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

Watercad - Looped System Modeling

I have a situation where a site is located at the intersection of two streets.  There are public mains in both streets and we are examining what would be the result if both mains were tapped.  We have hydrant flow tests for hydrants on both streets.  Typically only one hydrant is used and converted to a pump to model the system.  Would the model be any more accurate if both hydrants were modeled as pumps.  The thought here is that water is coming from at least two sources (directions).  Modeling back to the water source (water plant) just isn't an option.  Any thoughts on this?
Parents
  • You can only replace a full system model with a pump/reservoir if the hydrant flow test is done under conditions similar to those under which the site will operate.  So, if you want to represent two hydrants simultaneously as pumps, then the flow test on which you base the pump curves had to have been performed with both hydrants open simultaneously.  As you know, two pumps working in parallel perform differently than either pump operating alone and cannot be simulated simply by adding the two pump curves together.  The same is true here: two hydrants (near each other) flowing simultaneously will not flow as the sum of the individual flows the hydrants produce when opened by themselves.

  • Great answers Kim, I have a few approaches to add - first let me rank the approaches in preference.

    • 1) The full model with all controls is always best to determine the most accurate performance (this involves a good deal of set up, controls, and data entry of course). - Higher degree of accuracy at the cost of time and/or budget.
    • 2) If the full model is not available, developing a skeletal model from the node of interest to a consistent boundary condition(s) is highly advisable. This way you can still answer from which of the connections on either side of the street will more water flow from? Which connection will be limiting? Will they be similar in delivery? However to do this, the modeler still needs to have an understanding of the limited network delivering water flow/pressure to the site of interest., or us the powerful Skeletonizer tool within WaterCAD/GEMS - slightly reduced accuracy at a potentially lower cost
    • 3) When no utility model and no skeletal model are available the fall back is the collection of hydrant flow data (at the appropriate time of day) to mimic the system behavior at that point(s). Once you have boiled an entire model down to 2 points of delivery, a lot of intelligence has been taken out (controls, boundary conditions, which source is limiting or dominating?, so to say that the modeler will be able to accurately determine which source will deliver more flow (maybe determine the size of the connection pipe) may be an overly simplified task. In this simplistic form, two connections modeled as a pump/reservoir have the potential to be misleading - are the fire hydrants on the same watermain? Are one on an 8" watermain and one on a 12" watermain; is one at a higher elevation entering the site of interest than the other? These are assumptions the modeler has to make given the absence of a full or skeletal utility model to arrive at a conclusion. - reduced accuracy and increased assumptions but has lower time investment.


    In summary, I have in the past simulated one connection at a time (pump/reservoir) if I did not have enough information for a full or skeletonized utility model and avoided modeling 2 or more connections at the same time due to the points mentioned in 3). I would then confirm each pipe could deliver the necessary fire flow volume independently as a degree of factor of safety. Also don't forget to make sure the fire flow data is current and taken during high demand periods (worst case scenario).


    This is a test

  • I have a similar situation.  I have a pretty large development that I am modeling & it connects to (2) different systems which come from 2 different water towers.  The connection points are separated by about 3 miles.  Is it realistic to model 2 pumps at each tie in location?  The pump curves would be modeled off of flow data from fire hydrants.  What would be the correct way to model this situation?
Reply
  • I have a similar situation.  I have a pretty large development that I am modeling & it connects to (2) different systems which come from 2 different water towers.  The connection points are separated by about 3 miles.  Is it realistic to model 2 pumps at each tie in location?  The pump curves would be modeled off of flow data from fire hydrants.  What would be the correct way to model this situation?
Children
No Data