المحاضرة في ادناه
RISK ANALYSIS METHODOLOGIES, TOOLS AND TECHNIQUES
In this appendix, a summary of qualitative and quantitative risk analysis methodologies proposed by PMI (2008) is overviewed with a number of additional common risk analysis tools and techniques that are not applied in this research and presented only as definitions to be a useful reference for other researchers.
1) Risk Analysis Methodologies
1.1) Qualitative Risk Analysis
PMI (2008: pp.290-294) presented a structured methodology for qualitative risk assessment and analysis, which can be summarized as follows:
Qualitative Risk Analysis: Inputs
1) Risk Register
2) Risk Management Plan
3) Project Scope Statement
4) Organizational Process Assets.
The organizational process assets include, but are not limited to:
• Information on prior, similar completed projects,
• Studies of similar projects by risk specialists, and
• Risk databases that may be available from industry or proprietary sources.
Qualitative Risk Analysis: Tools and Techniques
1) Risk Probability and Impact Assessment for each identified risk.
2) Probability and Impact Matrix
3) Risk Categorization
4) Risk Urgency Assessment
5) Expert Judgment
Qualitative Risk Analysis: Outputs
The main output of the qualitative risk analysis process is Risk Register Updates.
1.2) Quantitative Risk Analysis
PMI (2008: pp.295-300) presented a structured methodology for quantitative risk assessment and analysis, which can be summarized as follows:
Quantitative Risk Analysis: Inputs
1) Risk Register
2) Risk Management Plan
3) Cost Management Plan
4) Schedule Management Plan
5) Organizational Process Assets
The organizational process assets include, but are not limited to:
• Information on prior, similar completed projects,
• Studies of similar projects by risk specialists, and
• Risk databases that may be available from industry or proprietary sources.
Quantitative Risk Analysis: Tools and Techniques
1) Data Gathering and Representation Techniques
• Interviewing techniques draw on experience and historical data to quantify the probability and impact of risks on project objectives.
• Probability distributions.
2) Quantitative Risk Analysis and Modelling Techniques
3) Commonly used techniques include both event-oriented and projectoriented analysis approaches including:
• Sensitivity analysis.
• Expected monetary value (EMV) analysis. Expected monetary value analysis is a statistical concept that calculates the average outcome when the future includes scenarios that may or may not happen. EMV for a project is calculated by multiplying the value of each possible outcome by its probability of occurrence and adding the products together.
• Modelling and Simulation. A project simulation uses a model that translates the specified detailed uncertainties of the project into their potential impact on project objectives. Iterative simulations are typically performed using the Monte Carlo technique. In a simulation, the project model is computed many times (iterated), with the input values (e.g., cost estimates or activity durations) chosen at random for each iteration from the probability distributions of these variables. A probability distribution (e.g., total cost or completion date) is calculated
from the iterations. For a cost risk analysis, a simulation uses cost estimates. For a schedule risk analysis, the schedule network diagram and duration estimates are used.
4) Expert Judgment
Perform Quantitative Risk Analysis: Outputs The main output of the qualitative risk analysis process is Risk Register Updates. The risk register is further updated to include a quantitative risk report detailing quantitative approaches, outputs, and recommendations.
2) Additional Tools and Techniques
The following group of tools and techniques are explored by the
researcher because they were used in risk management literature. The researcher collected some details about them, but since they are not used in this research, only brief definitions will be overviewed in this appendix as a quick reference for other researchers.
2.1) Algorithms
An algorithm contains a sequence of instructions for problem solving. These are the steps in a task and the responses determine the route to be followed. Algorithms have often been used as a prelude to computer programs as they are logical and easy to follow [Flanagan & Norman, 1999: p.76].
2.2) Fuzzy Systems
The concepts of fuzzy theory and fuzzy systems were proposed by Zadeh1. Generally, a fuzzy system is any system whose variables (or at least some of them) range over states that are fuzzy sets. Fuzzy sets here are fuzzy numbers,
and the associated variables are linguistic variables. Representing states of variables using fuzzy sets is a way of quantifying the variables. The method
allows translating the subjective judgment given in linguistic expressions (i.e.,
"low", "high", etc.) into mathematical measures [Lin et al., 2011: p.386 and
Rezakhani, 2012: p.18 & p.20].
A general fuzzy controller consists of four modules: a fuzzy rule base, a fuzzy inference engine, and fuzzification/ defuzzification modules. The input and output variables of each fuzzy controller are all linguistic variables and the state of each variable is a linguistic expression of each expert opinion from a macroscopic point of view. One advantage of using linguistic variables is that such expressions are more intuitive and make it easier for experts to give their opinions in ambiguous and complex situations in which numerical estimations are hard to obtain [Lin et al., 2011: pp.386-387].
