energy gradeline has inverse slope in unsteady analysis

[popriver]

I have a fairly steep slope in my model (18%). The energy gradeline is sloping up instead of down, as the water flows down this steep area (in a very bad way). It acually starts to go up at he top of the steep slope (or one cross-section up). I'm trying contraction coefficients in unsteady modeling. I don't know what the range is to use though. Any thoughts? No recommendations in RAS on what to use for unsteady, only steady. Wonder if we shoud use the same values.