| A community under the threat of natural hazards must contend with the possibility of damage to its transportation infrastructure. Such damage can result in significant repair costs and loss of life. Equally as serious, a disruption in the transportation network can have serious impacts on a community's ability to recover from a disaster, long after the initial emergency has ended. Engineering advances have made it possible to retrofit key infrastructure elements, such as highway bridges. If budgets are insufficient to retrofit all key bridges immediately, a prioritization scheme must be employed to help select which will receive attention first.; Presently, the prioritization schemes used by engineers and planning officials rely on assessments of the vulnerability and importance of individual bridges, without consideration of how each relates to the entire transportation network as a whole. The earthquake engineering community describes this as a serious flaw in the present approach. This research adapts a technique developed for evaluating potential network expansions, the Discrete Network Design Problem (DNDP), to retrofit prioritization. In order to capture the complex stochastic nature of natural hazards, an existing DNDP algorithm is modified to find the optimal set of retrofit projects across a representative range of scenarios. The new algorithm is applied to a classic test network, under several varying natural hazard conditions. The performance of the method against competing approaches is evaluated, as is the algorithm's computational performance. |
