SEISMIC RISK ASSESSMENT OF INTERDEPENDENT ELECTRIC POWER AND WATER SUPPLY SYSTEMS IN CHRISTCHURCH, NEW ZEALAND

This paper applies a generalized framework for modelling interdependent infrastructure system performance to a case study
\nof the electric power and water supply systems in Christchurch, New Zealand. Based on geospatial exposure data provided
\nby infrastructure operators, the case study identifies the key assets in each system and selects appropriate fragility functions.
\nThese are applied along with existing hazard prediction methods to first measure the performance of each system
\nindependently and then interdependencies are accounted for by two methods: supply area and proximity. A new
\ninterdependency response metric is proposed to measure the magnitude of the interdependency effect of electric power on
\nwater supply with minimal computation. For a 10,000-year stochastic earthquake catalogue the electric power network
\nexhibits high resilience with system performance issues arising in less than 3% of events. Exceedance probability curves
\nshow that the water supply system is less resilient, due to the effect of ground shaking and liquefaction on pipelines, and
\nthat for this case study, the method for modelling interdependency makes little difference to the results. The
\ninterdependency metric calculated for each event is low in most cases. However, regression analysis shows a strong linear
\ntrend between the interdependency metric and the loss metric for the electric power network and this observation may be
\nexploited in future to develop a simple method for predicting the interdependency effect between any two systems without
\nfull simulation.