Abstract:
This study adopted a mixed research design in order to derive a risk-minimizing preventive maintenance strategy for structural roofing leakage problems. The purpose of this study is to: (a) determine current waterproofing systems and components of concrete roofs at airports, (b) explore frequently used techniques to design optimal preventive maintenance policies, and (c) derive an optimal preventive multi-objective risk-based maintenance strategy via simulation that can diminish failure costs, thereby decreasing the system’s overall maintenance costs. Unique to the simulation presented in this thesis is the inclusion of external influences on failure probability via a predictive model derived using a random forest machine learning strategy. A case study on a prototype airport serves as the platform to test this simulation methodology. The results of the simulation reveal the prognostic influence of several maintenance strategies on the total risk-based cost of the waterproofing system. These outcomes showed a negative impact on the reduction of the failure-risk costs, and eventually augmented the overall cost. Thus, demonstrating that for assets whose failure is primarily dictated by external factors, a corrective maintenance strategy can be optimal.
