Simulating Controlled Dam Release

[jwrivenb]

Hi,

I have a question about modeling controlled flow through an inline structure (dam). I have 2 types of flows moving through the structure, one coming from a powerhouse with 2 identical turbines and 3 identical irrigation jet valves. I have stage vs flow rating curves for both. I am attempting to model these as gates, with 2 separate gate groups one for the two turbines and the other for the 3 jet valves. I have specified a user defined rating curve and have set up the rating tables for both gate groups. For the unsteady flow data I have specified TS gate openings as a boundary condition at the dam and have them both completely open.

I intend to model these gates as submerged flow, since the intakes and the turbines are NOT designed to be flood control structures. The problem is when I run the simulation with the gates open, I get a draw down from the reservior expected water surface elevation (from the closest upstream XS) to the invert of the gate at the upstream end of the structure. This seems unrealistic to me since gates are not supposed to be unsubmerged. My guess is that the gates in ras are intended to be flood control structures not hydroelectric or irrigation structures.

My question is what is causing this drawdown? What should I be using to model hydroelectric and irrigation (submerged) flows through a dam? Thanks.

[jwrivenb]

I have decided to add some output data to help explain my situation. I have 2 model runs: 1) With gates open and 2) All gates closed.

Run 1: Gates Open
XS UpStream
WSE: 1020.43
Qtot: 151.83

Inline Structure
WSE: 982.00 (also the gate invert) **very low**
Qtot: 152.26 (expected flow from gates)
Qgates: 0.00

XS DownStream
WSE: 949.98
Qtot: 152.26

Run 2: Gates Closed
XS US
WSE: 1020.68
Qtot: 5.00 (pilot flow)

Inline Structure
WSE: 1020.68
Qtot: 5.00 (pilot)
Qgates: (blank)

XS DS
WSE: 949.91
Qtot: 5.00

So you can see that, with the gates on, I am getting the correct flow but the WSE and Qgates have incorrect values. The only situation I can get a resonable WSE is when the gates are off…but I can’t get the flow I want. Is this normal for what I am trying to do? Should I ignore the WSE at the dam if I am getting the correct flow?

[Chris G.]

If you are indeed only using the rating curves, then the problem lies there. It’s very simple, RAS takes the computed flow at the structure and uses the curve(s) to get the upstream water surface elevation. That’s it. You could even do some hand calculations to verify your rating curves. Not sure if this has anything to do with it, but be careful using the gates with low-level outlets, since the user-defined curves are NOT tailwater dependant (i.e. no way to account for high tailwater (submerged ocnditions).

[jwrivenb]

Aparently I am confused. I was under the impression that it was the other way around that the flow is computed based elevation by using the rating curve on the upstream wse. I am interested in computing flow through the gates based on a known wse.

I’m really just trying to make dummy gates to simulate irrigation jet valve and hydroelectric turbine discharge under normal flow conditions. Since these structures are designed to operate only under water, my gates will have to account for submerged conditions. How would you simulate this? If I cannot use gates at the inline structure, will I have to use a lateral structures with a RC for each process?

I can’t get you a screenshot but I can draw you a picture in paint of what’s going on.

[Chris G.]

If you are okay using a vertical lift gate to simulate these features, then RAS will take into account the submergence. However, when you set up your gate as a “user defined” gate, you put in a family of rating curves that are independant of tailwater. It’s a limitation, Otherwise you’d have to have a family of family of rating curves :). One way around this is to use Rules as your gate boundary condition. You should be able to put in multiple rating curves, for a range of tailwater elevations, and direct RAS which one to use under a variety of conditions. Using Rules is tricky, requires some understanding of writing code/scripts, and has very little documentation to go with it. But this is an option if you have the time/patience to figure it out.

[Author]

Posts