This discussion has been locked.
You can no longer post new replies to this discussion. If you have a question you can start a new discussion

Why does adding a spillway as a separate pond outlet all of a sudden cause it to engage when it previously wasn't?

I am modeling a pond with a composite outlet structure (riser with orifice openings). My pond has a spillway and at first i was modeling the riser/ orifice and spillway as a single composite structure connected to an outlet pipe. After some research on the Forum I realized that I should model the spillway as it's own separate composite structure as the spillway is emergency only and flow over it leaves the pond without getting into the riser outlet pipe.

However when I deactivate the spillway and just let the riser handle the flow, the max HGL in the pond is below the spillway invert, but when I turn the spillway on and run the model, the max HGL increases and becomes active. I am not sure why this occurring as all other model parameters remain the same, the only difference is that the spillway is active in one and off in the other.

      

  • Hi Cecily,

    It is hard to say what is causing this without seeing the model. Can you provide a copy? See: Sharing Hydraulic Model Files on the OpenFlows Forum

    Is it possible that the elevation of the outfall downstream of the spillway is set too high and somehow reverse flow is occurring when the spillway is activated?

    There may be some instability occurring and interfering with the results. Which solver are you using? I recommend the Explicit (SWMM) with the Routing Step set to 1 second, for models with ponds. Check the Continuity Error in the Calculation Summary and review graphs of the results to check for instability. The following articles have guidance on troubleshooting:

    Troubleshooting unstable SewerGEMS and CivilStorm results using the implicit solver

    Troubleshooting unstable SewerGEMS and CivilStorm model results using the Explicit SWMM Solver

    If this does not help, please provide a copy of the model for review, and reply here with the file name and any steps needed to reproduce.


    Regards,

    Jesse Dringoli
    Technical Support Manager, OpenFlows
    Bentley Communities Site Administrator
    Bentley Systems, Inc.

  • Thanks Jesse! Switching to Explicit solver does seem to work, but now at the most DS end of my model there appears to be an issue.

    The pond outfall pipe is connected to a DS channel that is connected to the outlet node. While there is flow leaving the pipe, it appears to never enter the channel or outlet node (maximum values for channel and outlet are 0 cfs). If flow is exiting the pipe, I would expect there to be flow in the DS elements. 

    Why does the model show zero discharge for the channel and outlet?

    I have used the secure link to provide you a copy of the model. The zip file I sent is titled PondCMolina and the actual model is titled Proposed_Effluent_5ft_Berm

    Running any of the Parent Scenarios should show that the channel and outlet DS of the pond outfall pipe are dry.

    If you would like to view my original spillway issue, please run the No Spillway Child Scenario (under the 100yr Parent) . Originally I was running it with the spillway NOT active with Implicit solver and results showed max HGL below spillway invert. When I would run it with an active spillway using Implicit, the max HGL increased by more than 1ft (but again using the Explicit seems to have fixed this). 

    I would like to note the while utilizing the Explicit solver fixed by spillway issue, it does yield different results than those found using the Implicit solver. 

    Can you please explain why I'm seeing Q and Max HGL differences between the two solvers and possibly provide guidance on how to determine which solver to use?

  • Thanks for providing a copy of the model.

    Why does the model show zero discharge for the channel and outlet?

    You have a user defined tailwater of 539.69 on the outfall, which prevents water from leaving the outfall (via the conduit) unless the water level rises above that elevation. The entire channel is below this, so the flow you see in the channel is the channel filling up, but the high downstream tailwater from the outfall prevents it from proceeding forward to the outfall. Once the channel fills up to an elevation of 531.4 ft, the water level does not rise further because it overtops headwall H-2. So, the flow in the channel is then lost to the system as overflow. Graph overflow for H-2 and you will see the positive values beginning when the HGL reaches 531.4 ft. You will also see in the "Routing Summary" section of the Calculation Summary that all the flow exits as overflow losses.

    Is the high tailwater on the outfall correct? Is the culvert allowed to surcharge instead of overflow? To allow flow to pass into the conduit, you will either need to lower (or remove) the outfall tailwater, or change the "Height" property of the headwall to a value above 13.29 (the outfall tailwater elevation minus the headwall invert elevation). For example if I set the Height to 15 ft, you will see the flow pass into the conduit.

