Asset Deterioration: Determining Probabilistic Maintenance Intervals

The municipality of Amsterdam is facing the challenge to deal with uncertainties accompanied with the maintenance planning of maintenance activities on bridges. Long term asset planning (LTAP) has the goal to show the proposed expenditures, capital and operational expenditures (CAPEX and OPEX), for the long term to be able to build up a monetary buffer to absorb the CAPEX which is necessary when a period of large expenditures due to multiple major overhauls and replacements comes up. LTAP also gives the opportunity to program the more expensive activities which create the advantage of spreading expenses over a longer term. The Long Term Asset Planning of Amsterdam makes use of a deterministic time value for activities on bridges, while in fact this value does not give a reliable image of the actual situation. The majority of maintenance activities seems to be performed after the deterministic time value has been reached. Thus, the long term asset planning copes with uncertainties which results in an unreliable outcome and prospect in the expenditures for maintenance. Uncertainties can partly be solved by the use of deterioration models. The parameters involved that predict the physical behavior are mostly an assumption and not known with certainty. Data can be derived from historical maintenance data. A problem faced by all municipalities in the Netherlands is the absence of reliable registered quantitative data. Even the presence of past data would not guarantee its future reliability due to the dynamic characteristic of an urban environment. A method has to be found to retrieve the necessary data alternatively to improve the reliability concerning maintenance intervals for bridges in order to improve the reliability of long term asset planning for municipalities like Amsterdam. Expert judgement is a method that has proven to be a mature and scientific tool for achieving the necessary data that is unavailable at this moment and will be used for this research to develop probabilistic intervals for maintenance activities. By qualifying and quantifying the uncertainties regarding the maintenance on bridges, improvements may be realized on the short- and long term for the costs in maintenance planning. A questionnaire reveals that experts in the field have trouble assessing condition over time. A recommendation is given to solve this issue in the future.Civil Engineering | Construction Management and Engineerin