Flow regulation problem

  • Thread starter De Maeyer Ronny
  • Start date

Thread Starter

De Maeyer Ronny


I am looking to the following control problem:

We have a water supply via a pump, the flow is regulated by a flowcontroller (frequency regulator on the pump) who gets his setpoint from a level controller. The water from the pump is fed to four vessels which have all there own flow controller. The goal is to open all four flowcontrollers on the vessels maximum and regulate the pump to minimum energy. The problem is that the flow to the four vessels is not equal due to different pressure losses in the collector.
Does any one has an idea how to solve this problem?

Thanks in advance,

Ronny De Maeyer

Ralph G. McDonald, P.E.

Without knowing more about the system dynamics ( static and dynamic heads, etc )here is my suggestion.

First, use the existing flow controllers with tank level transmitters at each of the 4 tanks to regulate and even shut off flow as each tank approaches its full level setpoint.

Retransmit the 4 tanks levels and use the current minimum level to regulate VFD pump speed. This will allow the system to adapt to various flow rates through each of the 4 tanks.

For example if Tank D is the controlling tank, and flow out of Tank B increases, it's flow control valve will open to maintain it's level, more total flow will be required to maintain level in Tank D and the pump speed will be increased. If flow through Tank B increases enough, it's level may become the speed controlling variable, and the Tank D flow control valve will then control Tank D level.

One flow control valve should be fully open at all times so you can minimize pump speed and pressure to that required to maintain total system flow, while maintaing all tank levels at or near setpoint.

Good Luck
Send the four valve opening signals to a high signal selector and let its output (the highest of the four) be the measurement of a valve position controller, which can be set at 90%, 95% or 100% depending on the nature of the process and let it control pump speed. This way, none of the valves will ever be fully open, but they will keep opening until the one in the line with the most restriction will reach nearly full opening. This configuration results in minimum pump speed and therefore saves pumping energy. It is a
standard these days on most water distribution systems and is shown in Figure 8.10m in the 2nd volume of my handbook.

De Maeyer Ronny

The solution you proposed is not so very clear for me, am I correct when I say it like this:
We take the highest valve output and regulate the pumpspeed to optain a valveposition of e.g. 95%, but are the flows then equally distributed over
the four lines? The main problem is the equilly distribution of the flow over the 4 (sometimes 2 or 3) take-off lines. Maybe a drawing could clear up the solution, can you provide me with more
information about the book you where talking about, title or ISBN Number
Thanks in advance

Yes Ronny,

The 4 flow controllers throttle the 4 valves to maintain the 4 flows the same AND the valve position controller keeps the "most open valve" at 95% opening, by throttling pump discharge pressure (speed).

Best regards, Bela


I belive that should revise the piping net diameters. Could be that some piping after collector has a different diameter and this might
create a restriction.

Ana Maria

Martin Rimmele

I think that if you want to have equal flow to the four vessels you must compensate the different pressure losses making a regulation
on the valves open percent. Take one to 100% open and regulate the others. The goal is compensate the differents restrictions that you might have.

Excuse my english.

Having solved your problem, no more help is needed.
Simple comment:
In one of the project, I was hired sizing all the valves only, I was the third person doing that (there were about 400 valves), I found quite few mistakes. Every time I was ready for one valve then I sat down with the pump/piping Engineer. He saved on pump size (energy) and some expensive stainless stell piping.

Just to show that conjoint optimum sizing was the right way and finaly technical.

In fact, having done your project using same approach is the solution because your valves work near 100% open. I call your system over instrumented. I may be wrong on grounds that escape to mind.

One detail may be worth mentionning.
The dynamic of a collector is not as simple as it looks, as each branching point develops a negative pressure (several psig in your case).Therefore, the presure each valves sees is not what one imagines first. It affects the valves flow significantly if the four valves are for instances butterfly valves.
[email protected]