Analyzing Risks in Business Process Models using a Deviational

Technique

Hanane Lhannaoui, Mohammed Issam Kabbaj and Zohra Bakkoury

AMIPS Research Group, Ecole Mohammadia d’Ingénieurs,

MOHAMMED V University, Agdal, Rabat, Morocco

Keywords: Business Process Management, HAZOP, Deviational Analysis, Risk Analysis.

Abstract: In a business environment, many processes are performing ineffectively due to different risks. Several

methods and techniques for the identification and management of risks in business processes have been

proposed. Some of them originate from other domains and have been adapted to the business environment

such as deviational approaches. Nevertheless, there are few examples for potential applications of those

methods in business process redesign. This paper addresses this gap through proposing an approach that

adapts HAZOP (HA-Zard OP-erability), a systematic deviational technique, to business process

environment. We discuss how this method can contribute to the improvement of business process models.

1 INTRODUCTION

In the last few years, organizations have become

more aware of the importance of managing risks.

They are increasingly implementing procedures that

aim to control and manage business risks. This leads

to an extra effort and costs that may sometimes

distract people from doing what they should focus

on: the business.

Therefore, creating and preserving value in

business cannot be assured without bringing the risk

management practices closer to business process

management (BPM) domain. This need had given

birth to the Risk aware business process. In fact, the

Risk-aware business process system is a system that

allows the reasoning about management of risks in

BPM from the design to the post execution of

business process. This integration has many

advantages including the ability to:

 analyze risks and incorporate risk mitigation

strategies in a business process model during

design time (Goluch et al., 2008),

 monitor the emergence of risks and apply

risk mitigation actions during run time

(Conforti et al., 2011),

 Identify risks from logs and other post-

execution artifacts (Jans et al., 2011).

This paper addresses the topic of risk-aware

business process management. It describes an

approach for analyzing risks in business process

during the design time by studying the deviational

behavior of the process. Since risk-aware business

process environment has many similarities with the

safety domain. The method described in this paper is

inspired from this domain. Concretely, it uses

HAZOP a deviational technique from the safety

domain. That would potentially improve the risk

analysis process. Thus, the results of risk analysis

(HAZOP output) can be used to propose a redesign

for the studied process. This is done by

incorporating risk controls into the design since the

analysis is considered earlier in process lifecycle.

This paper is structured as follows: Section 2 covers

related work while Section 3 summarizes some

preliminaries needed for the understanding of the

concepts introduced in the paper. Section 4 describes

EPC-Based HAZOP Analysis which is an adaptation

of HAZOP a traditional method for deviational

analysis. Section 5 discusses how our method can be

used to motivate business process improvement.

Finally, Section 6 concludes the paper.

2 RELATED WORK

The use of risk management techniques in design is

a concept that had been initially adapted in safety

critical systems. Many of those tools and methods,

for example FMEA (Failure Mode and Effects

189

Lhannaoui H., Kabbaj M. and Bakkoury Z..

Analyzing Risks in Business Process Models using a Deviational Technique.

DOI: 10.5220/0005109801890194

In Proceedings of the 9th International Conference on Software Engineering and Applications (ICSOFT-EA-2014), pages 189-194

ISBN: 978-989-758-036-9

 2014 SCITEPRESS (Science and Technology Publications, Lda.)

Analysis), FTA (Fault Tree Analysis) and HAZOP

(Hazardous Operability Analysis), have been used in

various domains such as the military, power plant,

aircraft and space industry. The objective is to

design reliable and safe systems.

In software engineering for example, HAZOP

studies have also been successfully performed on

several kinds of models such as object-oriented

models (Lano et al., 2002), data flow models and

CORE models (McDermid, 1995). HAZOP has

proved that it is useful for software hazard analysis,

and safety-certification authorities recommend its

use during software development. Subsequently,

many adaptations have been proposed such as UML-

BASED HAZOP analysis (Martin-Guillerez et al.,

2010).

The application of risk management techniques

in business environment is a concept that has been

introduced in the past ten years. Most approaches

that have been proposed in the area of risk-aware

business process management tried to address the

issue of risk in business processes at design time

(Suriadi et al., 2012) including design time risk-

analysis.

Suriadi et al. enumerated, in (Suriadi et al.,

2012) risk-Aware BPM approaches that integrate

risk analysis partially or comprehensively. However,

the research efforts in this area are distributed across

various types of risk analysis, including risk

probability analysis, risk impact analysis, risk

identification/discovery analysis, and risk mitigation

analysis (Fenz and Neubauer, 2009), (Mock and

Corvo, ). They lack the technical precision to afford

a convincing design-time risk analysis approach

(Suriadi et al., 2012). In addition, even if the risk

analysis is carried in the design-time, its output is

not exploited to improve the reliability of the

process. In fact, these approaches do not propose

mechanisms that attach risk analysis to the design

quality.

