Generating network representations of small-scale infrastructure using generally available data

Risk analysis (including resilience analysis) of infrastructure requires models that describe the connection of components and subsequent flow dynamics. However, the detailed information needed to define these models may not be available, especially for small-scale infrastructure that connect to every building. In this paper, we generate location-specific small-scale networks using detailed data that should always be available. We propose a general framework where we generate the network topology, we estimate the resource demand at each building, and we design the network components to meet the demands. This general framework is applicable to all types of infrastructure, but many procedures are specific to the type of network being generated. This paper develops the necessary procedures to generate sewer networks and illustrates the usage for an example network in a small study area in Seaside, Oregon. The proposed sewer network generator produces realistic sewer networks as compared to the real network of Seaside.