Irrigation Headgate Modeling Problem

[tatraxle]

I am modeling an irrigation diversion in Hec-Ras. The diversion consist of a headgate located on an outside bend of the channel. Immediately downstream of the headgate is a pin/plank structure which will act as a grade control structure. There is about a .5' drop from the floor of the pin/plank to the downstream channel. In addition, the floor elevation of the headgate is about 1' lower than the pin/plank floor. The system is being designed so that the pin/plank is only an insurance policy, and will not have boards in it unless there is a problem.

I initially modeled the headgates using hy-8 to determine the headwater elevation needed to get the required flow in the ditch. I then performed an at-a-station calc at the design channel cross section immediately upstream of the diversion to determine what the actual water surface elevation would be. This seemed like a pretty straightforward approach.

I then decided to model the system in Hec-Ras. I initially modeled the headgate as a lateral weir with sluice gates. The results indicate that the watersurface upstream of the lateral weir (headgate) begins to drop about 50' upstream. The water surface immediately upstream of the structure is about .5' lower than the normal depth elevation further upstream. This seems unusual, as the channel upstream is 18' wide and the headgates total 12' in width. There is a drop in elevation from the channel to the headgate floor, but it seems that the pin/plank structure immediately downstream in the main channel should act as grade control and maintain the water surface elevation.

I also remodeled the system using a junction and in-line structures. I got pretty similar results, with the water surface still dropping immediately upstream of the headgate. I've been beating my head against this for 3 days now and can't come to a conclusion as to what is going on.

I can imagine that there would be a drop in elevation immediately (within 2-3') upstream of the headgate, but I would assume the headwater elevation would still be a function of the normal depth of the stream channel. Is my approach incorrect here? Any help on this would be greatly appreciated. Thanks

 

[eea]

Not a direct solution / answer by any stretch, but I'm curious as to your cross-section frequency along the channel.

 

[gbam]

First off, I noticed that the gate is located on the outside of a bend. Is the radius of the channel small enough to induce super-elevation in the water surface. Have you considered that in your calculations for your gate? For the change in water surface, it is expected due to the reduction in discharge and lower downstream water surface (Not as much back water; sub regime I presume). You may want to increase the number of cross sections if your system is that sensative as eea mentioned.

 

[tatraxle]

Thanks for the reply. I have not considered super elevation. I don't think superelvation will be a large factor because most of the water will be going into the canal during irrigation season. I have numerous cross sections through each reach to try to accurately model the system. I expect to see a drop in the water surface at the headgate since the channel drops, but the model is showing a significant drop starting 10' upstream of the gate. That seems unlikely to me that w.s.e. would be affected that far upstream.