Flooding issue in Tanks Evacuation

I am modelling evacuation tanks in SewerGEMS, the tank presented as ponds in the model for volume purposes and to use the control structure, the model is shared in the Bentley sharing server is showing the Evacuation line scenario, in that scenario for example in case evacuate tank 3 to tank 5 the line between the two tanks are flooded in spite the water level balance in the tanks when they balanced is entirely below the ground elevation and the manholes, in reality, should not be flooded.

I am looking for the reason why that flood happened in order to solve the flood.

Regards,

Mohamad Azzam

Parents
  • Mohamad, this appears to be due to some initial instability. If you advance to the second timestep, the profile results will make more sense.

    It is common to encounter initial instability with dynamic modeling, and you may be able to reduce the impact by adjusting the advanced calculation options: Troubleshooting unstable SewerGEMS and CivilStorm model results using the Explicit SWMM Solver

    For example if you change "Use Bentley Transition Equation?" to "True" and set the Routing Step back to the default of 5 seconds, the initial instability is no longer seen in that profile and the %Non-convergence (seen in the Calculation Summary) will be improved.


    Regards,

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

    Answer Verified By: Mohamad Azzam 

  • Thank you Jesse for the support - "Use Bentley Transition Equation?" to "True" with a 5-sec timestep is solving the flood and giving reasonable results, however, I am getting flow continuity errors around 70% and I guess that because the downstream last tank has no outfall, do you think that is the reason?

    Alternatively; for the line connecting tank 3 and tank 5, if I use the initial water depth in the manholes equal to the water depth in the tank (without using "Transition Equation" will also solve the flooding issue, is this approach correct?

    As per modelling results, there is no flooding in this case but I do not know why flooding occurs and why these two approaches solve the issue, do you have a theoretical justification for that?

    Regards,

    Mohamad

  • Thank you Jesse for the support - "Use Bentley Transition Equation?" to "True" with a 5-sec timestep is solving the flood and giving reasonable results, however, I am getting flow continuity errors around 70% and I guess that because the downstream last tank has no outfall, do you think that is the reason?

    My suggestion was just an example to illustrate the sensitivity of the results to calculation option adjustments in certain circumstances. If your flow continuity error is still high you may need to use a combination of other adjustments as seen in the troubleshooting article provided. For example you may need to further refine the Routing Step, Surcharge Method (Slot vs. Extran), number of trials, and other options until the results are more reasonable. You may also need to take a closer look at the physical configuration of the model and avoid any extra complication where possible. Generally speaking the simpler the model, the more stable and reliable it will be.

    Alternatively; for the line connecting tank 3 and tank 5, if I use the initial water depth in the manholes equal to the water depth in the tank (without using "Transition Equation" will also solve the flooding issue, is this approach correct?

    Yes, if this approach also yields more stable results then it could be an acceptable alternative approach. There can be multiple possible ways to stabilize a model. You will want to look carefully at all the results to check for any adverse impact of any changes. A typical starting point is the %non-converged, flow continuity error and flow/HGL graphs at key points in the model.

    As per modelling results, there is no flooding in this case but I do not know why flooding occurs and why these two approaches solve the issue, do you have a theoretical justification for that?

    The flooding issue that you originally asked about appears to be due to initial instability as mentioned in my previous reply, which is typical at the beginning of the simulation. Combating instability is common practice with dynamic modeling and some engineering judgement will need to be used to decide whether results are acceptable and to decide how to proceed. For example knowing that the instability is only for the first few minutes of the simulation and accordingly the flooding is relatively insignificant, the dynamic modeler may choose to overlook this and advance the time browser to the next timestep after the initial instability. Refinement of the advanced calculation options is also very common in dynamic modeling, especially when working with hydraulically challenging situations. In your case you are starting the simulation with an initial depth in the ponds and dry pipes that suddenly need to fill up and balance with the pond levels, so it is not surprising to have to make a few adjustments to account for this challenging situation.


    Regards,

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

    Answer Verified By: Mohamad Azzam 

Reply
  • Thank you Jesse for the support - "Use Bentley Transition Equation?" to "True" with a 5-sec timestep is solving the flood and giving reasonable results, however, I am getting flow continuity errors around 70% and I guess that because the downstream last tank has no outfall, do you think that is the reason?

    My suggestion was just an example to illustrate the sensitivity of the results to calculation option adjustments in certain circumstances. If your flow continuity error is still high you may need to use a combination of other adjustments as seen in the troubleshooting article provided. For example you may need to further refine the Routing Step, Surcharge Method (Slot vs. Extran), number of trials, and other options until the results are more reasonable. You may also need to take a closer look at the physical configuration of the model and avoid any extra complication where possible. Generally speaking the simpler the model, the more stable and reliable it will be.

    Alternatively; for the line connecting tank 3 and tank 5, if I use the initial water depth in the manholes equal to the water depth in the tank (without using "Transition Equation" will also solve the flooding issue, is this approach correct?

    Yes, if this approach also yields more stable results then it could be an acceptable alternative approach. There can be multiple possible ways to stabilize a model. You will want to look carefully at all the results to check for any adverse impact of any changes. A typical starting point is the %non-converged, flow continuity error and flow/HGL graphs at key points in the model.

    As per modelling results, there is no flooding in this case but I do not know why flooding occurs and why these two approaches solve the issue, do you have a theoretical justification for that?

    The flooding issue that you originally asked about appears to be due to initial instability as mentioned in my previous reply, which is typical at the beginning of the simulation. Combating instability is common practice with dynamic modeling and some engineering judgement will need to be used to decide whether results are acceptable and to decide how to proceed. For example knowing that the instability is only for the first few minutes of the simulation and accordingly the flooding is relatively insignificant, the dynamic modeler may choose to overlook this and advance the time browser to the next timestep after the initial instability. Refinement of the advanced calculation options is also very common in dynamic modeling, especially when working with hydraulically challenging situations. In your case you are starting the simulation with an initial depth in the ponds and dry pipes that suddenly need to fill up and balance with the pond levels, so it is not surprising to have to make a few adjustments to account for this challenging situation.


    Regards,

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

    Answer Verified By: Mohamad Azzam 

Children
No Data