That error code is not useful as it is basically saying it is just crashing.
My guess is that the 1D portion of your model is going unstable causing odd results. If the 1D channel is quite different, make sure ineffective flow areas are set correctly, maybe add some interpolated cross-sections to see if that fixes the issue.
Do you have the 1D/2D iterations turned on?
Are the weir coefficients for your lateral structure reasonable?