Bridge service life estimation considering inspection reliability

Bridges begin to deteriorate as soon as they are put into service. Effective bridge management requires sound understanding of the deterioration mechanism as well as the expected service life. Decision makers design and execute programs that safely maintain or extend bridge service life at low cost. Key maintenance practices include inspections, repairs, and rehabilitation, among which inspections may be the most important since the other maintenance interventions are conducted based on the inspection report. This paper presents a methodology for determining the expected service life of a bridge or a bridge network based on a deterministic condition-based model associated with inspection quality. This study used almost 10-year condition rating results of bridges and developed a non-linear regression model that takes into account the Inspection Reliability Index (IRI). The IRI represents the relative inspection accuracy with respect to a tolerable error of one grade by comparing condition ratings between routine and in-depth safety inspections. In the evaluation of the IRI, significant variability in condition ratings between routine and in-depth safety inspections was observed. Specifically, approximately 42% of routine inspection results deviated within ± 1.0 grade with respect to results from in-depth safety inspections. A modified non-linear regression model combined with the IRI is proposed to predict the remaining service life of a bridge. The proposed model can estimate the service life of a bridge by combining the bridge age and its condition rating.