Journal of Loss Prevention in the Process Industries 44 (2016) 44e52

Contents lists available at ScienceDirect

Journal of Loss Prevention in the Process Industries

journal homepage: www.elsevier.com/locate/jlp

Risk assessment of a compound feed process based on HAZOP analysis

and linguistic terms

Luis Fuentes-Bargues a, d, *, Cristina Gonza
Jose
lez-Gaya b, Ma Carmen Gonza

lez-Cruz a,
Vero
nica Cabrelles-Ramírez c
a
Departamento de Proyectos de Ingeniería, Universitat Politecnica de Valencia, Camino de Vera s/n, 46022, Valencia, Spain
b
Departamento de Ingeniería de Construccio
n y Fabricacio

n, ETSII, UNED, C/Ciudad Universitaria S/N, 28040, Madrid, Spain
c
Universitat de Val
encia, Avda. de la Universidad s/n, 46100, Burjassot, Valencia, Spain
d
Departamento de Ingeniería Química, Universitat de Val encia, Avda de la Universidad s/n, 46100, Burjassot, Spain

a r t i c l e i n f o a b s t r a c t

Article history: The size and complexity of industrial plants, along with the characteristics of the products used, require a
Received 27 July 2015 study, analysis and control of the existing risks in every industrial process.
Received in revised form In this paper, a methodology for risk assessment in industrial plants, based on the combination of risks
22 July 2016
identification through the Hazard and Operability (HAZOP) analysis and the risks evaluation through
Accepted 28 August 2016
linguistic variables and fuzzy numbers is applied to a case study consisting on a compound feed plant
Available online 29 August 2016
located in the town of Silla (Valencia, Spain).
The results from this study show that the main risk in the compound feed production process is the
Keywords:
Risks
formation of explosive atmospheres (ATEX). Therefore, the corrective measures will focus on reducing
HAZOP the concentration of dust in the atmosphere and eliminating the possible sources of ignition, such as
Fuzzy numbers electrostatic discharges or sparks during the different phases of the process (the grinding, the transport
Explosive atmospheres of the raw materials, etc.)
Linguistic variables © 2016 Elsevier Ltd. All rights reserved.
Compound feed

1. Introduction been developed (Suokas, 1988; Papazoglou et al., 1992; Lees, 1996;
Papazoglou et al., 1996; Tixier et al., 2002; Papazoglou and Aneziris,
Technological and social development has led to an increase in 2003; Kim et al., 2005; Yun et al., 2009; Casamirra et al., 2009; Zhao
the size and complexity of industrial plants. These changes involve et al., 2009; Taveau, 2010; Kim et al., 2011; Demichela and
certain risks that need to be controlled and minimized. Camuncoli, 2014). This has led to the development of a scientific
Risk is understood as the possibility that someone or something discipline known as process safety that focuses on the prevention
is adversely affected by a hazard (Woodruff, 2005), while danger is of fires, explosions, and accidental chemical releases in chemical
defined as any unsafe situation or potential source of an undesir- processing facilities (Center for Chemical Process Safety (CCPS),
able and damaging event (Reniers et al., 2005). Other definitions of 2000); International Standard Organization (ISO), 2000) or what
risk are the measure of the severity of a hazard (Høj and Kro €ger, are known as serious accidents.
2002), or the measure of the probability and severity of adverse Directive 2012/18/EU (or Seveso III) (European Union, 2012)
effects (Haimes, 2009). defines as a serious accident an event (such as a major leak, fire, or
In recent decades interest in the safety of industrial plants has explosion) resulting from an uncontrolled process during the operation
greatly increased and many risk assessment methodologies have of any plant and producing a serious danger, whether immediate or
delayed, to human health or the environment, inside or outside the
plant, and involving one or more hazardous substances. Examples of
serious accidents in industrial processes include: Flixborough in
* Corresponding author. Departamento de Proyectos de Ingeniería, Universitat
cnica de Vale
Polite ncia, Camino de Vera s/n, 46022, Valencia, Spain. Britain (1974), Seveso in Italy (1976), Bhopal in India (1984),
E-mail addresses: [email protected], [email protected] (J.L. Fuentes- Enschede in the Netherlands (2000), Toulouse in France (2001) and
Bargues), [email protected] (C. Gonza
lez-Gaya), [email protected] Buncefield in Britain (2005). In Spain, examples include an accident
(M.C. Gonza
lez-Cruz), [email protected] (V. Cabrelles-Ramírez).

http://dx.doi.org/10.1016/j.jlp.2016.08.019
0950-4230/© 2016 Elsevier Ltd. All rights reserved.
 J.L. Fuentes-Bargues et al. / Journal of Loss Prevention in the Process Industries 44 (2016) 44e52 45