In this paper, we propose an approach of

analyzing risks that can facilitate a design change in

an earlier phase process lifecycle. This objective has

been proposed formerly in (lhannaoui et al., 2013).

Nevertheless, the present paper introduces more

comprehensive version of the risk analysis method.

It also details the concepts related to the risk analysis

output.

3 PRELIMINARIES

In this section, we describe basic concepts that are

used in this paper.

3.1 Event-Driven Process Chain

EPCs (Event-driven Process Chain) are a graphical

business process description language introduced by

Keller, Nuttgens and Scheer in 1992 (Govermatori,

2006). It was developed at the Institute for

Information Systems of the University of Saarland,

Germany, in collaboration with SAP AG. The EPC

is a core part of the ARIS-framework and combines

different views towards the description of

enterprises and information systems in the control

view on the conceptual level (ISO/DIS 31000,

2006).

EPCs describe processes on the level of their

business logic. The name represents the control flow

structure of the process as a chain of events and

functions. Actually, the EPC describes processes by

the use of alternating functions and events as time-

referring state changes. Events and functions are

linked by the control flow as directional edges

(Govermatori, 2006).

An event-driven process chain consists of the

following elements:

 Functions: The basic building blocks.

Functions are active elements used to

describe the tasks or activities of a business

process that needs to be executed.

 Events: Passive elements used to describe

under which circumstances a process (or a

function) works or which state a process (or

a function) results in (like pre- / post-

conditions).

 Logical connectors: They can be used to

connect activities and events. This is the way

how the flow of control is specified. There

are three types of connectors: AND, XOR

(exclusive or) and OR.

The extended EPC includes the elements

described below:

 The Organization Unit or Role is responsible

for performing an activity or function.

 The Information Objects portray input data

serving as the basis for a function, or output

data produced by a function.

 The deliverables represent results (services

or products) functions produce or input

functions require.

In this paper, we will use EPC as a modeling

language for describing business process.

ICSOFT-EA2014-9thInternationalConferenceonSoftwareEngineeringandApplications

190

3.2 HAZOP

The HAZard and Operability (HAZOP) study was

initially developed by the company ICI in 1974 for

chemical developing facilities but has later been

extended to other types of systems and also to

complex operations and to software systems

(Ministry of Defense, 2000). It is typically

conducted by a team consisting of four to eight

persons with a detailed knowledge of the analyzed

system. HAZOP is performed using a set of

guidewords and attributes. It is based on a theory

that assumes risk events are caused by deviations

from design or operating intentions.

Table 1: HAZOP Guidewords.

Guidewords Interpretation

No This is a complete negation of the design

intention. No part of the intention is

achieved and nothing else happens.

More This is a quantitative increase.

Less This is a quantitative increase.

As well as All the design intention is achieved

together with additions.

Part of Only some of the design intention is

achieved.

Reverse The logical opposite of the intention is

achieved.

Other than Complete substitution, where no part of the

original intention is achieved but

something quite different happens.

Early Something happens earlier than expected

relative to clock time.

Late Something happens later than expected

relative to clock time.

Before Something happens before it is expected,

relating to order or sequence.

After Something happens after it is expected,

relating to order or sequence.

The purpose of a HAZOP study is to identify

what potentially hazardous variations from the

design intent could occur in components and in the

interactions between components of a system

(Ministry of Defense, 2000). Consequently, this will

permit us to avoid continuing development of

designs with potential hazards (Ministry of Defense,

2000). The technique uses “guidewords” to promote

creative thinking about the ways in which hazardous

situations might occur. A guideword is used to

express a particular kind of deviation (Table 1).

In this paper, we use HAZOP technique for risk

analysis since it is:

 A qualitative method. In fact, our approach is

based on the study of risk behavior rather

than its appreciation;

 An inductive risk assessment tool, meaning

that it is a “bottom-up” risk identification

approach where we start from a particular

fault to the general effect of the fault;

 Risk-focused. Actually, it concentrates on

how the design will cope with abnormal

conditions rather on how it will perform

under normal conditions;

 Design-oriented because apart from being

used for identifying hazards, it proposes

recommendations with low-level details on

the design.

4 EPC-BASED HAZOP

In this section, we present our method to analyze

risks based on an EPC model representing a business

process description. The risk analysis is then

performed on this description using an adaptation of

the HAZOP method. For this, we consider that we

have an EPC business process model. The design

should be done as early as possible in the

development process to allow early identification of

major risks and to program consequent adaptation of

the design to incorporate risk controls.