    If you would like to view my original spillway issue, please run the No Spillway Child Scenario (under the 100yr Parent) . Originally I was running it with the spillway NOT active with Implicit solver and results showed max HGL below spillway invert. When I would run it with an active spillway using Implicit, the max HGL increased by more than 1ft (but again using the Explicit seems to have fixed this). 

    First, I think you might be using the wrong inflow alternative in the "no spillway" scenario. If you use the Scenario Comparison tool, it reveals that the "100 yr" scenario is using an alternative called "100 yr". where the "no spillway" scenario is using an alternative called "500 yr". I am guessing you meant to choose the 100 yr inflow alternative.

    Next, it appears that the advanced calculation options for the Implicit solver were adjusted to an extreme, which may be producing unreliable results. However, I was able to see the issue even with the default options, and even when making some further adjustments. But, if you change the top outfall to free outfall, the issue no longer occurs with the Implicit solver and the results are quite close to what you see with the SWMM solver after fixing the headwall. So, my conclusion is that this particular model configuration is not well suited for the Implicit solver, because it is known to struggle with ponds that have significant backwater effects (like in this case with the outfall tailwater). If you do not actually need the outfall tailwater, remove it and the Implicit solver should produce OK results, otherwise the SWMM solver is a better choice in this model.

    I would like to note the while utilizing the Explicit solver fixed by spillway issue, it does yield different results than those found using the Implicit solver. 

    Can you please explain why I'm seeing Q and Max HGL differences between the two solvers and possibly provide guidance on how to determine which solver to use?

    Generally speaking difference solvers have different limitations and different underlying calculation methodology and assumptions, though in this case you can match the results if you remove the outfall tailwater. As mentioned earlier, the SWMM solver works better with ponds and with tailwater and this is documented in the article previously provided: Troubleshooting unstable SewerGEMS and CivilStorm results using the implicit solver

    See also: Differences between solvers: GVF-Convex vs. GVF-Rational vs. Implicit vs. Explicit (SWMM)

    Some other observations:

    • The ground elevation for CS-6 is not reasonable - it is set to 6000 ft when it should be closer to 530 ft.
    • You appear to be using a slightly older version, 10.03.01.08. We have made several improvements since this version which you can read about here. Information on upgrading can be found here.


    Regards,

    Jesse Dringoli
    Technical Support Manager, OpenFlows
    Bentley Communities Site Administrator
    Bentley Systems, Inc.

  • Thanks Jesse! Your explanations help fill in some gaps and help me better understand how the model operates. 

    I did have one last question regarding the model I sent you...

    The tailwater is included because this system outfalls into a large creek that we are trying to account for the possible backwater effects of the creek on the pond. In my initial model, the outfall was set at the end of the pipe [CO-4(1)] and I was having issues with the Implicit solver due to the high tailwater which is why I added the outfall channel [CH-2] and adjusted the Base Calculation options. 

    Since the channel and outlet show no flow using the Explicit solver, I decided to turn those elements off [CH-2 and O-4]. When I turn those elements off and H-2 is the new outlet of the system with the tailwater set to 536.69 ft, the maximum flow through MH-4, C)-4(1) and H-2, according to results section in the properties table, does not match the maximum flow of the immediate upstream element [C)-3] and is approximately double the maximum inflow to the pond. I'm not sure where this additional flow is coming from as the only inflow input I have is the inflow hydrograph to the pond.

    However, when I look at Flow graph for these elements, their peaks align with the rest of the system and make sense. 

    Is it a know issue with the older version of the program I'm using to have the Results section in the Properties table not match the Graphs for the elements? 

  • Cecily, the values shown in the "Results" section are dependent on the timestep selected. In the Time Browser tool, you can select which time you would like to view, and the results in the properties will change to show the results at that time. By default you will be looking at time zero.

    If you are referring to the statistical results like maximum flow, try adding this to e graph of flow to see that they align. If the flow graph never reaches the horizontal maximum flow line, it could mean that your output increment is set too high and a momentary spike/peak in flow occurs between two reporting timesteps. See: Wetwell level drops when pumps are off, Result never reaches Maximum, or other unexpected results in graphs

    If this does not help, please provide a new copy of the model with steps to reproduce.


    Regards,

    Jesse Dringoli
    Technical Support Manager, OpenFlows
    Bentley Communities Site Administrator
    Bentley Systems, Inc.