2.3) Markov Chains
Markov chains have the special property that probabilities involving how the process will evolve in the future depend only on the present state of the 1 Zadah, L. A. (1965) Fuzzy sets and systems, System Theory, Polytechnic Press, Brooklyn, NY, pp. 29-37; and Zadeh, L. A. (1965) Fuzzy sets, Information and Control, process, and so are independent of events in the past. Many processes fit this description, so Markov chains provide an especially important kind of probability model [Hillier & Liberman, 2001: p.802].
2.4) Bayesian Networks
Bayesian Theory: Using the theory of probability developed by Bayes, we can examine the sensitivity of decisions. Bayes developed a theory of prior and posterior probabilities: prior probabilities can be revised in the light of additional information to form posterior probabilities. Hence, Bayesian theory is used to incorporate new information into the analysis [Flanagan & Norman, 1999: p.85] (see the example in the same reference (pp.85-87)).
Bayesian Networks are graphical representations of knowledge for reasoning under uncertainty. They can be used at any stage of a risk analysis, and may substitute both fault trees and event trees in logical tree analysis.
While common cause or more general dependency phenomena pose significant complications in classical fault tree analysis, this is not the case
with Bayesian Networks. They are in fact designed to facilitate the modelling
of such dependencies [Sousa & Einstein, 2011: p.87].
2.5) Nominal Group Technique [URL1]
The nominal group technique (NGT) is a group problem solving process involving problem identification, solution generation, and decision making1.
It can be used in groups of many sizes, who want to make their decision quickly, as by a vote, but want everyone's opinions taken into account (as
opposed to traditional voting, where only the largest group is considered)2.
1 Delbecq A. L. and VandeVen A. H, (1971). "A Group Process Model for Problem Identification and Program Planning," Journal Of Applied Behavioral Science VII
(July/August, 1971), 466 -91.
2 Dunnette M D., Campbell J. D, and Jaastad K., (1963). "The Effect of Group
Participation on Brainstorming Effectiveness for Two Industrial Samples, Journal of Applied Psychology, XLVII (February), 30-37.
The method of tallying is the difference. First, every member of the group gives their view of the solution, with a short explanation. Then, duplicate solutions are eliminated from the list of all solutions, and the members proceed to rank the solutions, 1st, 2nd, 3rd, 4th, and so on.
Some facilitators will encourage the sharing and discussion of reasons for the choices made by each group member, thereby identifying common ground, and a plurality of ideas and approaches. This diversity often allows the creation of a hybrid idea (combining parts of two or more ideas), often found to be even better than those ideas being initially considered.
2.6) Delphi Method
The Delphi method is an established technique for obtaining consensus estimates from several experts, and this technique can be applied to the assessment of risks. The general procedure for this technique is that an estimate of the variable(s), or risks, is obtained from each of the experts. This
estimate can relate to the probability of occurrence or the likely impact of a variable. The experts are then informed of all the estimates and asked to give a revised estimate. This process continues until a consensus estimate is produced. This method can be viewed as a qualitative or quantitative technique, since the experts may or may not be asked to provide a quantitative
estimate of a variable [Smith et al., 2006: p.97].
2.7) Cause and Effect Diagrams
Cause and effect diagrams, also called Ishikawa diagrams or fishbone diagrams, illustrate how various factors might be linked to potential problems or effects. A possible root cause can be uncovered by continuing to ask "why" or "how" along one of the lines. "Why-Why" and "How-How" diagrams may
be used in root cause analysis. Cause and effect diagrams are also used in risk
analysis [PMI, 2008: p.208].
2.8) How-How and Why-Why Diagrams
How-How Diagram
A How-How Diagram is a Tree Diagram where each child is determined by asking "how?" the parent can be achieved. It is thus useful for creating solutions to problems [URL2].
The How-How diagram uses cards that can be Post-it Notes, Index cards or boxes on a computer application such as PowerPoint. When working with a group, a wall area on which to stick up notes will be needed with a large sheet of paper or a whiteboard on which to draw. The steps are [URL3]:
• State problem.
• Ask "How can this be done?".
• Repeat and conclude.
Why-Why Diagram
A Why-Why Diagram is a Tree Diagram where each child statement is determined simply by asking "why?" the parent occurs. It is thus very similar in use to a Cause-Effect Diagram, and techniques may be borrowed from Cause-Effect Diagram usage. Its simplicity can make it useful in less formal
situations [URL4].
The "Why-Why" and "How-How" Diagrams together make a very simple toolset for finding causes of and solutions to problems.
