Modeling a small stream and culvert, seeing higher water surface elevation for the 10-year profile than the 25, 50, and 100-year flows. The flow data is all correct. The model was functioning properly (or at least realistically) until I updated the roadway elevations.
Does the higher flows have a higher energy grade? If the higher flows have a higher velocity, then you get a higher velocity head (V^2/2g). And it is possible that the water surface can actually be lower. So if the energy grade looks realistic, I wouldn’t automatically assume that there is a problem with the results.
Doesnt matter if its a post. Its always a cause for concern, and the engineer should evaluate all and any possibilities for change including the transition of a steady state model to an unsteady state model. But there are instances when velocity head will occur as described above. This is a common comment by districts reviewing scour depths as well. Many people think that just because the 500-yr flow is higher than the 100-yr flow that the 500-yr scour depth MUST be greater; and its simply not the case. Bottom line; its something as the engineer you should look into, but there are explanations and combinations of numerical inputs and geometric features for which a profile will slightly cross.