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

Transient Model Boundary Conditions from EPS

Hello,

I have a working 36hr EPS hydraulic scenario and would like to setup a transient scenario for a pump trip.  My attempt in doing so has not been successful and I was seeing my zone draining and my pump station (that feeds the zone) dead heading, with flow reversing through the pump towards the boundary conditions.  

  • What are the steps to go from 36hr hydraulics EPS to 120sec transient EPS.  I tried using "initialize transient runs at time hours" for the desired initial conditions but my transient scenario has not been working.  I hope to get some guidance.  Do I need to remove my boundary conditions in my transient scenario and manually setup the initial conditions for the tanks and pumps based on the hydraulic scenario?  Also, I have surge relief valves that have pressure thresholds but the transient scenario is showing flow through them all the time.  
  • The model is giving me an error that the pump is not connected to a downstream pipe, even though, as shown in the linked screenshot, they are connected.  Please advise.  /resized-image/__size/320x240/__key/communityserver-discussions-components-files/5925/8446.Capture.JPG

Regards,

Youssef 

  • Thanks, Scott.  I will be on the lookout for upcoming training sessions on your calendar.  I have been referring to the Youtube videos, which have been helpful.  

    Got it, that explains it.  Thanks again.  

    Regards,

    Youssef 

  • Hello Youssef,

    Bentley offers training for all products, including HAMMER. We post upcoming instructor-led classes on our calendar. You can also find HAMMER training material on our Youtube channel and Learnserver.

    For the results reporting, you likely have a user notification that says "The period between path histories has been increased by a factor of X to limit the number of path records to Y." This occurs because there is a limit to the size of the report file HAMMER will generate. To account for this, HAMMER will increase the reporting time step by some factor. You can reduce this in a number of ways, such as running the model for a shorter time or decreasing the number of report points. This link has more information on that.

    Regards,

    Scott

    Answer Verified By: Larry Abla 

  • Hi Tom,

    I thought I replied but I am not able to see my response.  Anyway, thank you so much for your help.  It ended up being the "smart pipe removal" skeletonization method that was messing up my output as I was not aware that it compromised the system hydraulic capacity.  Speaking of which, does Bentley offer any "rigorous" courses for us to truly grasp the software and its computational methods?  

    Another thing , following the computation of the Hammer model, the transient results viewer is showing me results every approximately 3.4 seconds in the "dynamic" graph.  Even though my report times are set to all and time step interval is set to 0.01 seconds.  Any thoughts?  Also, the model takes approximately 30 minutes to compute even though I selected the "critical" report points (approximately 1,000 pipes and junctions versus the approximately 3,000 in the skelebrated model).  Any thoughts on that as well?  

    Regards,

    Youssef 

  • A lot depends on which method you used to skeletonize the system. if you used anything but Smart Pipe Removal, you should have virtually the same results for the full and skeletonized model because the methods preserve capacity. Smart Pipe Removal doesn't do that.

    You apparently removed some important capacity.Try running some standard WaterGEMS runs with the full and skeletonized model and see where the differences occur.

  • Hi Scott,

    That was helpful, thank you.  Now I am stuck with a dilemma.  I tried skelebrating one zone in my model to decrease computation time of the transient model, however, once I do that and try to initialize initial conditions, I get a lot of errors with negative pressures at junctions.  This error was never present before skelebration.  But if I do not skelebrate it, the computation fails due to the model being too large.  Any thoughts on this? 

    Regards,
    Youssef