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

SewerGEMS High Continuity Error during specific storm that causes flooding of catch basins

Hello, I have a SewerGEMS model set up to review 2, 10, 100, and 2500 year storm scenario's. For this model, the explicit solver achieved much better convergence and continuity error results for the 2, 10, and 100 year storms. However, I am seeing very poor results currently when running the 2500 year storm. This storm event results in a few of the catch basins having "overflow" which I think could be a portion of my problem.

The 2500 year storm has to be run so that I can look at the ponding results at the various flooding catchments and verify that the ponded depth does not impact nearby roads/facilities. I am new to reviewing this ponding depth and would be open to any suggestions on how to best set up and review this portion of the model.

Another aspect of the model that appears to be incorrect is the maximum inflow/outflow for each of Basin A and Basin B. 50-60cfs for each basin outflow is expected for each and that roughly appears to coincide with what the flow in the outfall pipes is. However, the pond flex table shows a much larger number for the outflow. Basins A and B also are showing much larger than expected inflows. When running the hydraulic reviewer volume balance on the ponds, the reviewer shows a huge difference in inflow volume vs. outflow volume. I am not sure if this could be related to the overflowing inlets or not, but I am 100% sure on what steps to take next to troubleshoot this model. The problems all seemed to start arising as soon as some flooding of the inlets began happens in the 2500-year event as the other rainfall event have little issue. I have attached the model for reviewing.

7002.New folder.zip

Parents
  • Hello Joshua,

    I was able to get a better continuity error by adjusting the Explicit solver calculation option filed "Surcharge Method" to Slot and the SWMM Output Increment to 0.1 hours. There were still instabilities related to pond outflow, but this may be related to reverse flow into the ponds which may occur depending on the tailwater result in comparison to the pond headwater. I mention this because if you set the composite outlet structure to be forward flow only, the results are improved as well.

    As for the structures that are flooding, given the amount of inflow, that may not be a surprise. I did not that some catch basins that are flooding at times to not have a surface storage method set. You may want to revisit this as well.

    Lastly, the surface storage will be right at the node itself and will not be considered to be spreading to other parts of the system. Our development team is looking into adding 2D modeling from OpenFlows FLOOD into SewerGEMS. If and when this is available, this may provide more accurate results in terms of the extent and depth of the flooding that occurs for the 2500 year storm.

    Regards,

    Scott

  • Additionally, from your last comment above I understand what you are saying about not being able to model the 2D flow. Is there a way to at least quantify a volume of ponded water at each catch basin or a ponded depth at each catch basin during the storm to have an idea of what that flooding quantity looks like?

    An additional question I have relates to the pond flex table column labeled (Flow (Out to Links Maximum). I'm not exactly sure what this column is showing. For example Basin A might show a flow out to links maximum of 150cfs, but the actual flow is limited by the pipe that the outflow structure is connected to, and for a 24" outfall pipe, will be something like 40-50cfs. Is the flow out to links maximum just the maximum flow that would leave the pond if allowed? Even though it is limited by the outfall pipe. Hope this question makes sense. Thank you for your time in reviewing this model.

  • Hello Joshua, 

    There is a Depth (Flooding) result, to get the depth of flooding. You would need to graph these results as there is no maximum flooding depth. I don't see a flooding volume in the properties, but if you look at the Calculation Summary and go to the Report tab, there is a Node Flooding Summary. 

    For the pond flextable, the maximum flow out to links is not the maximum allowable flow. if there are instabilities in the results, you may see something like this. I would try graphing the pond results, including the flow out of the pond. You may see some spikes in the flow, which is where I noticed some of the flow instability. There were spikes in the flow that will max the maximum flow appear to be higher than what it might appear to be based on other results, like the flow in the downstream conduit.

    Regards,

    Scott

  • Thank you for your reply. I have posted an updated model in a reply above, and I see what you are saying about the outflow curve for the basins showing instability (See example attached image). Do you have some suggestions to try with the model in order to get the basin outflow curves to show stability. The suggestions from previous replies have help get the continuity error looking good. There are still some issues with convergence and also with the outflow from both Basin A and Basin B.

  • Upon further review I was able to get the Basin B outflow curve to look much more stable by increasing my top elevation for the outlet structure weir a couple of feet. The same effect was not achieved by raising the Basin A composite outflow structure, and upon further review I noticed that the outflow curve for Basin A look unstable in all of the proposed model scenarios (2, 10, 100, 2500). Not sure what else to try with Basin A.

  • Hello Joshua,

    With a routing time step of 0.5 seconds and an output increment of 0.1 hours, the results are a little better for Basin A. Basin B does look much improved. 

    Based on your note about changing the weir elevation for Basin B, I tried that with Basin A as well. I set the weir elevation to 3180 feet. When using the calculation options mentioned above, the results were improved. I am not sure if this is a viable workaround, but the stability does appear to be sensitive to this. 

    If that is not a viable workaround please let me know.

    Regards,

    Scott

Reply
  • Hello Joshua,

    With a routing time step of 0.5 seconds and an output increment of 0.1 hours, the results are a little better for Basin A. Basin B does look much improved. 

    Based on your note about changing the weir elevation for Basin B, I tried that with Basin A as well. I set the weir elevation to 3180 feet. When using the calculation options mentioned above, the results were improved. I am not sure if this is a viable workaround, but the stability does appear to be sensitive to this. 

    If that is not a viable workaround please let me know.

    Regards,

    Scott

Children
No Data