One could argue that drift loads are the most important snow load since they account for roughly 75% of all snow-related structural problems.  The various types of roof snowdrifts are reasonably well understood.  However, a new snowdrift was recently observed downwind of a run of roof-top refrigeration piping, which did not seem consistent with our current understanding of snowdrift formation.  In this article, the drift formation processes for four common snowdrifts – leeward roof step drifts, gable roof drifts, windward roof step drift, and parapet wall/roof projection drifts – are reviewed as well as the apparent formation process for the new roof-top piping-run drift.  Finally, an explanation for the apparent inconsistency between the new piping-run drift and common drifts is provided.