Hello,
In many models, we have instabilities when we use culverts (option = 2) and when Z1 is different from Z2.
Recently, I have a very small model so I tried to debbug this without success unfortunately.
I print every time step to listing and graphic output but I can't observe any instabilities but the model is unstable
GRACJG (BIEF) : EXCEEDING MAXIMUM ITERATIONS: 200 RELATIVE PRECISION: NaN
If I change Z1 and Z2 in my culvert file with Z1=Z2=max(Z1, Z2), the model runs smoothly without issues.
Fun fact, if I restart the computation from the latest valid output, the computation finishes without issues.
I can't share the model in public but I could email it to someone.