A probability based decision support methodology for seismic retrofit of bridges

The need for a rational method of project selection for bridge retrofit cannot be overemphasized. This need has long been recognized by various government agencies with the result that many departments of transportation around the country have devised their methodology for project selection. This need is even more crucial in seismic prone regions where an earthquake could hit many bridges at the same time. Which bridges get retrofitted and to what extent is the retrofit? If one bridge is selected, what parts of the bridge get retrofitted and how best can the retrofit activity be carried out to get the most out of the money spent? These types of questions become more important since bridge retrofit is not only a major capital public expenditure, but it also directly influences the economic activities in the region of interest.; From California to New York; from Texas to Minnesota various methodologies for project selection have been adopted by States' Departments of Transportation for project selection. The techniques involved in most of these procedures usually involves the calculation of some type of prioritization index. The calculated indices are then ranked and selection is made based on the ranking from the list. Usually, a bridge with the higher priority index stands a better chance of being selected for retrofit while those with lower priority indices may not have a chance of being selected.; Researchers in the field of transportation and bridge engineering have also proposed many methodologies for project selection. Some researchers also based their methodologies on the computation of a ranking index, while others seek to find the true weight of a decision parameter by the use of fuzzy number theory. In calculating the priority index, factors such as importance, seismicity, age, vulnerability etc. are usually incorporated into the model. What many of these methodologies lack is the direct application of the earthquake magnitude or the probabilistic nature of the decision problem in their formulations for calculating the priority or ranking index.; The methodology proposed herein addresses the decision problem by using a probability tree diagram to calculate the Net Present Value which is the decision parameter. This procedure allows one to incorporate any pertinent decision making parameter into the tree diagram with the result that a more rational decision for project selection is made. Some of the parameters included in the probability tree diagram are probability of failure, probability of closure, expected duration of closure, user costs, annual equivalent closure costs etc. A spreadsheet is used to calculate the Net Present Value of the project from which a decision is made for project selection.