[Red Ember]

Hi to all.
I’m trying to estimate kinetic energy loss for turbine rotor wheel (I have one half of stage only).
Calculation residuals proved to be good as well as monitor points, so I plot a chart for kinetic energy loss spanwise distribution. I used two formulas:
1. dzita = (prel*in-prel*out)/(prel*in-p out)
2.dzita = (prel*in-prel*out)/(prel*out-p out),
where prel* is total pressure in rel frame. Actually I don’t know which one is correct exactly.
But there's considerable part in the chart that yields negative loss!
What's wrong?
I attached domain picture, expected chart and calculated one (via both formulas).