Hello,
I'm currently working on developing scatter graphs using a link discharge time series (force main) in gpm vs a water level time series in feet (at the node where the force main begins) minus the node invert elevation. Part I (preliminary phase) of my analysis will have the flow loaded directly on the force main to develop the first scatter graph. This model will only include a force main and outfall (with a water level time series boundary condition). From the scatter graph, an estimated system curve is developed for the wet weather and dry weather flows. These system curves will then be provided to a pump station designer, the pump station designer will use this information to come up with a pump curve that meet the conditions provided by the system curve in phase I. Once that is complete, the pump station designer will then provide the modeler a pump curve using the system curve developed in part I (plus any additional losses). The pump curve will then we added into a revised model, Part II. This revised model will include the pump station modeled as a node, a pump with the pump curve from the designer, a gravity pipe and manhole upstream from the wet well where the flow will now be loaded (instead of being loaded directly into the force main). The pump and pump station attributes (diameter, depth, pump on/off elevations, etc.) will all be included. For this simulation run, instead of using the link discharge (more erratic when pumps are on) the pump discharge will be used vs a water level time series in feet (at the node where the force main begins, same place as part I) minus the pump operation level to develop the new scatter plot and system curves.
The results from part II are not what was expected. Once the pumps are added, there appears to be 2 distinct patterns (system curves) but they do not share the same shutoff head (as the curves during part I did). Also, the system curves before the pumps were added (part I) are much steeper than after the pumps are added (part II).
The flows are the same in both models and use a diurnal pattern. The model simulation duration is approximately 6 days and it includes 1 wet weather event (equivalent to a 2 yr storm) around day 3.
Has anyone else tried to use MU to develop system curves?
I know that when the flows are added directly to the force main, the impacts from the boundary condition in the more clearly observed.
Any suggestions will be greatly appreciated !!!
Thanks!