View Full Version : Instability in River Reaches

January 16, 2017, 03:24 AM
I have a client supplied model with river reaches that surcharge and flood onto the 2D mesh with further exchange of flow as one reach spills into the next one downstream via the mesh. However, when surcharging of the reaches occurs, there is significant instability with flows that are nominally ~2m3/s oscillating wildly between -10m3/s and +60m3/s even when running with a 1 second timestep. The river reaches need to flood as there is a dam and throttle at the end of the valley and both the reaches and mesh provide storage. The question I have is: does the software allow for flooding in this manner to occur, or is it a quirk of the model that I am using that generates this instability.

Duncan Kitts
January 16, 2017, 03:43 AM
The software will allow flow to pass overbank in this situation. It does sound like your model is unstable.

If you haven't already done so, I would recommend selecting the option to 'Link 1D and 2D Calculations at minor timesteps' in the 2D parameters. Also, check that you are using the bank linearization threshold in the simulation parameters. Both options are explained in the help.

You may also want to review your modular limit for the banklines. If the results suggest the banks should be drowning out during the run, a high modular limit would inhibit this from happening correctly and instead maintaining free flow (Mode 1 shown on the “River Reach - Bank Flows” help page) which could lead to instabilities. You could try reducing the modular limit to a value in the region of 0.5/0.6 to 0.7. This will allow the banks to drown out when necessary and will also allow the “Drowned bank linearisation threshold” simulation parameter to work effectively, further reducing flow reversals. If the instabilities were as a result of the banks drowning out, then you should find that this helps a lot to prevent the instabilities.

January 18, 2017, 04:03 AM
Thanks Duncan. That has addressed the instabilities.