Planning and scheduling models live from their accuracy. In the present work a gap analysis between planned data of simulations from the program “Aspen Petroleum Scheduler” and actual data has been conducted. The reasons for the discrepancies in the cuts of the plants and the in- and output analysis of tanks are diverse – nevertheless error sources could be identified in many cases. On one hand differences arise due to planning changes that can be improved by better communication between planning, scheduling and plant personal, but will always be present to some degree. On the other hand deviations in the yield structure of plants evolve due to differences of the modelled to the real plant. In addition to other proposals for measures this deviation analysis should be carried out again for some plants, and a corresponding adaption of the models should be conducted. In the second part of the work a concept for an algorithm has been created, which has the aim to initiate the automation of mixture of semi-finished to ready for delivery products and the pipeline transfer of these products to another depot in the program. Restrictions could be identified and taken into account in the programming structure.