Risk simulation in a portfolio of port and rail capital projects

Abstract

There are some advantages of using quantitative risk assessment methods over more traditional qualitative risk assessment methods. The cost and schedule impacts of project risks can be better described when using quantitative methods. This in turn allows contingency calculations to be more scientific than when using more traditional methods. In many cases, these quantitative risk registers are standalone MS Excel based entities. This represents a problem in that it is difficult and impractical to use these separate risk registers to do a concurrent Monte Carlo simulation. This thesis therefore presents a model which uses the Monte Carlo method to quantify certain risk and project categories in a portfolio of 86 port and rail capital projects. The purpose of the model is to provide a portfolio-wide view of risks to answer the questions What matters most? and Where should the focus be regarding risk treatment plans? . The answers to these questions should then be used to identify policies and procedures which need to be changed to improve the project delivery and execution process. The model was based on the principles of the ISO31000:2009 risk management process, MS Excel spreadsheets and @Risk simulation software to generate output distributions which are ranked using various methodologies. The risk and project categories which were used in the model included the following: Project type: Each of the 86 projects in the project portfolio was assigned to one of 15 different project categories. The initial expectation was that certain risk names in the project portfolio would cause the most uncertainty. Risk type: This refers to the control the project owner and the project team have over influencing the likelihood and consequences which are associated with specific risks. Five different types were used: External uncontrollable, External Influencable, Internal Owner Requirement, Internal Operational and Internal Project Processes. Risk name: A total of 165 risk names were used to describe 1063 different risks which belonged to the 86 projects. Project start delays: Certain risks delay the execution start of projects and therefore caused the escalation of project cost due to inflation. Risks associated with programmes: The model classified each risk in terms of three types defined by Aritua (2011, 311): Generic Project Risks, risks which are Amplified in Programmes and risks which are Common to Programmes. The initial assumption that certain risk names drive uncertainty in the project portfolio was disproved using the unique risk simulation approach developed in this thesis. It was also shown in a unique manner, using various risk categories, that uncertainty in the project portfolio was driven by eight large, complex, multistakeholder projects. The next risk category which caused the most uncertainty was controllable risks, followed by start delay risks, planning risks and lastly policy related risks. The main contributions of the thesis are identified as: Amount and quality of the unique data which was gathered for this research. Limited information was available regarding risk simulation in a portfolio or program of projects, especially for a large, complex portfolio. A total of 165 different risk names were identified during the research. Each of the risks were assigned to various risk categories in a unique manner as part of a detailed risk analysis to determine What Matters most/ and Where to focus? . The way in which the simulation model and accompanied framework was developed. The literature review identified a gap in how simulation models related to the ranking of risks in portfolios of projects can be developed and which questions to ask during the risk analysis process. This gap was filled by a detailed description of how such a model can be built and how risk aggregation can take place in a project portfolio, using unique combinations of functions in spreadsheets and risk simulation software standards MS Excel and @Risk functions.