Filter Flow/Level Control

[Bigginhk]

Does anyone have experience of two stage filter flow/level control where control on second stage is unstable.

Control basis is inlet distribution channel level control cascading to common filter outlet flow control (this is not a physical hydraulic split).

I am looking for some advice on how PI controllers should be set up on second stage if anyone has an opinion.

First stage filter control is fine but second stage control is very sensitive to slight upstream variations and tends to oscillate levels and flows for several hours.

Thanks in advance.

 

[brianslater]

Hi Big in hk?
Its difficult on this type of forum to examine these problems without some kind of P&ID sketch, I can only hazrd a guess at your problem. I assume this must be a very slow acting process? If the flow rates are reasonably stable over the long term (say +/-10%) you could try turning off the integral action altogether and running on proportional only. That should remove any tendency to oscillate. you may have to adjust the set point to compensate for any offset (in the old days of puff puff instruments you had a manual loading station for this job). Its a bodge, not very elegant but might work for you?

 

[alexkoo]

Hello, I am an electrical engineer of Water Supplies Department(Hong Kong) I may not interpret your question fully. The normal control for individual filter outlet is based on a PID controller(Bailey Fisher or any supplier) with constant flow setpoint with output signal modulating the filter outlet valve so as to compensate any loss of head across the filter media.

The master setpoint of these individual PID controllers will then be cascaded to a common inlet level PID controller.

This would achieve constant flow in individual filter and yet a constant level in the common water flow inlet.

 

[Bigginhk]

Thanks Brian, the system is very slow, up to 10 x 150m2 filters which take a long time to register any change in level. The period of the incoming flow from the outlet of the upstream filter plant is long as is the second plants response which I believe contributes to the instability. The flow variations can be large depending on upstream load changes, especially during filter washing on the first stage. I hope to look into trying proportional only and see what this does.
I am sorry that I cant be more specific but as you rightly point out, without P&ID's, FDS's etc it is difficult to get a full understanding of the problem. However, thanks anyway.
Alex, the filter control is inlet level cascading to filter outlet flow controllers. Due to the slow response of this system the control of level is slow which leads to a certain degree of instability when the inlet conditions vary for long periods. The instability is not severe in my opinion but I am faced with a others who disagree and I need to try and identify any avenues that could help reduce it.
You might be interested to know that this is one of your own plants (Tai Po WTW). If you want to discuss it in more detail I would be willing to chat with you.