Conceptual Quantitative Model to Group Risks in Fast-Track Construction Projects

Overlapping of activities on fast-track projects generates unique risks which can be grouped using a new quantitative model. The model employs clustering analysis to classify the risks according to the risk categories and activity evolution, sensitivity, production reliability, and constraints. A hierarchical method is first applied using the centroid method and square Euclidean distance to define the number of clusters. Then, we apply a nonhierarchical method using K-means and a convergence criterion of zero to generate the final groups. A base case with data collected through a case study using a commercial construction project illustrates the clustering approach. Results showed that the risks could be combined into ten clusters. Although the results will vary by project, the model can be applied to group risks of any fast-track construction project and support the definition of mitigation actions and allocation of responsibilities.