at the Repsol refinery in Puertollano (2003) in which an explosion has been combined with other methods, as a set pair-analysis to
in a gas storage area killed nine workers and injured many others, assess the major hazard of installations of storing flammable gas
as well as causing property damage. (Zhou, 2010), or with Layer of Protection Analysis (LOPA) in natural
On the other hand, the presence in industrial plants of dust from gas industry (Khalil et al., 2012) or with the fault tree analysis (FTA)
manipulating small particles or commodities such as cereals, along in the simulation of the causes of an accident for fire and explosion
with the combination of oxygen from air, can create potentially in crude oil tanks (Wang et al., 2013).
explosive atmospheres. These situations, with a source of ignition, The purpose of this article is the identification and analysis of
can cause explosions with catastrophic consequences. To prevent risks in a compound feed industry, with major hazards, using a
such accidents the European Union drafted the ATEX Directives, methodology based on a combination of risk identification through
Directive 1994/9/EC (European Union, 1994) concerning equipment HAZOP analysis and risk assessment using fuzzy numbers and
and protective systems intended for use in potentially explosive linguistic variables. The proposed method was tested previously in
atmospheres, and Directive 1999/92/EC (European Union, 1999), academic works with theoretical industrial cases and has subse-
which sets minimum requirements for the protection of the health quently been applied to the case of the paper: a compound feed
and safety of workers potentially at risk from explosive atmo- industry.
spheres in the workplace. The HAZOP technique allows to identify the deviations that may
The complexity and severity of accidents at these plants requires take place with regard to the intended functioning, as well as their
the implementation of risk management systems. The ISO 31000: causes and consequences, but does not permit to prioritize the risks
2010 (International Standard Organization (ISO), 2010) standard and the actions to mitigate the identified risks. This is a disadvan-
defines risk management as coordinated activities to manage and tage when the HAZOP technique is applied in industries at opera-
control an organization with regard to risk and comprises the tion phase. For this purpose HAZOP is combined with other
following steps (Fig. 1): communication and consultation, estab- techniques (quantitative or semi-quantitative), but in many times,
lishing the context, risk assessment (identification, analysis, and data obtained of the probability or the consequences or both are
evaluation), risk treatment, monitoring, and review. not adequate or accurate (for example, a value of probability of an
Risk assessment is the process of identifying, analyzing, and event in a fault tree analysis). The use of fuzzy logic or fuzzy set
evaluating the risk posed by an industrial plant and the main aim is theory allows to solve these problems, and can work with uncer-
the prevention and mitigation of accidents in potentially hazardous tainty and imprecision and solve problems where there are no

et al., 2009; Demichela and Camuncoli, 2014).
facilities (Dunjo sharp boundaries and precise values or the qualitative values do not
The phase of risk identification is the process in which risks are defined the whole spectrum of possibilities.
discovered and recorded. The analysis phase involves developing This paper shows an improvement of the in-depth knowledge of
an understanding of the risk and providing information for evalu- the combination of these tools. That way, researchers could use this
ation. The evaluation phase involves comparing the estimated risk case of study for future works and applications.
levels with predefined criteria to define the importance of the level
of risk and decide whether it is necessary to address the risk e as 2. Definitions
well as the most appropriate strategies and methods of risk
treatment. 2.1. HAZOP method
Choosing the appropriate risk assessment techniques is a diffi-
cult decision that will depend on factors such as the complexity of HAZOP studies evolved from the Imperial Chemical Industries
the problem, the methods for analysis of the amount of information (ICI) as a “Critical Examination” technique formulated in the mid-
available, the need for quantitative data, and available resources. 1960s. One decade later, HAZOP was published formally as a
One of the most used techniques for risk assessment is the disciplined procedure to identify deviations from the design intent.
HAZOP study (HAZard and OPerability) (Knowlton, 1981; Kletz, Lawley in 1974 (Lawley, 1974) defined and delineated the principles
1983). It is a qualitative technique that carries out a structured needed to carry out operability studies and hazard analysis, the
analysis of the process and allows identifying the deviations that Chemical Industry Association (CIA) published a guide in 1977
may take place with regard to the intended functioning, as well as (Chemical Industry Association (CIA), 1977) and many studies were
their causes and consequences. HAZOP does not try to provide developed in this field during the 1980s, as the guides written by
quantitative results but in many situations it is necessary to rank Knowlton (Knowlton, 1981) and Kletz (Kletz, 1983).
the identified risks, mainly to priorize the actions to mitigate them The HAZOP technique (International Standard Organization
because this decision depends of the risk level. For this purpose (ISO), 2011) is a structured and systematic methodology to iden-
HAZOP is combined with other techniques. In these cases, quanti- tify and document hazards through imaginative thinking. The
tative techniques like the fault trees or the event trees, or semi- sequence of typical HAZOP study is shown in Fig. 2. It involves a
quantitative techniques as the probability-consequence matrices very systematic examination of design documents that describe the
have been applied. However, the uncertainty of the information installation or the facility under investigation. The study is per-
quality or the evaluations' subjectivity has made possible the formed by a multidisciplinary team of technicians and engineers
introduction of the linguistic variables in the risk assessment with extensive knowledge on design, operation, and maintenance
techniques. By applying the principles of the fuzzy sets theory, the of the process plants (HAZOP team).
use of linguistic terms instead of exact number shave been inte- In HAZOP study, the process documents, the instrument dia-
grated into classic methodologies (Huang et al., 2001; Cho et al., grams and the design documents such as piping and instrumen-
2002; Markowski and Mannan, 2008; Shapiro and Koissi, 2015) tation diagram, cause and effect charts, etc., are examined
or have been used to develop new methodologies (Carr and Tah, systematically by the HAZOP team, and the abnormal causes and
2001; Loyd, 2004; Zhang and Zou, 2007; Wang and Elhag, 2007; adverse consequences for all possible deviations from normal
Nieto-Morote and Ruz-Vila, 2011). operation that could arise are found for every section of the plant.
Fuzzy assessment also has been used for risk assessment at in- Thus, the potential problems in the process plant are identified.
dustries, for example to assess the risk of process operations in the The HAZOP team members try to imagine ways in which haz-
oil and gas refineries (Sa'idi et al., 2014) or to evaluate the risk from ards and operating problems might arise in a process plant. To
explosive atmospheres (Markowski et al., 2011). This method also cover all the possible malfunctions in the plant, the HAZOP study
46 J.L. Fuentes-Bargues et al. / Journal of Loss Prevention in the Process Industries 44 (2016) 44e52

team members use a set of “guide words” for generating the pro-
cess variable deviations to be considered in the HAZOP study. A list
of guide words with their meaning and the parameters where they
can be applied is presented in Table 1.
When these guide words are applied to the process variables in
each line or unit of the plant, we get the corresponding process
variable deviation to be considered in the HAZOP study. The guide
words and process variables should be combined in such a way that
they lead to meaningful process variable deviations. Hence, all
guide words cannot be applied to all process variables. For example,
when the process variable under consideration is temperature, the
guide words MORE OF and LESS OF lead to meaningful process
variable deviations.

