Fuzzy Set-based Risk Management for Construction Projects

Efficient and comprehensive risk management is critical for successful delivery of engineering, procurement, and construction management (EPCM) projects. Complexity of construction projects is on the rise, which makes it necessary to model uncertainties and to manage risk items related to this class of projects. For decades, researchers and construction practitioners worked together to introduce methods for risk identification and assessment. Considerably less effort was directed towards the development of methods for mitigation, monitoring, and control. The respective individual limitations of these methods prevent the development of comprehensive model which satisfies the needs of practitioners. In this research a comprehensive risk management model “CRMM” is developed to address the limitations of existing methods and to fill the gap between research and practice. The developed model implements a micro system approach to introduce a novel risk identification methodology that provides a systematic procedure to identify risk associated with construction projects. The identification procedure implements root cause analysis and brainstorming technique to identify risk items, consequences, and root causes. The developed CRMM also introduces new method for determination of risk ownership utilizing fuzzy set theory and “One Risk – One Owner” concept. The ownership determination method allocates risk to the owner with highest ability, effectiveness, and capacity to deal with that risk. It also introduces a new qualitative and quantitative evaluation process that utilizes fuzzy set theory and fuzzy probability theory, as well as a new risk mapping procedure which allows for the determination of risk level associated with any project component (e.g., category). The quantitative assessment methodology allows for the use of linguistic and numeric fuzzy evaluations. Fuzzy Linguistic-Numeric Conversion Scheme (FLNCS) is introduced to convert the linguistic evaluations into numeric. The quantitative assessment methodology also introduces the pre-mitigation contingency that represents the contingency fund required for a risk in case no mitigation strategy is implemented. In this respect a novel risk mitigation framework is developed to generate and evaluate possible mitigation strategies for each risk being considered. It also provides a selection procedure which allows users to select the most effective mitigation strategy; making use of fuzzy set theory. The mitigation methodology introduces the post-mitigation contingency that quantifies the contingency required for the selected mitigation strategy. Performance of selected mitigation strategy is monitored using a newly developed risk monitoring method that compares the actually depleted contingency to the post mitigation contingency. The developed monitoring method provides an early warning that alerts users of detected possible failure of selected mitigation strategy. It also determines the correct time for initiation of control process based on a set of qualitative factors. Once risk control process is initiated, the developed control method identifies, evaluates, and selects the most effective control action(s) to support the selected mitigation strategy. In cases where the selected control action fails, the developed control method notifies the user to revise the risk management plan. These notifications allows user to avoid potential failures of similar risk items which are expected to occur in the future. The developed CRMM was coded using VB.Net under Microsoft® windows and .NET framework environment to facilitate its application. A set of case studies are collected from literature and analysed to validate the developed methods within CRMM and to illustrate their essential features. Also, a numerical example elucidates the complete computational processes of the developed comprehensive model.