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

pumps in sewerGems

Hello everybody,

I use the Explicit SWMM solver dynamic wave as a solver in SewerGems to model two pumps in parallel. But I see that the solver doesn't account for losses and uses only the last point of the pump curve. My question is to know if is it possible that the solver takes into account the curve of the pumps as input (taken into account the pressure drops) and is not limited only to the last point of the curve?

Parents
  • Hello Julia,

    The pump operation follows the pump curve setup in the pump definition. How are the parallel pumps set up? Pumps in parallel should have the same curve so as to work in parallel. What pump definition are you using?

    If you can share a snapshot of how you have set up your pumps in parallel it would help us understand the operation.

    Hope this helps.


    Regards,

    Yashodhan Joshi

  • Hello Yashodhan,

    I use the pump definition type : multiple point (type 3). The two pumps don't have the same starting and stopping levels. Here are the parameters of each pumps. They don't have the same pump curve.

    Thanks

  • Julia,

    Although not typical, it is OK to have pumps in parallel with different pump curves. The pumps should operate based on the system conditions, even with the SWMM solver.

    When you say that they operate at the "last point", what do you mean exactly? Are you saying that they are operating at the highest flow point? This might indicate that the pumps do not need to add much headloss to overcome friction losses in the downstream pipes. Try looking at a profile of the system from the pump to the downstream gravity discharge point (or intermediate high point if one exists).

    If the pumps are not operating as expected, try toggling the "is virtual" option in the parallel piping, as this can sometimes work better, with the SWMM solver. You may also need to use a smaller Routing Step. Lastly, try using the 3-point pump curve type, with values set to approximate your multi-point curve.

    See the second-to-last bullet point in this article: Troubleshooting unstable SewerGEMS and CivilStorm model results using the Explicit SWMM Solver

    There are some additional troubleshooting tips in the above article. If you need further assistance, you can provide the model file for review: Sharing Hydraulic Model Files on the Haestad Forum


    Regards,

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

  • Hi Jesse,

    I share my model "Farmer.zip"  as confidential files for more informations. 

    Thanks,

  • Hello Julia,

    I took a look at the model you uploaded and here are some of my observations and recommendations.

    • There are 4 scenarios that use a the SWMM solver and the parallel pump configuration. I looked at the scenario "DynamicSWMM" as that was the current scenario when you last saved the model.
    • The pumps that are in a parallel configuration are using pump definitions "Farmer 1P" and "Copy of Farmer 1P".
    • When I validated the scenario "Dynamic SWMM", a red user notification for catchment "CM-023" was returned stating "Percent Impervious must be greater than or equal to 0.0 percent or less than or equal to 100 percent". The field "Percent Impervious" for this catchment was set to 433.4%. I changed this value to 0%.
    • There were also 10 yellow user notifications for conduits stating "The crown of the link adjacent to the Start Node is above the top of the node". I recommend checking all of the input data in your model.

    Try making the changes listed above and then compute the model again. If you are still seeing unexpected results, please upload an updated model and elaborate on what the current issue is (ie. what you are seeing and what you expect to see). It would also help if you would provide scenario you are looking at, relevant screenshots and steps to reproduce.

    As Jesse mentioned, the pumps should operate based on the system conditions.

     

    Regards,

    Craig Calvin

    Bentley Technical Support

Reply
  • Hello Julia,

    I took a look at the model you uploaded and here are some of my observations and recommendations.

    • There are 4 scenarios that use a the SWMM solver and the parallel pump configuration. I looked at the scenario "DynamicSWMM" as that was the current scenario when you last saved the model.
    • The pumps that are in a parallel configuration are using pump definitions "Farmer 1P" and "Copy of Farmer 1P".
    • When I validated the scenario "Dynamic SWMM", a red user notification for catchment "CM-023" was returned stating "Percent Impervious must be greater than or equal to 0.0 percent or less than or equal to 100 percent". The field "Percent Impervious" for this catchment was set to 433.4%. I changed this value to 0%.
    • There were also 10 yellow user notifications for conduits stating "The crown of the link adjacent to the Start Node is above the top of the node". I recommend checking all of the input data in your model.

    Try making the changes listed above and then compute the model again. If you are still seeing unexpected results, please upload an updated model and elaborate on what the current issue is (ie. what you are seeing and what you expect to see). It would also help if you would provide scenario you are looking at, relevant screenshots and steps to reproduce.

    As Jesse mentioned, the pumps should operate based on the system conditions.

     

    Regards,

    Craig Calvin

    Bentley Technical Support

Children
  • Hello Craig,

    I share my model "Farmer.zip"  as confidential files. Here, my first concern is with the PMP-1 and PMP-2 pumps. There is no water in the wet wells and the pumps do not output any flow which. This is normal. But a look towards the pipes after the pressure pipe shows that there is a flow equal to the maximum capacity of each pumps. I would like to know how I can fix this problem?

    My second problem is with the PMP-4, after the cross-section PJ-4, the pipe has negative flow why? Is there a way to reduce these imperfections?

    Regards,

  • Hello Julia,

    Your pump PMP-2 is set to "On" but in reality, there is no inflow to the downstream wet-well (PosPom-005) i.e. there is no flow accumulating in the wet-well for the pump to activate and pump flow downstream. From your response to Craig, I think this is expected in your model. Am I right?

    However, since the pump status is "On", I think this unaccounted (additional) flow might be considered. Here is a reference article for this;

    Unaccounted for flow in SewerGEMS

    You can try the options provided in the above article and check the results.

    When the pump status is set to "Off" though, these results are not seen i.e. no initial spike in flow in downstream conduit.

    For the negative flow at PJ-4, I can suggest using a slightly larger output time-step ("SWMM Output Increment" under Calculation Options). This can give you betterr results.

    Hope this helps. 


    Regards,

    Yashodhan Joshi

    Answer Verified By: Sushma Choure 

  • Hello Yashodhan,

    Indeed, in my model, I do not expect water in the wet well for this scenario. I would wait to see my results when there will be more water.

    Thank you team for help.

    Regards,