2.2. Fuzzy sets theory

The Fuzzy Set Theory introduced by (Zadeh, 1965; Zadeh, 1975a;

Zadeh, 1975b; Zadeh, 1975c) is suitable for dealing with imprecision Fig. 1. Model of risk management. Source: ISO 31.000:2010 (European Union, 1999).
and uncertainty associated with data in risk assessment problems.
In a universal set of discourse X, a fuzzy subset A of X is defined by a
membership function mA(x), which maps each element x in X to a
real number in the interval [0, 1]. The function value of mA(x) sig-
nifies the grade of membership of x in A. When mA(x) is large, its Fuzzy subtraction : A1.A2 ¼ ða1  d2 ; b1  c2 ; c1  b2 ; d1
grade of membership of x in A is strong (Kaufmann and Gupta,  a2 Þ
1991). (3)
Among the various types of fuzzy sets of special significance are
fuzzy numbers (Dubois and Prade, 1978) defined as A ¼ {x, mA(x)}
where x takes its number on the real line R and membership Fuzzy multiplication : A1 5A2 zða1 $a2 ; b1 $b2 ; c1 $c2 ; d1 $d2 Þ
function mA: R / [0, 1], which have the following properties: (4)

(i) A continuous mapping from R to the closed interval [0, 1].  

a1 b1 c1 d1
(ii) Constant on (-∞, a]: mA(x) ¼ 0 c x (-∞, a]. Fuzzy division : A1 /A2 z ; ; ; (5)
d2 c2 b2 a2
(iii) Strictly increasing on [a, b].
(iv) Constant on [b, c]: mA(x) ¼ 1 c x [b, c]. The fuzzy addition or the fuzzy subtraction of any two fuzzy
(v) Strictly decreasing on [c, d]. trapezoidal numbers is also a trapezoidal fuzzy number. But the
(vi) Constant on [d, ∞): mA(x) ¼ 0 c x [d, ∞). fuzzy multiplication or the fuzzy division is only approximate a
trapezoidal fuzzy number.
Where a, b, c, d are real numbers and eventually a ¼ - ∞, or b ¼ c, The scalar multiplication of a trapezoidal fuzzy number is also a
or a ¼ b, or c ¼ d or d ¼ ∞. trapezoidal fuzzy number defined as:
For convenience, mLA is named as left membership function of a
fuzzy number, defining mLA(x) ¼ mA(x), for all x [a,b]; mRA is named as A1 $k ¼ ða1 $k; b1 $k; c1 $k; d1 $kÞif k > 0 (6)
right membership function of a fuzzy number A, defining
mRA(x) ¼ mA(x), for all x [c, d].
A1 $k ¼ ðd1 $k; c1 $k; b1 $k; a1 $kÞif k < 0 (7)
A trapezoidal fuzzy number A is a fuzzy number denoted as A ¼
(a, b, c, d) which membership function is defined as:
2
0 for x < a
6 xa
6 L
6 mA ðxÞ ¼ for a  x b 3. Methodology
6 ba
6
mA ðxÞ ¼ 6
61 for b  x  c (1) The methodology has been structured in four phases:
6
6 R
6 m ðxÞ ¼ x  d for c  x  d
6 A cd ▪ Phase (A) studies the industrial process, where the production
4
0 for x > d system, the equipment, machinery and products used will be
analyzed in detail.
where a, b, c and d are real numbers and a < b < c < d. If b ¼ c, it is ▪ Phase (B) includes risks identification through HAZOP analysis.
defined as a triangular fuzzy number. ▪ Phase (C) comprises risks analysis (of the identified risks at
By the extension principle, the fuzzy arithmetic operations of phase B) by using fuzzy numbers and linguistic variables,
any two trapezoidal fuzzy numbers follow these operational laws: ▪ Phase (D) consists of risks assessment with the levels initially
defined.
Fuzzy addition : A1 4A2 ¼ ða1 þ a2 ; b1 þ b2 ; c1 þ c2 ; d1 þ d2 Þ
For phase (C), the identification of the risk factor function and of
(2)
the linguistic scales will be needed, and for phase (D), the identi-
fication of the scale for risk assessment will be essential.
 J.L. Fuentes-Bargues et al. / Journal of Loss Prevention in the Process Industries 44 (2016) 44e52 47

3.1. Definition of the Risk factor function

The risk factor (RF) is a function of the risk probability (RP) and
its impact (RI). The risk probability is the probability that the un-
wanted event occurs, and the risk impact is the effect that the
unwanted event or realization of the hazard has on the safety,
environment and other objectives of the project. This relation is
represented by the following equation:

RF ¼ RI$RP (8)

3.2. Definition of the linguistic scales and their associated fuzzy

numbers

According to Chen and Hwang (1992) when the members in a

risk assessment group obtain inexact information about risk asso-
ciated with the project, the assessments cannot be exact but
approximate. In these circumstances, the judgements of the
members in a risk assessment group are expressed by means of
linguistic terms instead of real numers according to the fuzzy
numbers of Table 2 and Fig. 3.
The linguistic terms are:

- For evaluating the Risk Impact (RI) a three-point scale is defined:

High (H), Medium (M) and Low Probability (L).
- For evaluation the Risk Probability (RP) a five-point scale is Fig. 2. HAZOP process. Source: Prepared by authors ISO 31.010:2011 (International
defined: Critical (C), Serious (S), Moderate (Mo), Minor (Mi) and Standard Organization (ISO), 2011).
Negligible (N).

