In some environments wireless communication with electromagnetic radiation is unfeasible or simply too dangerous. Molecular Communication (MolCom) works by the sender emitting particles instead and is therefore a promising alternative. A carrier medium, such as a gas or liquid, physically moves the particles through space. The presence of particles is then sensed at receivers, which recover the transmitted data. In an industrial macro-scale MolCom system, existing pipe systems could be used, into which magnetic particles are injected. Current simulators for this either lack the capability to consider the complex flow inside the pipe system or suffer from poor performance.I present the "Pogona" simulator which uses Computational Fluid Dynamics (CFD) to model the flow of liquid inside the pipe system. The flow is imported into the simulator and used to predict the movement of particles. Additionally, models for the injection and sensing of particles are proposed. To improve performance and flexibility, a scene approach that allows the configuration of scenarios consisting of independent objects is proposed. I verified my simulator prototype with simple scenarios that are compared to analytical results and known-good implementations. Simulation scenarios that match an existing testbed were created and their output compared to empirical measurements. Overall, a low degree of agreement was achieved and room for improvement remains. My analysis shows that the interpolation algorithm applied to the CFD flow output is of importance. However, finding well-suited interpolation algorithms remains an open research task.