Rate Controller

[controlnovice]

I am trying to controller the upheat of a batch reactor which is also exothermic. I'd appreciate any input.

The issue I have now is:
One batch process requires a rate of rise to be around 1 Celcius up to one temperature, then 0.6 up to the next higher temperature.

I copied a control scheme from a previous project from another engineer. In that scheme, the temperature setpoint is ramped up at the desired rate of rise, so the setpoint is constantly moving up until it reaches the final desired temperature. The controller tries to match the PV to the moving SP.

In this instance, the problem is when the PV exceeds the SP, cooling comes on to try to 'match' the SP,which dramatically removes heat. Then, because the PV is well below the SP, the heat comes on again, but now the rate of rise of the PV can be undesireable becuase the controller is trying to reach the SP which now has no relation to the desired PV rate of rise.

Does that make any sense?

Is this something I can control using a controller with the SP being the actual desired rate? So that the PV is the actual rate of rise? If so, how do I stop this controller once the temperature SP is reached? Change the rate to 0?

Controlnovice

 

[jsummerfield]

If I understand your question, you may eliminate the ramp/soak algorithm with a simple setpoint. If this is what you are considering, you probably need this loop to be tuned "over damped". This will prevent the temperature overshoot but is slow to reach setpoint.

John

 

[controlnovice]

John,

Thanks for the reply. Yes, I thought of this also. If I overdamp the loop, what will happen when it starts to exotherm? I need controlled cooling at that time (not emergency cooling). I don't think I want to have this overdamped also.

I'm using the same controller for both heating and cooling in a split range configuration.

 

[m777182]

Try to search the web on TECHNOLOGIC TLK48. It is a complex controller that allows you to approach the set point through many ramp steps with auto tunning for optimum performance. Do not give up reading the instructions manual!
Good luck!
m777182

 

[FrancisL]

Ramping is not difficult, you can write your own algorithm or if the system has standard ones then use one of those.
As I read what you say, however it sounds like a tuning problem. The process gain when cooling is too high.
Damping is not the right solution, getting the tuning correct is. And that means ensuring that as the control switches from heating to cooling it does so with the same process gain.