RF i ¼ RP i 5RI i (11)
3.3. Judgement of the RP and RI parameters
where i is each of the identified risks and 5 represents the fuzzy
multiplication.
This phase is divided in two steps:

a) Individual judgement. Using the linguistic terms defined in 3.5. Defuzzification

Table 2, each member in the risk assessment group provides
their judgement of the parameters RP and RI of each identified Defuzzification is the operation of transforming a fuzzy number
risk. These linguistic measures are converted into their corre- into a real number. There are several methods proposed for
sponding fuzzy numbers RPm m
i and RIi , where i is the number of
defuzzification process. In this research, the centroid method
identified risks and m is the number of members in the risk (Yager, 1980) is proposed:
assessment group. Z 1
b) Global judgement. The measures of each member in the risk xRFi ðxÞdðxÞ
assessment group are aggregated into a group fuzzy number by ðRFi ÞT ¼ Z0 (12)
1
using the fuzzy arithmetic average which is defined as:
RFi ðxÞdðxÞ
0
1 Xm
1  
RPi ¼ $ RP m ¼ $ RPi1 4RPi2 4…4RPim (9)
m n¼1 i m

3.6. Risks classification

1 Xm
1  
RIi ¼ $ RIim ¼ $ RIi1 4RIi2 4…4RIim (10)
m n¼1 m The last step of the risk assessment procedure is the classifica-
tion in categories of the risks depending on their risk impact factor.
where i is each of the identified risks, m is the number of risk The defined categories and the risks included are:
assessment group, $ is the scalar multiplication defined in equa-
tions (6) and (7) and 4 is the fuzzy addition defined in equation (2). - If (RFi)T 2[0, 0.1], Ri belongs to category I and is classified as
“Negligible”.
- If (RFi)T 2[0.1, 0.4], Ri belongs to category II and is classified as
3.4. Fuzzy judgement of the risk factor “Acceptable”.
- If (RFi)T 2[0.4, 0.8], Ri belongs to category III and is classified as
Once the parameters RI and RP are expressed in fuzzy numbers, “Non Acceptable”.
the risk factor of each risk is defined as indicated in the following - If (RFi)T 2[0.8, 1], Ri belongs to category IV and is classified as
equation: “Intolerable”.
48 J.L. Fuentes-Bargues et al. / Journal of Loss Prevention in the Process Industries 44 (2016) 44e52

Table 1
HAZOP guide word method. Source: ISO 31010: 2011 (International Standard Organization (ISO), 2011).

Guide word Meaning Example of deviation

NO Absence of the variable to which it applies No flow in line

LESS Quantitative reduction Less flow
MORE Quantitative increase Higher temperature
OTHER Partial or total replacement Presence of impurities
INVERSE Opposite function to design intention Return flow
PART OF Qualitative decline. Only part of what should happen occurs Decrease in the composition in a mixture
IN ADDITION Qualitative increase. More is produced than intended Impurities or an extraordinary phase

Table 2 Table 3
Description of linguistic terms of RI and RP. Source: Chen et al. [53]. Nodes for the HAZOP analysis. Source: Prepared by authors.

General interpretation Fuzzy number Sub-system Node

Description of RI 1 Reception of raw materials (Reception hoppers)

Critical (C) Involved very highly impact (0.8, 0.9, 1, 1) 2 Transport of raw materials (Conveyor belt)
Serious (S) Involved highly impact (0.6, 0.75, 0.75, 0.9) 3 Transport of raw materials (Bucket elevator)
Moderate (Mo) Involved moderate impact (0.3, 0.5, 0.5, 0.7) 4 Storage of raw materials (Silos)
Minor (Mi) Involved only small impact (0.1, 0.25, 0.25, 0.4) 5 Grinding (Hammer mill)
Negligible (N) Involved no substantive impact (0, 0, 0.1, 0.2) 6 Mixing of raw materials (Horizontal paddle mixer)
Description of RP 7 Vapor addition (Conditioner)
High (H) Very likely to occur (0.7, 0.9, 1, 1) 8 Granulation of feed in flour (molding grinding press)
Medium (M) Likely to occur (0.2, 0.5, 0.5, 0.8) 9 Cooling of the granulated feed (Countercurrent vertical cooler)
Low (L) Occurrence is unlikely (0, 0, 0.1, 0.3) 10 Selection of the adequate size of the grain (Filter/Sieve)
11 Dry extrusion (Crusher/Vertical roller mill)
12 Feed bagging (Bagger)

