Large amounts of dangerous chemicals are stored near cities or are transported via truck and rail near and through cities. Many of these chemicals are heavier-than-air when released (e.g., chlorine, ammonia, hydrogen fluoride). Few dense gas models exist for rapidly computing the transport and dispersion of a heavier-than-air cloud in a city accounting for the effects of buildings. In this presentation, we describe a new approach using the Quick Urban & Industrial Complex (QUIC) dispersion modeling system with an embedded shallow water model. Shallow water equations are well suited to model dense gas behavior in a complex environment because it is possible to include ground slopes and obstacles. The model integrates the partial differential equations to describe the cloud height and spreading velocity in the horizontal directions. The entrainment is evaluated with the Eidsvik approach (1980) and, in principle, it can differ in each part of the dense gas cloud. The shallow water model uses the wind field of the QUIC wind solver which allows for incorporation of complex wind patterns in the presence of obstacles and to have a more representative reference velocity to quantify the entrainment. The Lagrangian random-walk model is used within the dense gas cloud and can switch to the neutral gas simulation when the cloud density approaches the air density. Multiple sources having the same initial density can also be simulated. The manuscript describes the shallow water theoretical framework, the parameterization of entrainment, the obstacle and ground slope modeling, and the model verification and validation.