You are here

Home › Forums › W2 Forum › Hydrodynamics

Dealing with drying cells

Wed, 04/12/2017 - 8:38am — JoseI
Hi all! My name is Jose and I am doing some tests with Ce Qual. I am trying to model hydrodynamics and water quality of a reservoir that is feeded by pumps, that means that the inflow is coastal water that is pumped to the reservoir during the day, and the outflow is through the same pump back to the ocean during night, plus evaporation. This is simulated using two files, one to pump in and one to sink the water out of the system (i am not modelling the interaction back to the ocean yet). During spring/summer i can pump up to the reservoir more water than in fall/winter, that means that in some dates, the reservoir goes dry in a bunch of segments. The thing is that the reservoir is almost symmetrical, so when the water level decrease i get to dry the downstream segments (with the upstream there is no problem because i noted that the model uses the upmost active segment to continue). Is there a way to define as "downstream" another segments that is not the last one? Considering that the "pump out/trib in" point is located near to the bottom of the segment 13 (in a total of 30 segments). The model crash when i get to dry the "downstream" segment always (Segment n° 30), the only way that i have found to fix this, is modifying the depth of the downstream segment so it wont dry, but that change the geometry of the reservoir. Best regards!
Forums: 
Hydrodynamics
Wed, 04/12/2017 - 3:16pm — Benjamin

Not sure if I completely understand the problem, but what about tricking the model by making your reservoir 3 branches. One branch (the downstream branch) would be a single segment in the center of your reservoir (the deepest part). Then the other two branches would flow into that branch, essentially being mirror images of each other. So something like:
    Branch 1 --> Branch 3 <-- Branch 2
Branch 3 = Downstream branch (center of reservoir)
Branch 1 = ~1/2 the reservoir
Branch 2 = ~1/2 the reservoir
Tue, 04/18/2017 - 3:11pm — JoseI

First of all. Thanks a lot for your reply Benjamin, and excuse my english. Based on what you posted i think you've got a good understanding on what my problem is.

I have tried that way already, the thing is that i can't feed the branch 3 (beeing 1 segment) with the other 2 branches, because in the page 46/827 of the manual, Grid restrictions, point 3 says: "Two branches may not connect at the same segment of another branch".
+++,,,\_,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,__/,,,,,+++
!!!,,,,,,,,,\_,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,__/
,,,,,,,,,,,,,,,,\_,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,__/
---,,,,,,,,,,,,,,,,\_,,,,,,,,,,,,,,,,,,,o,__/,,,,,,,,,,,,,,,,,,,,,,---
,,,,,,,,,,,,,,,,,,,,,,,,\_________/
The "+/-" sign would be the max/min level of the reservoir (summer/winter), the "o" is the pump (in/outflow). And when the "!!!" dries, then it crash.

My model actually looks like this.

,+++,,|,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,__/,,+++
,,,,,,,,,,|,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,__/
,,,,,,,,,,|,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,__/
---,,,,,,|,,,,,,,,,,,,,,,,,,,o,,,,,,__/,,,,,,,,,,,,,,,,,,,,---
,,,,,,,,,,,\____________/

Thanks in advance.

Wed, 04/19/2017 - 11:24am — Benjamin

Ahh, I forgot about that restriction of 1 connected branch per segment. What if you simply omitted branch 3 from the previous example then? In that case, choose one of the branches to be the "downstream" branch which ends at the center, and the other branch connects at that downstream interface. Alternatively, branch 3 could be two segments, so branch 1 and branch 2 could connect to different segments of branch 3. The branch 3 segment you choose to be downstream would likely be the one closer to your pump, but I don't know if it will matter.
Thu, 04/20/2017 - 1:11pm — JoseI

Thanks again for your reply Benjamin! I will do some tests later this month following your recomendations to see how it goes! I will keep in touch!