View Full Version : reservoir with reverse flow?

January 6, 2017, 12:31 AM
I have a network "A" that receives water from a pipe from the city. The network is like a lateral loop that receives water from the pipe "M"
I have pressure data measured at "M"
I was planning to assume "M" as a reservoir with constant head
However, with such assumption results in water flowing from "A" to "M".; it should be the opposite.

How can I solve this problem? Is it ok to simulate the input node as rservoir, or it has to be a junction with negative flow (water input)

Patrick Moore
January 6, 2017, 04:07 PM

If there is reverse flow it is because the system head at "A" is higher than at your reservoir. Check your HGL assigned to your reservoir is correct as a first step, or that the HGL at "A" is not getting unusually high due to something somewhere else in the system. You may also wish to control the supply form that reservoir using a control valve (PRV, FCV) etc, if supply from that source is controlled in some way. Otherwise the model will allow a potential infinite flow from the reservoir, based upon how much the head in the reservoir would allow. Using a reservoir and a control valve is a fairly typical way most people model a purchased water connection as a PRV can set a downstream HGL to maintain and a FCV can specify an actual flow if flow controlled.

Negative demands could be used, but have drawbacks. For negative demands the model would calculate how much head would be needed to push the negative demand in, but this can easily create unrealistic head conditions if you are not careful. Using a Reservoir and a FCV allows you to regulate the supply and the HGL that is allowable. It also has the advantage that you can easily "turn off" the supply by closing the valve while a negative demand is not so easy to "turn off".

Hope this helps. Let us know if you have additional questions in this matter.

Patrick Moore

January 8, 2017, 09:49 PM
The PRV works fine