Modelling railway bridge asset management. (November 2013)

• Record Type: Journal Article
• Authors: Le, Bryant; Andrews, John
• Other Names: Iwnicki Simon guest-editor.; Dennis Colin guest-editor.
• Is Part Of: Proceedings of the Institution of Mechanical Engineers. Volume 227:Number 6(2013)
• Page Start: 644
• Page End: 656
• Publication Date: 2013-11
• Subjects: Bridges -- asset management -- maintenance modelling -- whole life costs -- Markov analysis; Railroads -- Periodicals; Personal rapid transit -- Periodicals; 625.1
• DOI: 10.1177/0954409713490924
• Languages: English
• ISSNs: 0954-4097

Abstract

Railway bridges form one of the major railway asset groups with more than 35, 000 bridges on the UK rail network. Additionally, the bridge structures are old with more than 50% of the population constructed over 100 years ago. Due to the unique nature of each bridge and their varied means of construction, the decision as to what type of maintenance actions should be performed and when to perform them is a complex problem. Models can be formulated to predict the future condition of assets along with the effect that interventions such as servicing, repair and element replacement will produce. This can be used to support this decision making process. This paper demonstrates a Markov modelling approach to predict the condition of individual bridge elements. For each bridge element the degradation process is determined by examining the maintenance records and analysing the times that each element takes to deteriorate to the point where maintenance of a certain severity classification is required. By combining the elemental models, an overall bridge model is formed which can be used to investigate different maintenance strategies. The model is capable of accounting for a bridge's current condition, material, route criticality, structural arrangement and environment. The maintenance, opportunistic maintenance and renewal strategy can also be varied in the model along with the service frequency, inspection frequency and repair delay time. Using the model the whole life costs can be predicted for any of the selected maintenance strategies.