parliament and of the Council of 16
December 1999 on minimum requirements for improving the safety and
health protection of workers potentially at risk from explosive atmospheres.
[11] Norma EN 60079–10 (CEI 31–30): Costruzioni elettriche per atmosfere
esplosive per la presenza di gas. Parte 10: Classificazione dei luoghi pericolosi.

253

[12] Guida CEI 31–35 Costruzioni elettriche potenzialmente esplosive per la
presenza di gas. Guida all’applicazione della norma CEI EN 60079–10 (CEI
31–30). Classificazione dei luoghi pericolosi.
[13] Fiorenza G, Molino A. Analysis of integration of a molten carbonate fuel cell in
a biomass steam gasification plant; 2008
[14] Hofbauer H, Veronik G, Fleck T, Rauch R. The FICFB gasification process. In:
Proceedings of the international conference developments in thermochemical
biomass conversion, vol. 2. Banff, Alberta, Canada; 1996. p. 1016–25.
[15] Rossetti A, Scagliotti M, Valli C, Giannotti A, Bertone R. Molten carbonate fuel
cells research capabilities. In: proceedings of international symposium on
diagnostic tools for fuel cell technologies, Trondheim, Norway; 2009.
[16] Lees FP. Loss prevention in the process industries, vols. 1–3, Butterworth; 1996.
[17] Lees FP. Loss prevention in the process industries, vol. 1–3, Butterworth; 2004.
[18] Zabetakis, MG. Flammability Characteristics of Combustible Gases and Vapors,
Bulletin 627, US Bureau of Mines, XMBUA, USA; 1965
[19] Gifford FA. Atmospheric dispersion calculations using generalized Garrison
plume model. Nucl Safety 1961;2:56.
[20] Gifford FA. Use of routine meteorological observations for estimating
atmospheric dispersion. Nucl Safety 1961;2:56.
[21] Pasquill F. The estimation of the dispersion of windborne materials. Metall
Mag 1961;90:33.
[22] Seinfeld JH, Pandis SN. Atmospheric chemistry and physics. New York: WileyInterscience; 1997.
[23] Sharan M, Modani M. Variable K-theory for the dispersion of air pollutant in
low wind condition in the surface-based inversion. Atmos Environ
2007;41(33):6951–63.