View Full Version : Distribution System with no Flow

April 16, 2014, 01:08 PM

I have a question related to modelling a distribution system with intermittent flow from the pumping station.

The station is setup such that the pumps only turn on when the discharge pressure from the station drops bellow a set head, otherwise it will be off.

The problem i am encountering is that when i attempt to run it based on this logic, the system disconnects because there doesn't seem to be any water supplied (obviously because the pumps are turned off, isolating the reservoir from the demand junctions).

Would anyone have a recommendation on how potentially to model this scenario? I need this logic to replicate my station, as the pressure unfortunately isn't constant (as could be achieved with a PRV).


Shawn Huang
April 17, 2014, 09:25 AM
Here is another question to think - how to maintain hydraulic grade line when the pump is off? Are there any tank?

Patrick Moore
July 16, 2015, 09:03 AM

You have probably long ago resolved your issue, but I want to add a response for the sake of the forum.

There are a few things you might consider in this situation. First as noted by Shawn Huang above, what controls the HGL when the pump is off in the real system? Is it fed by PRV's or other tanks, or is it a closed system with a hydropneumatic tank at the discharge of the station? Knowing what the real system does when the pumps are off is the first consideration to understand before developing a model solution. If it is a closed system (sounds like it from the description) adding a hydro tank to the discharge of the facility will allow a "source" on the discharge side that will keep the system from disconnecting when the pumps shut off. Modeling a hydropneumatic tank One suggestion I would have is to put a small pipe with a check valve from the suction to discharge side of the pumps. This tank generally is small volume tank with a small diameter but tall height. It should be tall enough that it can account for the max pressures that shut off the pumps. The other idea would be to simply put a small pipe from the suction to the discharge side of the pump station and assign a check valve to that pipe. This pipe would allow a connection to a source to avoid disconnection, but would shut off when the pumps run. Making it small would limit the supply available through the line.

One thing I have noticed is that pressure based controls are tricky due to the mathematics in the model. Real world systems will require pressure drops to be observed over a small period before a pump is kicked on, but the model does not. This instantaneous nature of the controls can lead to pumps turning on and off in a model, so set your on and off pressures carefully. If the Pump station is small and feeding a small area, you could also simply put a pressure sustaining valve from the discharge back to the suction side that would cap the max pressure and potentially just leave the pump running especially if the controls are too problematic.

But based on your described setup using either a hydro tank or a small bypass line with a check valve could resolve your issue.

Here are Notes from the InfoWater Help on modeling a Hydro Tank:
http://forums.innovyze.com/bigquestion2.gifHow do I model a
pressurized pneumatic tank?

If the pressure variation in the tank is negligible, use a very short, very
wide cylindrical tank whose elevation is set close to the pressure head rating
of the tank. Select the tank dimensions so that changes in volume produce only
very small changes in water surface elevation.

If the pressure head developed in the tank ranges between H1 and H2, with
corresponding volumes V1 and V2, then use a cylindrical tank whose
cross-sectional area equals (V2-V1)/(H2-H1).

Innovyze Support