Aircraft departures often follow standardized and restrictive routes intended to guarantee a safe transition to the en-route network. Since the procedures must take the flight performance of many aircraft types into account, they represent a compromise between numerous optima and must be consistent with noise abatement strategies. This paper investigates the concept of departure funnels, in which flights can adopt their optimal profile within a procedural space based on actual flight performance to replace standard routes. For this, an algorithm based on DBSCAN identifies typical traffic flow funnels for a set of radar tracks as reference and individually optimized flight trajectories as preferred funnels. For the latter, an innovative 3D pathfinding grid is developed, which expands dynamically using the specific flight performance of the aircraft type and enables evaluation of operating costs due to wind and fuel consumption. From the clustered traffic flows, a funnel starting at the runway is determined based on the variance of the flight profiles along their mean trajectory. This funnel provides a restricted space for individual trajectory optimization for the day of operation. The procedure is applied using the example of Munich Airport, where the funnel size and the associated fuel-saving potential are determined. The results indicate an average fuel-saving potential of 0.4% with respect to the trip fuel.