Predictive Maintenance of Stormwater Infrastructure Using Internet-of-Things Technology

Abstract

Operations and maintenance (O&M) practices are becoming increasingly scrutinized as storm events become more frequent and intense, surpassing the design storms that dictated the original capacity. In order to get the best performance from infrastructure investments, communities are looking to establish effective operations and maintenance programs. In this study, an approach to assessing system vulnerability was developed using real-time sensors. By leveraging an adapted risk-based assessment model (RBAM), an alert rating system was developed to prioritize maintenance of ponded stormwater facilities. The degradation in performance of stormwater systems was progressively monitored using indicators for probability and consequence of failure. Failure in this context is a stormwater facility’s vulnerability to meet the design performance due to decreased capacity. Probability of failure was rated using a hydrologic model calibration statistic to assess the difference between the observed data set and a drawdown model. Consequence of failure was rated by evaluating the duration of drawdown periods based on regional regulatory criteria and operator schedules. The predictive maintenance alert methodology enables management of distributed assets more effectively through maintenance alerting and builds resilience into stormwater networks because issues are identified early and program resources are optimized.