Integrative modeling of performance deterioration and maintenance effectiveness for infrastructure assets with missing condition data

Deterioration modeling is an important analytical component in infrastructure asset management. It concerns the prediction of performance and remaining service life of assets of different designs under omnifarious working environments. For long‐term prediction, it also requires to characterize maintenance effectiveness because maintenance activities do not necessarily bring an asset to a completely renewed status. Deterioration modeling research has for decades been largely focusing on the modeling of the natural deterioration process per se, whereas the modeling of maintenance effectiveness is only a recent focus of investigation, mainly in pavement research. In practice, the asset conditions immediately before and after a given maintenance treatment both are not often known. This has made the modeling of maintenance effectiveness and long‐term deterioration prediction a challenging task. To bridge the gap, this paper presents a novel approach that integrates the modeling of deterioration and maintenance effectiveness into one process. The natural deterioration of asset performance is modeled as a continuous‐time Markov chain, whereas the effectiveness of a maintenance measure is modeled as a discrete‐time Markov chain. To account for missing condition data before and after the maintenance event, the paper also develops a robust statistical method based on Markov chain Monte Carlo simulation. A real‐life case study on a municipal sewer pipe system is carried out for demonstration of the proposed integrated modeling approach. The functional deterioration of sewer pipes and the effectiveness of flushing operations that target to bring flow capacity to intact state are modeled. Influences of pipe length, diameter, slope, and sewershed area are examined. The present work is a valuable step toward development of evidence‐based risk‐informed asset management framework.