This thesis presents an analysis and evaluation of wells drilled for Wintershall GmbH in the Emlichheim Oilfield in Germany from 1997 to 2012. Recommendations for the drilling program of a future campaign are developed based on the results of previous campaigns. For this purpose, learning within the drilling process is evaluated, different learning curve models are applied to the data and a model for the future campaign is generated. Non-productive time is identified from daily drilling reports. These events are summarized according to type, duration and causes studied for further analysis. Daily drilling reports are the prime data source used as they are the most continuous documentation throughout all the wells drilled. Invisible lost time analysis is performed exemplarily on three wells based on sensor data. The three wells were picked according to their well construction time: the longest and the shortest duration and one randomly picked well. From this analysis, time savings potentials are derived and applied on an exemplary well. Grouping of similar steps is used as assisting measure to portray the performance and capture trends to allow interpretation. The time estimates for the new drilling program are calculated based on a risk assessed evaluation.