Planning Restoration of Vital Infrastructure Services Following Hurricane Sandy: Lessons Learned for Energy and Transportation

Abstract

Hurricane Sandy arrived on October 29, 2012 with enormous adverse impacts on one of the largest metropolitan areas in the world, the New York region, including its electric power and transportation systems that disrupted the social and economic fabric of a wide geographic area. How these services were restored is an important basis for proactively designing robust public services not only in areas affected by Hurricane Sandy but also for many similar areas and events worldwide. Recovery rates are one important indicator of consequences of disaster. Selected physical and social measures of recovery are analyzed for New York City electricity and rail transit services following Hurricane Sandy to capture the ability of these services to be restored over various time periods, and to illustrate the use of indicators of restoration for infrastructure service resilience analyses for disaster management and planning. Electricity impacts and recovery are presented in user-oriented terms as the ratio of customers affected and restored to both customers without service and total customers served in the period following the hurricane. Results showed that restoration rates varied among the boroughs of NYC potentially reflecting differences in infrastructure resilience and exposure conditions. Rail transit impacts and recovery were analyzed in terms of both the restoration of the two dozen subway lines as a physical measure and ridership by station as a user or social measure. Line recovery analyses showed that within about a couple of weeks of the storm almost 80 percent of the lines were fully restored, almost 20 percent were partially restored, and complete suspensions were rare indicating a high degree of restoration capacity. Ridership measured at stations was compared to levels and rates of change over various time periods around the storm period, as well as from the previous year. High correlations in ridership between single days before and after the storm indicated little redistribution of ridership across stations between those time periods, however, rates of change indicated about a 14–16 percent decline in ridership system-wide within a two week period following the storm relative to the same time period the year before, recovering later, possibly reflecting slower rates of social recovery relative to the recovery of the physical transit systems. A longer period of service quality effects occurred due to lingering equipment damage potentially attributed for example to water damage and corrosion from saltwater. These findings point to the importance of recovery time as at least an initial guide to ways of systematizing service restoration measures and are a foundation for physical, social, and institutional adaptation strategies some of which are already being put in place. Larger considerations remain however in interpreting recovery restoration goals and who decides them.