View Full Version : How flow control valves are simulated Negative pressure downstream

February 28, 2017, 11:29 PM
I assume that maybe the lossess in a flow control valve FCV are a function of velocity head (v^2/2g)

I have a 18m circular tank with 4.5 m initial water level and elevation 100m.
It is connected to a node with elevation 99.72 bia a 300 mm pipe.
The flow is regulated by a FCV D=300mm and flow setting to 462m3/d; velocity in pipe 0.08 m/s)
When I see the valve, Upstream pressure is 6.71 psi, and downstream pressure is -6.86 psi.

I'm not sure why do I get negative downstream pressure

Patrick Moore
March 1, 2017, 11:06 PM

A FCV methodology comes from EPANET and operates on a similar basis as a negative demand would, in that it determines what Head is necessary for that much flow to be inserted downstream of the valve. The model then compares if the head downstream is compared to the available head upstream. If the Rqd Head is < Avail Head the valve initiates headloss through the valve so the downstream head matches what was needed to induce that flow. If the Rqd Head is > Avail Head a warning is issued that the valve cannot maintain flow but will run fully open and deliver as much flow as the head upstream can supply.

So if the downstream head is not controlled by any known head or pressure sustaining valve you can get low pressures as that is the minimum head needed to induce that flow.

Generally if you have a single junction you would not want to do a FCV like this as it would have no control over the downstream head and results could be unusual as you saw. It would be better to simply model this as a junction demand which will force water at a specified value and pressure will be based on the system HGL. If the downstream head of the valve represents a known system with something controlling the HGL, then put a reservoir downstream set to the downstream HGL and you can then use the FCV to feed it and the downstream HGL of the FCV will be the water level in that reservoir plus the headloss from that location to the FCV. You can decide what you think works best, but to use a FCV to a single junction requires more than you have currently included.

If you have further questions please let us know.

March 1, 2017, 11:42 PM

The tank is connected to a loop. What I want is to control the outlet of the storage tank that goes to the loop to 462 cmd, so that the water in the tank will be empited (renewed) in 3 days), like the figure.
I don't know the downstream head, only the flow outlet the tank that goes through the pipe.Is it ok my assumption of FCV, or is there other alternative?

Patrick Moore
March 2, 2017, 01:45 PM

It sounds like the problem is that nothing is controlling the HGL of the downstream system. Are there any other sources feeding the "loop" area or is 100% of the supply coming from the tank upstream of the FCV?

If something is controlling the HGL downstream it will impact what the FCV calculated downstream pressure is based on the known head downstream, but if the head is not constrained by any other known head this could explain why the pressures are so low. Check into what else is controlling the downstream head as a first start and then that ma resolve your issue. If you need to you can always insert a small dummy tank downstream of the FCV and set it with a small diameter but with an HGL such that is reasonable for the region but connect it to your system with a very small pipe such as 0.1 in diameter, C=50 ML = 250 such that flow into the tank should generally be near zero. But this will allow something to stabilize the hydraulic grade line so that the FCV makes sense.