The work with automated CFD-simulations of nozzles and propellers has resulted in important understanding for new designs of nozzles optimized for free running, towing and bollard pull conditions. The process includes automated process as well as manual adaptation to obtain an optimum design and to understand the underlying physical processes. In order to get realistic resulting trends from CFD-simulations it is important to include the 3D propeller geometry. This was concluded from using both a fully automated 2D axisymmetric simulation model and an automated 3D simulation model including propeller geometry. For simulations at bollard pull conditions it is effective to use a quarter of a sphere as boundary for the computational domain. The work of optimizing the performance of the nozzle-propeller system results in two new nozzle designs, one for free run and one for bollard pull conditions. The nozzle optimized for free running conditions improves the nozzle-propeller system by 3% in the computations at free running condition as compared to 19A. Also, the free running nozzle maintains the efficiency at bollard pull condition as compared to 19A. Larger efficiency increases can be achieved by changing the design point. In the computations, the nozzle optimized for bollard pull condition improves the bollard pull merit with 5% and decreases performance with 1% at free run condition as compared to the 19A nozzle. Parts of these gains comes from profile itself and part of the improvement comes from changing the outer dimensions of the nozzle. It is concluded that without changing the propeller design point, the 19A nozzle has a good performance and changing the profile has a limited effect if remaining at the same design point. However, if the propeller design point is adapted for the designed nozzle, the improvement will be greater. Generally for the nozzles, the thickness of the profile is the most important feature for free running condition and the geometry at the diffuser part of the nozzle is very important. For bollard pull, the outer dimensions is of large importance, a large nozzle, both in length and dimension, with an optimized profiles produces higher bollard pull as compared to the 19A nozzle setup. The experimental investigation of the BEN-HD and BEN-HS nozzle, which are based on the present investigation, verifies the results presented in the computational investigation. At the 19A free running design point, the difference between the nozzles is limited, but if the design point is adapted for the new nozzle design, the improvement will be greater.