A data-driven approach to identify-quantify-analyse construction risk for Hong Kong NEC projects

• Main Authors: Ming-Fung Francis Siu, Wing-Yan Jacqueline Leung, Wai-Ming Daniel Chan
• Format: Article
• Language: English
• Published: Vilnius Gediminas Technical University 2018-12-01
• Series: Journal of Civil Engineering and Management
• Subjects: risk identification; risk quantification; risk analysis; risk register; risk category; risk rating
• Online Access: https://journals.vgtu.lt/index.php/JCEM/article/view/6483

Project risks must be managed to deliver construction projects on time and within budget. In recent years, the New Engineering Contract (NEC) provides an alternate contracting method for procuring construction projects. As stipulated in the NEC contract, NEC risk register must be used to record any project risks. The risk register is designed to record each risk item in the context of textual description, likelihood, and consequence. However, it is time-consuming to identify, quantify, and analyse NEC project risks based on experience, questionnaire, simulation, and data-mining approach. Any method to fully utilise the records of NEC risk registers of past projects for managing NEC project risks remains unexplored. As such, a data-driven approach is proposed to categorise common risks of NEC projects and to analyse risk rating of risk categories by combining the use of text mining analysis and decision tree analysis. A practical case study in Hong Kong is used to illustrate the method of application. Top four common types of NEC project risks, which are ground and utilities, design information, structures, and workmanship, were identified, quantified, and analysed. The new approach helps NEC project planners to identify, quantify, and analyse NEC project risks time-efficiently.