Railway networks already play a major role in today's transportation of passengers and goods. With the growing efforts to protect the environment and climate, its role is likely to become even more important in the near future. To handle this appropriately we need a way to schedule trains in a fast and efficient way. A recent publication for computer network scheduling made conflict graphs in their field viable, by building it up dynamically till a solution can be found. We to took this idea and implemented it for joint routing and railroad scheduling. We added a Potential Conflict mechanism to their algorithm to improve on its performance and compare different strategies to find the best way to extend the conflict graph for our domain. The evaluation shows that with those two adaptions we are able to scale the conflict graph solution from a single train station to the German railroad network.