2.9) Influence Diagrams
The influence diagram is a simple visual representation of a decision problem. Influence diagrams offer an intuitive way to identify and display the essential elements, including decisions, uncertainties, and objectives, and how they influence each other [Marks, 2004: p.1].
The Influence diagram method was formalised as a methodology in the early eighties. An influence diagram is simply a technique involves mapping out the project, identifying the sources of risk and possible responses to these
risks. This information is then represented in a diagram, which consists of nodes reflecting variables and decisions, and influences are reflected by arrows [Smith et al., 2006: p.98 & p.106].
An influence diagram contains at least three elements, linked with arrows to show the specific relationships among them. Influence diagrams typically flow from decisions to uncertainties to value. Arrows indicate relevance and
show relationships [Marks, 2004: p. 2].
Combining influence diagrams and the Monte-Carlo technique results in a very powerful tool for risk analysis. The influence diagram method is a very flexible way of building the risk model, and it allows to add all risks to the risk model, not only those that affect time and cost estimates [Smith et al., 2006: p.107].
2.10) Decision Matrix
A decision matrix is a representation of the options that are open to the decision-maker, the factors that are relevant, and the outcomes [Flanagan & Norman, 1999: p.80].
2.11) Decision Trees
Decision trees, also known as decision networks, are diagrams that depict a sequence of decisions and chance events, as they are understood by the decision-maker. The decision tree is made up of two types of nodes, decision nodes and chance event nodes. A decision node represents a decision that has to be made and a chance event node represents an event that has a chance of occurring, possibly a risk. Each decision node should have at least one branch, or arrow, coming from it and these branches represent the decision alternatives [Smith et al., 2006: p.99].
The main advantage of using a decision tree, whether it used as a qualitative or quantitative technique, is that it requires the entire project to be set out in a logical sequence. This ensures that the decision-maker has considered all the options available in the project at an early stage. This technique does not identify the best alternative or course of action to be taken;
it merely sets out all the possible alternatives [Smith et al., 2006: p.100].
Stochastic decision tree analysis: The stochastic decision tree approach, originally developed by Hespos and Strassman (1965), provides another
method for analysing decision problems over time. It combines the logic of the decision tree analysis just discussed with the Monte-Carlo simulation approach used in risk analysis [Flanagan & Norman, 1999: p.87].
Decision Trees vs. Influence Diagrams Decision Trees structure the timing of the decisions and the revelation of the uncertainties, while Influence Diagrams structure the influence: on decisions, on values, on uncertainties, as well as structuring the timing [Marks, 2004: p.20].
2.12) Fault Tree Analysis
Fault Tree Analysis is a widely used method in industry to perform reliability analyses of engineering systems. The method was developed in the early 1960s at the Bell Laboratories to evaluate the reliability of the Minuteman Launch Control System1. The fault tree analysis begins by
identifying an undesirable event, called the top event, associated with a
system or problem. The events that could cause the occurrence of this top event are generated and connected by logic gates called OR, AND, and so on. A fault tree's construction proceeds by generating fault events in a successive manner until no further fault events need to be developed. The method is
called fault tree analysis because it only considers failure or negative events, and a fault tree may simply be described as the logic structure relating the top
event to the basic fault events [Dhillon, 2002: pp.63-65].
1 Dhillon, B. S., and C. Singh (1981) Engineering Reliability: New Techniques and Applications, New York: John Wiley and Sons.
E-Appendix T4
(E-App T4 Pg 10)
2.13) Bow-Tie Diagram [URL5]
The bowtie diagram provides a powerful graphical representation of the risk assessment process which is readily understood by the non-specialist.
The idea of the diagram is combining the cause (fault tree) and the consequence (event tree). When the fault tree is drawn on the left hand side and the event tree is drawn on the right hand side with the hazard drawn as a "knot" in the middle, the diagram looks a bit like a bowtie, as shown in the
sample figure below.
The exact starting point of the Bowtie Methodology has been lost in time but it is believed that they were originally called "Butterfly diagrams" and
evolved from the Cause Consequence Diagram of the 1970s.
It is then thought that David Gill of ICI plc developed the methodology and called them bowties in the late 70's. It is generally accepted that the earliest mention of the bowtie methodology appears in the ICI Hazan Course Notes 1979, presented by The University of Queensland, Australia.
The technique was given a huge boost in the early to mid 90’s when the Royal Dutch/Shell Group developed the technique as a result of the Piper Alpha disaster.
The structured approach of the bowtie methodology was particularly popular in risk analysis within safety cases where quantification is not possible or desirable.
Bow-Tie Diagram [Book, 2007: p.7]
ليست هناك تعليقات:
إرسال تعليق