The current state-of-the-art in timetable analysis in the presence of disruptions is to use railway microsimulation, which typically yields detailed results on infrastructure or timetable performance. However, micro-simulation is time-consuming and requires a detailed infrastructure model. This paper outlines a macroscopic approach which aims at reducing execution time by restricting the level of detail to high-level relations between significant events. In particular, the effect of disruptions is modelled by sampling delay times from probability distributions obtained from historical data. In this paper, we test whether this approach, given common disruption scenarios, still allows accurate results on delays to be obtained. Two disruption scenarios were simulated in RailSys and with the new method, using limited parameter tuning. In the results, visually similar delay distributions were observed. Although there is some room for improvements in accuracy, the new approach appears promising, and we found no evidence against its suitability in the presence of disruptions.