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 Flow Distribution

Good day,

I'm using Sewergems to simulate flow through a sewer treatment plant (using the Explicit SWMM solver). The flow split at some nodes do not make sense to me as it does not balance (see snip below). The example below shows a flow of 111.94 plus 0.32 MLD entering a node and 143.01 MLD going out. There are no additional inflows at the nodes and I'm not sure where the additional 37 MLD comes from. Does anyone have an idea why this is?

Im using SewerGEMS Connect Edition Update 3 (10.03.00.78)

Parents
  • Hello Johan,

    I would check the continuity error and other information in the calculation summary that is generated after computing the model. You can access this by going the Home > Summary as well. 

    If you have high values here and the setup of the system is correct, open the calculation options in the properties (Home > Options, double-click the active calculation options from the list). Find the field Routing Time Step and set this to a value like 1 second. This may improve the results. This and other troubleshooting steps for the Explicit solver can be found here: Troubleshooting unstable SewerGEMS and CivilStorm model results using the Explicit solver.

    If that doesn't help, we may need to see a copy of the model files.

    Regards,

    Scott

  • File is uploaded as WWTW.stsw_1_1182.zip

  • Hello Johan,

    Thanks for providing a copy of the model. We are still looking into this, but there may be some things for you to review. 

    First, there are some very steep slopes, such as CO-248(2)(2)(2)(2) and CO-248(2)(1) with slopes at or over 100%. While these and other less extreme steep slopes may not fix the issue, this sort of thing may have an impact of the stability of the calculation. One of these conduits (CO-248(2)(2)(2)(2)) as a length of 7.2 centimeters. Fixing this would help with the slope, but a review of other conduit lengths may be needed to, just in case.

    There is a diversion on conduit EffCH_4(1). This is likely no longer needed if you are using the Explicit solver. 

    There are several conduits where the crown is above the top of the start or stop node, or the invert is below the invert of the start or stop node. Like the slope calculations, this may not have a large impact on the results, but it is possible that fixing these may help with the stability. Conduit PST_1 is an example of this.

    There are also user notifications about headloss methods not being supported by the Explicit solver. This may have no impact on the results, but if there should be s headloss at the node, choosing a headloss method that the Explicit solver supports is recommended.

    If you do make updates to the model, please send an updated copy to us. Otherwise, we will continue to look into this.

    Regards,

    Scott

Reply
  • Hello Johan,

    Thanks for providing a copy of the model. We are still looking into this, but there may be some things for you to review. 

    First, there are some very steep slopes, such as CO-248(2)(2)(2)(2) and CO-248(2)(1) with slopes at or over 100%. While these and other less extreme steep slopes may not fix the issue, this sort of thing may have an impact of the stability of the calculation. One of these conduits (CO-248(2)(2)(2)(2)) as a length of 7.2 centimeters. Fixing this would help with the slope, but a review of other conduit lengths may be needed to, just in case.

    There is a diversion on conduit EffCH_4(1). This is likely no longer needed if you are using the Explicit solver. 

    There are several conduits where the crown is above the top of the start or stop node, or the invert is below the invert of the start or stop node. Like the slope calculations, this may not have a large impact on the results, but it is possible that fixing these may help with the stability. Conduit PST_1 is an example of this.

    There are also user notifications about headloss methods not being supported by the Explicit solver. This may have no impact on the results, but if there should be s headloss at the node, choosing a headloss method that the Explicit solver supports is recommended.

    If you do make updates to the model, please send an updated copy to us. Otherwise, we will continue to look into this.

    Regards,

    Scott

Children
  • Hi Scott. Thank you for the suggestions. I have attended to most of the issues as you mentioned but at some places in the model the flow balance is still not correct.  I have uploaded the new file as WWTW_REV1.stsw_1_1201.zip

    best regards

    Johan

  • Hello Johan, 

    After computing the model, flow continuity error is 4 % which is acceptable as it is less than 5%. However there are several conduits operating under pressure, e.g. CO-171, because these conduits are surcharged so they are operating under pressure. However none of the manholes are overflowing but conduits are surcharged, which means that system is not overflowing anywhere. 

    See below profile to see which conduits are surcharged and operating under pressure, if required you can correct the conduit sizes to reduce/avoid surcharging of pipes. The active scenario was 674 MLD SCADA. 

    As mentioned by Scott There are several conduits where the crown is above the top of the start or stop node, or the invert is below the invert of the start or stop node, the case is still there. It appears you modeling weir structures on conduit stop nodes

    Regards,

    Sushma Choure

    Bentley Technical Suppport

  • Hi Sushma, thank you for the reply.

    The conduits operating under pressure is how it currently operates, so at the moment it is not required to change the sizes. Regarding the crown above the start or stop of the invert I have changed all except 4 nodes (CO-47, CO-48, C0-49 and PST_1. If I drop the manhole inverts to the conduit invert the results does not make sense. Yes these conduits have weirs at the end that discharges into a channel. My biggest concern at the moment is the flow split at some nodes that doesn't make sense, i.e.MH-110 (although the overall continuity is ok). Do you perhaps know why this is?

  • Hello Johan,

    The slope on conduits CO-248(2)(1) and CO-248(2)(2)(2)(2) still have slopes greater than 100% and conduit CO-248(2)(2)(2)(2) still has a slope of 7.2 centimeters. If these are not accurate, please adjust the invert elevations for these. If they are accurate, what do they represent?

    If this doesn't fix the issue, please upload a new copy of the model with the changes.

    Regards,

    Scott

  • Hi Scott.. apologies, I rectified it (but its disabled in the current scenario so should not make a difference). Only look at the current 674MLD scenario. What I found is that if one changes the conduit entrance loss coefficients on CO-194, CO-197, CO-198 and CO-199 to 1 the flows balance out. I don't know why it doesn't balance with say an entrance loss coefficient of 0.25. I have uploaded the file again as WWTW_REV1.stsw_1_1201.zip