Improving the resilience of urban infrastructure, such as transport, energy and telecommunication networks, is a key aspect of disaster risk management. These networks are interdependent in complex ways and the overall risk has to be assessed from a system-of-systems perspective rather than by looking at individual networks. The theory of complex networks has been instrumental in modelling such interdependent infrastructure systems and identifying conditions for catastrophic cascades of failures. Many of the coupling phenomena, however, cannot be captured with models that consider only connectivity. This paper presents a method that models the transmission of failure across system boundaries using measures that represent the capacity and load at the level of individual network components. This enables a more comprehensive analysis, for example by taking into account the dynamic effects of demand fluctuations, redistribution of network loads and recovery mechanisms.