4.1 Hazop Method Adaptation

This section aims to propose an adaptation of

HAZOP for EPC-extended models. Actually, a

business process model fragment can be considered

as a HAZOP entity and its EPC elements as HAZOP

attributes. Therefore, once the EPC model is

completed, the EPC-based HAZOP method is

applied by selecting model’s elements and applying

the corresponding guidewords to them.

Since “Function”, is the only active element in

the EPC models, we will only consider guidewords

interpretation for this element.

For this, we simulate the function in an EPC

model to a system in the safety domain. Its

parameters are simulated to the other EPC elements:

passive elements (events) and annotations. Therefore,

we divide those parameters to three groups: the Role,

the input parameters and the output parameters

(Table 2).

Taking this into account, we then propose new

definitions of hazard related to guidewords

deviations which will suit for applying in the

business environment. We call, consequently, EPC-

based HAZOP the re-interpretation of HAZOP

guidewords previously presented in Table 1 in the

context of EPC modeling. Table 4 represents then the

AnalyzingRisksinBusinessProcessModelsusingaDeviationalTechnique

191

proposed definitions of the possible deviations that

can be detected for the described guidewords and the

related parameters.

Table 2: Groups of parameters.

Group Parameters Description

Role The

Organization

Unit or Role

The Organization Unit or

Role is responsible for

performing an activity or

function.

Input Event

(precondition)

Circumstances under

which a function or a

process works.

The Information

Objects

Input data serving as the

basis for a function.

The deliverables Input services or products

that functions require.

Output Event

(postcondition)

The state that a process or

a function results in.

The Information

Objects

Output data produced by a

function.

The deliverables The deliverables represent

results (services or

products) functions

produce.

Table 3: Correlation of guidewords and parameters.

GW Event Role Deliverables

Information

objects

Precon

dition

Postco

ndition

Inp

Out

put

Inp

Output

No + + + + + + +

Less + + + +

More + + + +

Part

+ + + + + +

well

+ + + +

Reve

rse

Other

than

+ + + + +

Once deviations have been identified, possible

consequences and causes are analyzed. We note that

some deviations can be caused by the deviation of

other categories of elements, such as events, the

organization role or unit, the deliverables and the

information objects. In those cases, the overall

deviation is associated to the function itself.

Consequently, Table 3 represents the mapping

between guidewords and the other process elements.

The final outcome of EPC-Based HAZOP

analysis consists of a list of recommendations and a

list of hazards, together with the possible deviations

leading to them. This list of hazards can be converted

to a list of risks. HAZOP inherently assumes that risk

events are caused by deviations from design or

operating intentions. In the next section, we propose

a new version of the HAZOP output in order to take

into account all the variables that are related to

business process risks.

4.2 EPC-based HAZOP Analysis Table

In this section, we will introduce EPC-based HAZOP

table, an adapted version of HAZOP table (the output

of HAZOP) that meets the purpose of our work. In

the literature, Kletz defines, in (Suriadi et al., 2012),

5 principal columns in HAZOP table: guideword,

deviation, possible causes, consequences action and

severity, which represents a preliminary risk

estimation of the impact of the deviation’s

consequences. and action required. Other columns

can be added to the HAZOP table such as safeguards,

comments, responsible for

In this paper, we use the term “recommendation”

that relates to the new security requirement or the

actions required to deal with the related deviation.

The recommendation represents changes that should

be applied to control risks.

We also introduce the severity and frequency

which represent respectively, the deviation impact

and the occurrence probability of the harm due to it.

Those two columns represent the risk valuation.

For the consequences column, generally two

levels are represented in the HAZOP table: The use

case effect which represents the consequences of the

deviation on the HAZOP element (the attribute) and

the system effect which associates the deviation to

the whole process and gives its effect in the real

world.

To sum up, we propose a deviation analysis table

with the following columns (c.f. Table 5):

 Attribute: the EPC element on which the

deviation is applied.

 GW: the applied guideword.

 Description: the deviation resulting from the

combination of the attribute and the

guideword.

 Composite: if the deviation is resulted for an

external reason or from the deviation of

other parameters.

 Possible Causes: possible causes of the

deviation that can be resulted from the

deviation of any of the parameters

predefined.

 Use Case Effect: effect at the use case level.

 System Effect: possible effect in the real

world.

 Severity: rating of effect of the worst case

scenario.

ICSOFT-EA2014-9thInternationalConferenceonSoftwareEngineeringandApplications

192

Table 4: List of deviations.

Element Guideword Interpretation Related parameters

Function

Not

The function is not executed and the output

is not generated