cascaded loops

I've got a chlorine dosing application where I'll use ML1100. The specification I'm working to defines that I use "compound loop control" I have two measurable variables - water flow and residual chlorine. Can anyone point me to an example or offer up any advice? I think that the loop update on the residual chlorine side would have to be very slow and that I'd use flow as a feed forward term.

You probably want to flow pace the chlorine feed rate for an initial dosing rate of so many ppm chlorine. You can then use feedback from the chlorine sensor to modify the dosing rate to maintain the correct residual. Keep trackof the deadtime in the process, as that will affect tuning.

Tom,
Thanks for the feedback! When you say "deadtime" are you referring to the time it takes for the chlorine level to "settle"? Pardon my ignorance.
Can I "flow pace" with a PID instruction? Would I then use another PID to modify the "flow pace" set point?
John

I am sure someone will correct me if I am wrong, but deadtime would be the amount of time it takes for you to see a change in the PV after increasing/decreasing the CV. As far as the term flow pace and the description above I take it as starting with a safe flow rate of chlorine and then modify that value (flow set-point) based on the reaction you see.

Here'e where I'm headed.....
I'm going to sample the residual every three seconds and load it into a 20 element stack. Each minute I'll average the stack and stuff the result into another 30 element file. Each hour I'll average that stack, compare to setpoint and adjust if needed.
I'm thinking I can use PID to "flow pace", but have not pecked that out yet.