Environment and health-related concerns mean that pedal-bicycles as an alternative mode of urban transport are gaining ground, with study of motorised/non-motorised traffic mix a topic of practical interest in transportation science and traffic modelling. This thesis reports on a simulation model, developed for heterogeneous traffic on city networks with AD HOC lane-sharing, characteristic of Dublin streets. While based on simple cellular automaton rules, the vehicle movement model also accounts for vehicle type heterogeneity and network-specific factors, including the resolution of conflicts and effects of driver decisions on movement dynamics. The model has been implemented as an agent-based simulation framework. Its spatial component is based on a modular design that facilitates straightforward scenario configuration and scalability. In order to perform large network simulations, the framework has been adapted for parallel processing. Issues of both static and dynamic load balancing are considered. While detailed field data are not available for heterogeneous traffic on urban networks, which precludes precise quantitative validation, sensitivity analysis of the model was performed with a wide range of parameters and values. Macroscopic whole-network measures are defined and used to study a number of scenarios, the most manifest property of which is the contrast between slow and fast, vulnerable and less vulnerable agents in the traffic mix.