4. Application to a case study

Table 4
The application of the proposed methodology is performed on a Table Summary of Risks Identification. Source: Prepared by authors.
plant dedicated to the manufacture of compound feed in the town
of Silla (Valencia, Spain). Products intended for animal feed are Id. Risk

produced from pork, beef, poultry, etc. Animal feed is a growing 1 Excess of dust during the unloading
sector of great importance for the Spanish economy. In fact, Spain is 2 Excess of speed
3 Dust rising during the storage
the third European producer with 17.5% of production in 2013
4 Excess of raw material at the entrance of the equipment
(European Feed Manufacturer' Federation (FEFAC)). Even though 5 Wrong functioning of the equipment
there are no chemicals or explosives in the raw materials used for 6 Lack of protection against foreign body contamination
the manufacture of compound feed, the use of cereals can generate 7 Reduction of production in the equipment
8 Absence of raw material at the entrance of the equipment
explosive atmospheres, so it is necessary to identify, assess and
9 Inappropriate mixing of raw materials
control the risks of the industrial process. 10 Lack of cleanliness
11 High temperature during the operation of the equipment
4.1. Risks identification through HAZOP analysis 12 High pressure during the operation of the equipment
13 Excessive speed during the sieving

The installation has been divided in twelve nodes that corre-

spond to each of the phases or sub-systems of the compound feed
production process (Table 3). the previous step (Table 4) in order to analyze the probability and
In Table 5, an example of the result of the HAZOP analysis for the impact according to the scales defined in Table 2. The results ob-
node “Grinding” is shown. In Table 4, the main risks of the HAZOP tained are shown in Table 6.
analysis are identified, upon which phase C of the methodology will These linguistic terms, according to the definitions in Table 2
be applied. and Fig. 3, are equivalent to trapezoidal fuzzy numbers, from
which the fuzzy values of the risk factor for each identified risk are
4.2. Risks evaluation through fuzzy numbers and linguistic terms obtained through expression (11) (Table 7).
Finally, the fuzzy number that represents the risk factor variable
Linguistic terms are assigned to each of the identified risks in is transformed into a real number using equation (12) and it is

Fig. 3. Membership functions of RI and RP. Source: Adapted by authors.

 J.L. Fuentes-Bargues et al. / Journal of Loss Prevention in the Process Industries 44 (2016) 44e52 49

Table 5
Example of HAZOP analysis for the node “Grinding”. Source: Prepared by authors.

Node: Grinding Sub-system 5: Hammer mill.

Guide word Variable Deviation Causes Consequences Corrective measures

MORE Flow Excess of raw material. Failure in the flow regulator at Overloading and obstruction in Inspection of the proper
the inlet of the mill. the interior of the chamber of functioning of the valve.
the mill. Installation of a level control
Acceleration in the system at the inlet of the mill.
deterioration of the hammers. Control of the temperature in
Heating of the surfaces. the interior of the mill.
WRONG Movement Oscillating hammers not well Wrong installation of the mill. Fire or explosion. Control of the temperature in
fixed to the rotary shaft. Deterioration due to usage. Friction between hammers and the interior of the grinding
sieve or perforated metal chamber.
screen. Installation of spark detection
Appearance of a great quantity and extinguishing systems.
of sparks. Correct maintenance of the
Heating of the surfaces. equipment.
Deterioration and breaking of Avoid the presence of
the hammers. flammable liquids near the
dryer.
Installation of explosion
protection systems such as
explosion venting and
flameless venting.
Installation of systems that
avoid the explosion
propagation to the rest of the
premises.
Utilization of chemical or
physical barriers like rotary
valves or anti- explosion valves.
LESS Production Reduction in ground product. Obstruction of the openings of Accumulation of ground Maintain depression in the
the sieve. product between the hammers interior of the mills.
Deterioration of the sieve. and the sieve producing a Use of hermetic bearings.
Deterioration of the hammers. continuous friction with the Control of the temperature in
Use of sieve with narrow rotor, which leads to an the interior of the grinding
openings. overheating of the product and chamber.
Grinding of raw materials with generates incandescent Control of the product's level at
excessive moisture. particles that can provoke a fire the exit of the sieve to avoid its
or explosion. obstruction.
Installation of localized
aspiration systems to facilitate
the evacuation and
refrigeration of the product in
the grinding chamber, reducing
the probability of generating
over heating and therefore,
incandescent particles.
Use of spark detection and
extinguishing systems.
Correct maintenance and
cleanliness of the equipment.
Protection of the mills with
systems that mitigate the
effects of an explosion and
systems that prevent the
explosion propagation to the
rest of the premises.
Avoid the presence of
flammable liquids near the
dryer.
Installation of explosion
protection systems such as
explosion venting and
flameless venting.
Utilization of chemical or
physical barriers like rotary
valves or anti- explosion valves.
(continued on next page)
50 J.L. Fuentes-Bargues et al. / Journal of Loss Prevention in the Process Industries 44 (2016) 44e52

Table 5 (continued )

Node: Grinding Sub-system 5: Hammer mill.

Guide word Variable Deviation Causes Consequences Corrective measures

NO Protection Entrance of foreign bodies such There is no sieving system Strong blows to the sieve and Installation of systems that
as metals and stones. previous to the grinding. hammers. prevent the entrance of foreign
Continuous friction producing bodies such as magnets or stone
the deterioration of the separators.
hammers, great quantity of Installation of temperature
sparks, overheating of the sensors in the interior of the
product and the presence of mill, since the product over
incandescent particles that may heating may cause the
provoke afire or explosion. exceeding of the temperature of
ignition in layer and cloud or
accelerate the ignition process.
Use of spark detection and
extinguishing systems.
Avoid the presence of
flammable liquids near the
dryer.
Installation of explosion
protection systems.

Table 6
Linguistic terms of the risk probability and its impact. Source: Prepared by authors.

