Chlorine gas dosing

My situation:



Two measurements 1. water flow in pipe (4-20mA), 2. residual chlorine 30m after place of dosing (4-20mA).

One motorized valve: two commands (open valve, close valve) two end contacts for valve position (valve open, valve closed).



Problem is, I need to control valve position and dose correct amount (0,3ppm/l) of chlorine gas into the pipe. I decided to use PI regulator from siemens FB42 CONT_S, the problem is that I don’t know what should be the input value and what set point. I noticed if I’m using chlorine as input usually dosing overshot set point because water flow has significant impact in measurement at the end of pipe. How to incorporate these two values into one regulator?



Any ideas about this regulation?

Hi there

As I see it you have identified 2 Process Variables (water flow, and gas concentration) therefore there would be 2 Setpoints (i.e. target values for these variables). You also seem to have only one output, and that controls the water flow.

If the Cl2 concentration is the key process parameter here, what actually controls that? You also mention a dosing point - is there a fixed rate of Cl2 dosing being performed and the only method of adjusting the concentration is to increase/decrease the flow of water?

I think we need more information about the process and interaction we can have with it before recommending a control strategy.

Ken

tgaljar said:

My situation:



Two measurements 1. water flow in pipe (4-20mA), 2. residual chlorine 30m after place of dosing (4-20mA).

One motorized valve: two commands (open valve, close valve) two end contacts for valve position (valve open, valve closed).



Problem is, I need to control valve position and dose correct amount (0,3ppm/l) of chlorine gas into the pipe. I decided to use PI regulator from siemens FB42 CONT_S, the problem is that I don’t know what should be the input value and what set point. I noticed if I’m using chlorine as input usually dosing overshot set point because water flow has significant impact in measurement at the end of pipe. How to incorporate these two values into one regulator?



Any ideas about this regulation?

Click to expand...

When dosing chlorine 30 meters away, the pipe line to sample point may hold 5 minutes of water, plus to get the chlorine from motorised valve to dosing point, may also be another 5 minutes, that could mean a 10 minute delay from dosing change at the motorised valve to seeing that change take place at the residual chlorine analizer, this requires an I component in the PI controller measured in minutes, to reduce the normal oscilation of the P component of the PI controller.
If the water flow is constant it will have little effect, but if it is constantly changing, then one will effect the other and will require care in implementing.

We dose our systems using sodium hypochlorite in liquid form using a metering pump for continuous dosing but the principal is still the same. The best way we have found to operate this system is to use a feed forward element proprtional to the flowrate as our main control strategy and a PID element to "fine tune" the concentration at the sampling point. The lengthy time delays will make simple PID control almost unworkable unless you have a very consistent water flowrate. I assume that the motorised valve is modulating rather than just on/off otherwise your dosing will be intermittent and the system can not work effectively.
Andybr