The infrastructures of modern society are connected in a complex network, relying on each other for their own functionality. When one or more infrastructure fails, it can also cause other dependent infrastructures to fail or suffer from reduced efficiency, which can lead to problems with further infrastructures in a cascade effect. By finding the most susceptible components of a network, we can prevent these initial failures from happening altogether. This paper describes the use of multi-agent simulation to simulate the infrastructural network of a virtual city experiencing a flood. With repeated iterations of the simulation, we can estimate the most likely outcomes of a flood on the infrastructural network through the Monte Carlo method.