Interdependent Infrastructure Systems: A Case Study of the St. Kitts Power and Water Distribution Systems

Abstract

Modelling interdependent infrastructure systems is a challenging task, and the available information about the infrastructures is often limited. Although a number of methods for modelling interdependent infrastructures have been developed, only a few detailed case studies of risk analysis of real-world interdependent infrastructure systems have been conducted. The objective of this thesis was to present such a case study, and show that it is possible to model interdependent infrastructures with limited amount of information.

This thesis contributes to and extends the work of an unpublished case study by Brunner et al. (n.d.), where a risk analysis of the power and water distribution systems on the island of St. Kitts is performed in relation to the hurricane threat to the island. In the case study, simulation models are used to perform analyses of the power and water distribution systems, as well as for simulating hurricanes and studying the effect of them on the systems. The water distribution system is dependent on the power system due to wells that require electrical power to pump water into the system. Thus, when hurricanes break power poles and cause power outages on the island, the water distribution system is also affected. The information about the infrastructures and the hurricanes that the models are based upon, were publicly available information.

Brunner et al. (n.d.) performed analyses of the power system to identify the most vulnerable parts of the system by simulating historical hurricanes to analyze how these hurricanes affected the power and water distribution systems. This thesis extended to the work of Brunner et al. (n.d.) by performing a synthetic hurricane analysis, where hurricanes were constructed and simulated in order to see if these could provide additional information beyond what the simulation of the historical hurricanes gave. Also, an individual water node analysis was performed, where the wells were analyzed by removing the dependency on electrical power during the simulations for one well at the time.

The results of the performed analyses in this case study have provided information that is valuable from a risk management perspective. It has been shown that risk analyses where interdependent infrastructures are modelled can be conducted, even when only publicly available information is used. The main findings of the risk analysis were that the most vulnerable part of the power system was the powerline that encompasses the eastern side of the island from the main city of Basseterre in the south to the northernmost part of the island. For power outages lasting for multiple days, also a central part of the power system located in Basseterre was found to be particularly vulnerable. In cases where the entire island is affected by a power outage, the most critical wells were found to be the wells located on the western side of the island. The simulation of historical and synthetic hurricanes identified which hurricanes that have the potential to cause the worst damages. Based on these findings, possible upgrades to the infrastructures were suggested, as well as preventative measures that can be implemented to reduce the damages in case of hurricane activity.