Methodology for determination of financial needs of gradually deteriorating bridges with only structure level data

Bridge managers, as other infrastructure managers, are increasingly required to justify the financial resources they need to perform preventative maintenance on bridges. This requires a demonstration of the current state of their bridges and the consequences if the preventative maintenance is not performed. The consequences can be expressed in terms of additional costs to the owner if unacceptable levels of failure risk are avoided. These demonstrations are complicated by the fact that many managers have only basic data with respect to the condition of their bridges, how deterioration processes will affect their bridges and on the type, cost and effectiveness of the interventions that will need to be performed on their bridges in the future. This, in many cases, makes it difficult to show that optimal intervention strategies are being followed and results in a less than convincing argument for the use of taxpayers' money. When less than the required funding is received, managers are forced to follow less than optimal intervention strategies, i.e. short term savings are exchanged for, often unknown, increased future costs. In this paper a general methodology is given that is currently being used in practice in Switzerland and France to determine the lowest cost intervention strategies of bridges with respect to gradual deterioration processes, their financial needs and the consequences, in terms of condition, if they are not followed. It is used when either sufficient detailed element level information or tools are not available to conduct such an analysis based on element level data. At a minimum, the methodology provides a way for managers to structure their technical knowledge and make logical, sound and repetitive estimates of required future financial resources. Additionally, however, it can be used to verify the base assumptions, e.g. speed of deterioration, used to make the estimates. As opposed to other methodologies used in similar situations, this methodology allows consideration of the uncertainty in the input parameters. An example is used to illustrate the type of results obtained using this methodology with normally existing data, as well as some potential uses of the results.