It recently happened to me, along a breakline, and it came out that computational interval was too high. reducing that (from 10 to 2 sec) has solved the problem.
I reduced the computation interval from 15s to 2s but I got the same result. Is there anything else I could change? What is the general approach for tackling this kind of error?
It looks like your breakline is missing a highpoint. Be as accurate as possible. If you break that location up into cross sections per cell youll notice your breakline has 6 faces and the US and DS faces are both at low points this can cause volume transfer