Risk RI RP

Excess of dust during the unloading Critical (C) High (H)

Excess of speed Serious (S) Medium (M)
Dust rising during the storage Critical (C) High (H)
Excess of raw material at the entrance of the equipment Serious (S) Low (L)
Wrong functioning of the equipment Moderate (Mo) Low (L)
Lack of protection against foreign body contamination Critical (C) Low (L)
Reduction of production in the equipment Critical (C) Medium (M)
Absence of raw material at the entrance of the equipment Serious (S) Low (L)
Inappropriate mixing of raw materials Moderate (Mo) Medium (M)
Lack of cleanliness Critical (C) High (H)
High temperature during the operation of the equipment Critical (C) Medium (M)
High pressure during the operation of the equipment Serious (S) Medium (M)
Excessive speed during the sieving Critical (C) Low (L)

Table 7
Risk factor values associated to each of the risks. Source: Prepared by authors.

Risk RI RP RF

Excess of dust during the unloading (0.8, 0.9, 1, 1) (0.7, 0.9, 1, 1) (0.56, 0.81, 1, 1)
Excess of speed (0.6, 0.75, 0.75, 0.9) (0.2, 0.5, 0.5, 0.8) (0.12, 0.375, 0.375, 0.72)
Dust rising during the storage (0.8, 0.9, 1, 1) (0.7, 0.9, 1, 1) (0.56, 0.81, 1, 1)
Excess of raw material at the entrance of the equipment (0.6, 0.75, 0.75, 0.9) (0, 0, 0.1, 0.3) (0, 0, 0.075, 0.27)
Wrong functioning of the equipment (0.3, 0.5, 0.5, 0.7) (0, 0, 0.1, 0.3) (0, 0, 0.05, 0.21)
Lack of protection against foreign body contamination (0.8, 0.9, 1, 1) (0, 0, 0.1, 0.3) (0, 0, 0.1, 0.3)
Reduction of production in the equipment (0.8, 0.9, 1, 1) (0.2, 0.5, 0.5, 0.8) (0.16, 0.45, 0.5, 0.8)
Absence of raw material at the entrance of the equipment (0.6, 0.75, 0.75, 0.9) (0, 0, 0.1, 0.3) (0, 0, 0.075, 0.27)
Inappropriate mixing of raw materials (0.3, 0.5, 0.5, 0.7) (0.2, 0.5, 0.5, 0.8) (0.06, 0.25, 0.25, 0.56)
Lack of cleanliness (0.8, 0.9, 1, 1) (0.7, 0.9, 1, 1) (0.56, 0.81, 1, 1)
High temperature during the operation of the equipment (0.6, 0.75, 0.75, 0.9) (0.2, 0.5, 0.5, 0.8) (0.16, 0.45, 0.5, 0.8)
High pressure during the operation of the equipment (0.6, 0.75, 0.75, 0.9) (0, 0, 0.1, 0.3) (0.12, 0.375, 0.375, 0.72)
Excessive speed during the sieving (0.6, 0.75, 0.75, 0.9) (0, 0, 0.1, 0.3) (0, 0, 0.1, 0.3)

classified according to the established scale (Table 8). - Implantation of localized aspiration systems to avoid the
Finally, the corrective measures to be introduced are established cleaning of the hoses during the unloading or the storage, since
and prioritized according to the risks classification. Regarding this would increase the risk of fire or explosion.
“Intolerable” risks, the corrective measures are introduced imme- - Revision and installation of earthing systems to avoid static
diately, no matter their cost. Some of the proposed measures for the discharges such as sparks.
case study are:
 J.L. Fuentes-Bargues et al. / Journal of Loss Prevention in the Process Industries 44 (2016) 44e52 51

