Boundary condition at the outflow of a river model

EFDC+
1

Hi, Actually, I’m building a river 2D vertical averaged model. The outflow of my river is throwing in the reservoir, I want to know what kind of flow boundary conditions or any other boundary conditions can I applied here.Thanks

2

Hi Dimitri,For a river flowing into a dam reservoir, I would suggest one of two alternatives:1) Open Boundary: This type of boundary condition would require a water level to be specified at the reservoir. The advantage of this boundary type is the simplicity since all you need to do is specify the water level at the reservoir. The corollary to this type of boundary condition is that the open boundary acts as an infinite source and sink of water. Therefore, if the mass balance in your model upstream of the reservoir does not approach closure, you may see some physically strange behaviors near the open boundary (water flowing into the reservoir from the open boundary cells).2) Flow Boundary: This type of boundary condition would require you to specify a discharge flow rate for the reservoir. The advantage of this boundary type is that it is physically closer to how a dam reservoir behaves. The main disadvantage is that it requires you to have very good mass balance closure to sustain appropriate water levels in the system (i.e. the mass in-out must be close to zero). So if you have good control over the mass in and out of your system, the flow boundary will produce the best results. However, the flow boundary may be more data-intensive depending on the length and number of inflows into your river. On the other hand, the open boundary is quick and easy to set up, but you may sacrifice overall accuracy and realism.Hope that helps,Tom

3

Thanks Tom, Actually, it’s like EFDC doesn’t take in account the outflow condition. My code is crashing a the first step for this reason. Thanks Ralph

4

Ralph,It looks like the root of your problem is that your simulation is predicting negative depths. This is likely related to your timestep being too large for the kind of boundary conditions you are applying. The first thing you should try is starting with a smaller timestep.From the look of it, this is also related to your other post.If you like, you can attach your model input files in a zip and we will have a look.Tom