water flow control valve

panthripu

Member
Join Date
Mar 2004
Posts
228
Hello everyone
I need some suggestion from you. We are controlling the flow of water in one of application. Our problem is the set point is fluctuating and due to this the valve is continued hunting and hence the flow is also fluctuating. Our requirement is not very critical that the flow should be exactly same as setpoint +/- 5 to 10% error is acceptable. Any ideas , how to tackle it.
Note: We tried filter in PLC for the input signal, tried to adjust damping the valve. But the results are not desirable.
 
the setpoint is fluctuating? You mean the flow is changing and your valve is not responding to adjust within a suitable time?

Sound like your PID settings need tuned better

Does the control valve have a positioner (intelligent controller) or just 4-20mA /3-15 psi ?
 
Hello everyone
I need some suggestion from you. We are controlling the flow of water in one of application. Our problem is the set point is fluctuating and due to this the valve is continued hunting and hence the flow is also fluctuating. Our requirement is not very critical that the flow should be exactly same as setpoint +/- 5 to 10% error is acceptable. Any ideas , how to tackle it.
Note: We tried filter in PLC for the input signal, tried to adjust damping the valve. But the results are not desirable.

I am going to assume that your SP actually is fluctuating, which would indicate a cascade loop in which a different loop controller is establishing the water flow loop's setpoint.

I have encountered a similar situation in which a water flow loop's setpoint was mathematically derived to maintain a specific ratio with the amount of dry material that was being added to a mixer by a weigh-feeder belt conveyor. I was able to achieve more stability in the water flow loop by deriving the water flow loop setpoint via the weigh-feeder's SP, rather than from the weigh-feeder's CV. I also implemented SP deviation alarms that would halt the process in the event that either the water flow or the solids delivery went too far out of spec.

I don't know if any of that helps in your situation, but if your SP is very unstable, I would think that you'd get better results by finding a way to stabilize the SP changes somewhat.
 
I don't know what the problem is - if the flow loop SP is changing, then it is right and proper that the valve should be fluctuating, and hence the water flow.

You could slow-down the valve fluctuations by giving the PID less Integral gain, and remove any Derivative, but it will still try to follow the SP, only slower.

You could "fix" your flow control valve at a steady position for a number of seconds at a time.

You will need to accumulate total flows over time, and step-change your fcv setpoint up or down according to the error.
 
If the problem is that the valve hunts after a change in position command you could try to open up the deadband on the positioner.
 
control valve

Set point is variable, its process requirement.
If a fixed set point is used, the valve responds very nicely. But in normal operation , when the valve tries to reach to the set point , by this time new set point is recieved. And this change in set point is within 2 to 3 seconds.
 
If these changes in set-point are not to big and given the tolerances you posted above, I would consider filtering the set-point changes as others have suggested.
 
How fast do update the PID loop?
There is the delay between sending a setpoint and seeing the first indication of a change in the feedback. This is called dead time.

You can measure deadtime. Let the loop stabilize. Then make a change to the control output. Then measure elapsed time until you see an observable change in the process variable (PV).

Fr'instance:
If the pump takes 1 second to change flow such that it can be detected by your PID loop (PV), then your PID loop cannot respond any faster than one second. After any update, your logic must wait at least the deadtime, preferably longer, before making another update.

If the PID loop updates faster than deadtime then it is responding more than one time to the original PV. This might cause the control output (CO) to oscillate or hunt.

If integral gain is applied, then the output is phase shifted, and it will be more likely to oscillate.
 

Similar Topics

I need to control DI water flow rate at about 45 kg/min. The water come out from a day tank and water will pass a centrifugal pump, a E+H Mass...
Replies
32
Views
17,819
Hi all- Looking for a stand alone valve/controller to regulate wastewater flow to 100gpm. would like to make it a self contained as possible...
Replies
6
Views
5,084
D
I'm looking for some general guidance on creating a flow meter controller in GTDesigner3. The pump supplier said that the inverter can be...
Replies
27
Views
7,016
Dear, I am looking for a simple solution showing the time of flow throughout the day, during the 24 hours, when is the flow passing through a...
Replies
2
Views
1,216
All, My client is looking to have me calculate the boiler feed water differential pressure in kPa to and actual flowrate in m3/hr. Once the...
Replies
2
Views
1,483
Back
Top Bottom