Table 8 explosive atmospheres and therefore, possible explosions, some

Risk Factor and Classification. Source: Prepared by authors. actions are proposed. On the one hand, the installation of filters and
Risk (Rfi)T Category sieves to avoid the entrance of foreign bodies and the installation of
Excess of dust during the unloading 0.8633 IV
aspiration systems to decrease the concentration of dust in the
Excess of speed 0.3900 III atmosphere. On the other hand, spark detection and extinguishing
Dust rising during the storage 0.8633 IV systems and earthing systems will be installed in order to reduce
Excess of raw material at the entrance of the equipment 0.0283 I the possible occurrence of electrostatic discharges or sparks that
Wrong functioning of the equipment 0.0517 I
may act as ignition sources.
Lack of protection against foreign body contamination 0.0833 I
Reduction of production in the equipment 0.4767 III
Absence of raw material at the entrance of the equipment 0.0700 I
Inappropriate mixing of raw materials 0.2700 II References
Lack of cleanliness 0.8633 IV
High temperature during the operation of the equipment 0.4767 III Tixier, J., Dusserre, G., Salvi, O., Gaston, D., 2002. Review of 62 analysis methodol-
High pressure during the operation of the equipment 0.3900 III ogies of industrial plants. J. Loss Prev. Process Industries 15, 291e303.
Excessive speed during the sieving 0.0833 I Carr, V., Tah, J.H.M., 2001. A fuzzy approach to construction project risk assessment
and analysis: construction project risk management system. Adv. Eng. Softw.
32, 847e857.
Casamirra, M., Castiglia, F., Giardina, M., Lombardo, C., 2009. Safety studies of a
- Use of filters and sieves to avoid the entrance of the majority of hydrogen refuelling station: determination of the occurrence frequency of the
accidental scenarios. Int. J. Hydrogen Energy 34, 5846e5854.
foreign bodies that may provoke dysfunctions in the equipment, Center for Chemical Process Safety (CCPS), 2000. Guidelines for Chemical Process
like friction, provoking sparks, or the presence of incandescent Quantitative Risk Analysis, second ed. AIChE, New York.
particles due to the clash of the raw material or finished feed Chemical Industry Association (CIA), 1977. A Guide to Hazard and Operability
Studies. Imperial Chemical Industries and Chemical Industries Associations Ltd.,
with those foreign bodies. London, UK.
- Prevention of spontaneous combustion through early detection Chen, S.J., Hwang, C.L., 1992. In: Fuzzy Multiple Attribute Decision Making: Methods
of rising temperatures, detection of combustion gases and and Applications. Springer-Verlag.
Cho, H.M., Choia, H.H., Kimb, Y.K., 2002. A risk assessment methodology for
infrared among other methods. incorporating uncertainties using fuzzy concepts. Reliab. Eng. Syst. Saf. 78,
- Installation of spark detection and extinguishing systems to 173e183.
avoid the entrance of incandescent particles in the reception Demichela, M., Camuncoli, G., 2014. Risk based decision making. Discussion on two
methodological milestones. J. Loss Prev. Process Industries 28, 101e108.
hoppers or in the silos.
Dubois, D., Prade, H., 1978. Operations on fuzzy numbers. Int. J. Syst. Sci. 9 (6),
613e626.
Those risks whose risk impact factor is 0.39 have been consid- Dunjo
, J., Fthenakis, V., Vílchez, J.A., Arnaldos, J., 2009. Hazard and operability
(HAZOP) analysis. A literature review. J. Hazard. Mater. 173, 19e32.
ered “Non acceptable” because they are placed in the limit between
European Feed Manufacturer
Federation (FEFAC). Compound Feed Production in the
one category and the other. This guarantees a greater precaution European Union. Disponible en: <http:\www.fefac.eu>. Access: July 2015.
margin regarding possible accidents. Moreover, a series of correc- European Union. Directive 1994/9/CE of the European parliament and the council of
tive measures for the “Non acceptable” risks are proposed, these 23th of March 1994 on the approximation of the laws of the Member states
concerning equipment and protective systems intended for the use in poten-
being progressively implemented according to an action program. tially explosive atmospheres. Official J. Eur. Union, L100, 1e29.
European Union. Directive 1999/92/CE of the European parliament and the council
of 16th of December on minimum requirements for improving the safety and
5. Conclusions health protection of workers potentially at risk from explosive atmospheres.
Official J. Eur. Union, L23, 57e64.
European Union. Directive 2012/18/EU of the European parliament and the council
The used methodology combines HAZOP analysis with risks of 4th of July 2012 on the control of major-accident hazards involving
assessment through linguistic variables and fuzzy numbers. HAZOP dangerous substances, amending and subsequently repealing directive 96/82/
EC. Official J. Eur. Union, L197, 1e37.
analysis allows identifying the risks and the causes that generate
Haimes, Y.Y., 2009. Risk Modelling, Assessment and Management, 3rd. ed. A John
them throughout the entire industrial process. The linguistic vari- Wiley & Sons Inc. Publication.
ables and fuzzy numbers permit risk assessment to be carried out in Høj, N.P., Kro€ger, W., 2002. Risk analyses of transportation on road and railway from
a European perspective. Saf. Sci. 40 (1e4), 337e357.
an intuitive and qualitative way, without need of numeric data of
Huang, D., Chen, T., Wang, M.J.J., 2001. A fuzzy set approach for event tree analysis.
probabilities or consequences like those needed for the application Fuzzy Sets Syst. 118, 153e165.
of techniques such as the fault trees or the event trees. Using fuzzy International Standard Organization (ISO), 2000. ISO 17776:2000 Petroleum and
numbers can be used instead of the conventional risk matrix Natural Gas Industries e Offshore Production Installations e Guidelines on
Tools and Techniques for Hazard Identification and Risk Assessment.
because they allow working in situations with uncertainty and International Standard Organization (ISO). ISO 31.000, 2010. Risk Management.
imprecision where the precise values cannot be obtained or Principles and Guidelines on Implementation.
deterministic values do not defined the whole spectrum of International Standard Organization (ISO). ISO 31.010, 2011. Risk Management. Risk
Assessment Techniques.
possibilities. Kaufmann, A., Gupta, M.M., 1991. Introduction to Fuzzy Arithmetic: Theory and
The methodology generates a classification of the risks in four Application. Van Nostrand Reinhold, New York.
different categories: Intolerable, Non acceptable, Acceptable and Khalil, M., Abdou, M.A., Mansour, M.S., Farag, H.A., Ossman, M.E., 2012. A cascaded
fuzzy-LOPA risk assessment model applied in natural gas industry. J. Loss Prev.
Negligible. If the risk belongs to the “Intolerable” category, the Process Industries 26 (6), 877e882.
corrective measures will be implanted immediately, no matter their Kim, H., Koh, J.S., Kim, Y., Theofanous, T.G., 2005. Risk assessment of membrane type
cost. On the other hand, if the risk belongs to the “Non acceptable” LNG storage tanks in Korea based on Fault Tree Analysis. Korean J. Chem. Eng.
22 (1), 1e8.
category, the company will develop a program for the implantation Kim, E., Lee, K., Kim, J., Lee, Y., Park, J., Moon, I., 2011. Development of Korean
of the proposed measures with the aim of eliminating the described hydrogen fueling station codes through risk analysis. Int. J. Hydrogen Energy 36,
risks or converting them into “Acceptable” risks. 13122e13131.
Kletz, T.A., 1983. HAZOP and HAZAN Notes on the Identification and Assessment of
During the compound feed production process, no dangerous
Hazards. Institution of Chemical Engineers, Rugby, UK.
substances are utilized, but the formation of dust during the Knowlton, R.E., 1981. Hazards and Operability Studies, the Guideword Approach.
transportation of the raw material (in this case, cereals) in different Chemetics International Company, Vancouver.
phases of the industrial process (storage, grinding, etc.) can Lawley, H.G., 1974. Operability studies and hazard analysis. Chem. Eng. Prog. 70,
45e56.
generate explosive atmospheres, making it necessary to minimize Lees, F.P., 1996. Loss Prevention in the Process Industries, 2nd. ed. Butterworths,
the conditions that may lead to an explosion. In order to avoid London and Baton.
52 J.L. Fuentes-Bargues et al. / Journal of Loss Prevention in the Process Industries 44 (2016) 44e52

