In dredging operation, the high-pressure water jet is widely used for the excavation of soil. To study the jetting process and optimize the dredging devices, the moving vertical water jet penetrating cohesive soil experiments were carried out by Nobel (2013). However, in terms of the design optimization for the dredging devices, it is not easy to change the jet scale and soil properties during the experiment due to time and economic constraints. Some detailed physics during the jetting process, e.g. pressure on the soil surface and shear plane inside the soil during jetting, were also not monitored during the experiment. Therefore, numerical simulation is chosen to optimize the design of dredging devices and study the physics of the jetting process. A CFD (computational fluid dynamics) numerical model is built to simulate the moving jet penetrating cohesive soil. The soil is modeled as a Bingham plastic. The sediment transport is modeled by using drift-flux model. The moving jet modeling is achieved by using dynamic mesh algorithms AMI (arbitrary mesh interface) and A/R (cell layer addition removal). The CFD numerical model has been validated with the experiment of Nobel. After the validation, an analysis of the jetting process based on this CFD model is accomplished proving that the CFD model can reveal the details of the soil failure process during jetting. This thesis work reveals that it is possible to describe the hydraulic excavation of cohesive soil with reasonable accuracy using CFD numerical model. The CFD model can also reveal the details of the soil failure process that could not be retrieved from the experiment. Since the model is generic, it can be applied for a jet bar with multiple nozzles. This is helpful to improve the design of dredging equipment, optimize the operational settings and estimate the production.