Keywords: Water supply systems. Lifelines. Seismic performance. Extreme events. Los Angeles Department of Water and Power (LADWP). Multi-hazard performance. Numerical simulations. Networks. Decision support. Damage. Displacement.
Abstract: This report presents the development of a functional Decision Support System for the seismic and multi-hazard performance of water supplies. An improved hydraulic network model of the full 2007 Los Angeles Department of Water and Power (LADWP) water distribution system is presented. The improved model includes an enhanced simulation of the time-dependent response, all sources of earthquake damage (loss of aqueducts, electric power outage, effects of permanent and transient ground deformations on pipelines), and fragility curves to characterize probabilistically the seismic damage to facilities such as tanks, reservoirs, regulation stations and pumps. The network model is validated through comparison of model results for the effects of the 1994 Northridge earthquake with actual areas of lost water service as well as pre- and post-earthquake flow measurements documented by LADWP. Parametric studies show that the most important factor affecting the post-earthquake system performance is the demand, followed by transient ground displacements. System serviceability is not influenced significantly by moderate changes in negative pressure tolerance, moderate variations in leakage rates, or changes in the percentages of breaks and leaks that cover the range of previous observations. An actual decision support problem faced by LADWP system management is used to demonstrate the application of the proposed methodology. The LADWP is modeled with and without several key reservoirs, which have been removed from service to meet water quality standards, to assess their influence on supplying water after an earthquake. It is demonstrated that opening the disconnected reservoirs immediately after a severe earthquake improves serviceability, with the most substantial impact in areas with the highest population densities.