Loyd, A.A., 2004. A Comparison of Fuzzy Indices with Monte Carlo Simulations for following the AZF disaster. J. Loss Prev. Process Industries 23, 813e823.
Risk Assessment at the Preliminary Stages of Transit Project Planning. Technical Wang, Y.M., Elhag, T.M.S., 2007. A fuzzy group decision making approach for bridge
University of Szczecin, Poland. Master Thesis of Science in Environmental risk assessment. Comput. Industrial Eng. 53, 137e148.
Protection. Wang, D., Zhang, P., Chen, L., 2013. Fuzzy fault tree analysis for fire and explosion of
Markowski, A.S., Mannan, M.S., 2008. Fuzzy risk matrix. J. Hazard. Mater. 159, crude oil tanks. Jorunal Loss Prev. Process Industries 26 (6), 1390e1398.
152e157. Woodruff, J.M., 2005. Consequence and likelihood in risk estimation: a matter of
Markowski, A.S., Mannan, M.S., Kotynia, A., Pawlak, H., 2011. Application of fuzzy balance in UK health and safety risk assessment practice. Saf. Sci. 43 (5e6),
logic to explosion risk assessment. J. Loss Prev. Process Industries 24 (6), 345e353.
780e790. Yager, R.R., 1980. On a general class of fuzzy connectives. Fuzzy Sets Syst. 4,
Nieto-Morote, A., Ruz-Vila, F., 2011. A fuzzy approach to construction Project risk 235e242.
assessment. Int. J. Proj. Manag. 29, 220e231. Yun, G., Rogers, W.J., Mannan, M.S., 2009. Risk assessment of LNG importation
Papazoglou, I.A., Aneziris, O., 2003. Master Logic Diagram: method for hazard and terminals using the Bayesian-LOPA methodology. J. Loss Prev. Process Industries
initiating event identification in process plants. J. Hazard. Mater. 97, 11e30. 22, 91e96.
Papazoglou, I.A., Nivolianitou, Z., Aneziris, O., Christou, M., 1992. Probabilistic safety Zadeh, L.A., 1965. Fuzzy sets. Inf. Control 8, 338e353.
analysis in chemical installation. J. Loss Prev. Process Industries 5 (3), 181e191. Zadeh, L.A., 1975a. The Concept of a Linguistic Variable and its Applications to
Papazoglou, I.A., Aneziris, O., Bonanos, G., Christou, M., 1996. SOCRATES: a Approximate Reasoning. Part I. Information Science, 8, 199-249; Part II. Infor-
computerized toolkit for quantification of the risk from accidental releases of mation Science Vol. 8, Pp 301-357; Part III, Information Science Vol. 9,
toxic and/or flammable substance. International Journal of Environment and pp. 43e80.
Pollution 6 (4e6), 500e533. Zadeh, L.A., 1975b. The concept of a linguistic variable and its applications to
Reniers, G.L.L., Dullaert, W., Ale, B.J.M., Soudan, K., 2005. Developing an external approximate reasoning. Part II. Inf. Sci. 8, 301e357.
domino prevention framework: Hazwin. J. Loss Prev. Process Industries 18, Zadeh, L.A., 1975c. The concept of a linguistic variable and its applications to
127e138. approximate reasoning. Part III. Inf. Sci. 9, 43e80.
Sa'idi, E., Anvaripour, B., Jaderi, F., Nabhani, N., 2014. Fuzzy risk modeling of process Zhang, G., Zou, P.X.W., 2007. Fuzzy analytical hierarchy process risk assessment
operations in the oil and gas refineries. J. Loss Prev. Process Industries 30, approach for joint venture construction projects in China. J. Constr. Eng. Manag.
63e73. 133, 771e779.
Shapiro, A.F., Koissi, M.C., 2015. Risk Assessment Applications of Fuzzy Logic. Ca- Zhao, J., Cui, L., Zhao, L., Qiu, T., Chen, B., 2009. Learning HAZOP expert system by
sualty Actuarial Society, Canadian Institute of Actuaries, Society of Actuaries. case-based reasoning and ontology. Comput. Chem. Eng. 33, 371e378.
Suokas, J., 1988. The role of safety analysis in accident prevention. Accid. Analysis Zhou, J., 2010. SPA-Fuzzy method based real-time risk assessment for major hazard
Prev. 20 (1), 67e85. installations storing flammable gas. Saf. Sci. 48 (6), 819e822.
Taveau, J., 2010. Risk assessment and land-use planning regulations in France