PID Loop on SIEMENS S7-200

Hi everybody,

I need some help on PID loops in a project. I have done this project for a pasteurizer and I control the temperatures in two compartments with PID loops controlling analogue pressure valves which in turn control two pneumatic steam valves. The project has been working perfectly well for more than a year. About 3 weeks ago I recieved a call from the customer and I visited the plant. Sometimes one, other times both PID loops behave in a very strange way, then the problem goes away for a couple of days and then back again.
The problem is that lets say I have a SP of 60 Celsius and the actual temperature is 59.9, the PID will start from 6400(closed valve) to 7000 then somewhere close to 8000 then suddenly 32000 then again back to 8000 and continues like this until temperature goes close to 60.8 where it goes gradually below 60 and starts again. The customer didnt change anything on the PID settings neither on his process.
The steam and air pressures are ok.
I am using Siemens S7-200 and the PID subroutines are locked by SIEMENS so I cannot see anything in this Subroutines.
Any ideas would be appreciated

How is your temperature measurement? Is it possible that the temperature measurement jumps to an invalid value, causing the PID to react. When the measured value jump back to where it should be, it might be within the dead band. Causing the PID to pause.

Just guessing here.

A possible solution, other than replacing the temperature measurement, could be to damp the temperature by a couple of seconds.

Hi Boneless
No the temperature is kept +/- 1 degree from SP. Is not the best behavior but its acceptable. What is very strange is that this behavior is not steady. The system may work perfectly one day then the next day one or both PID loops may have this strange behavior as if there is no Integral part in the PID and the next day everything works perfect again and so on. Please note that the system has been working perfectly for more than a year and this behavior started about three weeks ago without any change to blame either in parameters or in process.

I stand by that this may be caused by invalid temperature measurement. Especially since you are only using P control.

If measurement goes to 0 for one cycle, the ER becomes big. Causing your output to jump up. When the measurement jumps back to the correct value, as you said you have a dead band of 1. Making the ER 0, causing the controller to pause.

I have ran into this before, it might be hard to detect if it only happens for a couple of cycles at a time.

I agree with boneless. It sounds like your temperature feedback is disappearing for a short time. I would look for things like broken wires and loose terminals. Possibly the sensor itself or the analog input. The easiest way to confirm this is to trend the input, if that is an available option. Another option is to put in logic to record high and low input values. In pseudocde the logic would be:

IF (Input > Max_Input) THEN
Max_Input = Input;
END_IF;

IF (Input < Min_Input) THEN
Min_Input = Input;
END_IF;

Keith

I have both P & I. I have monitored the PLC program and the temperature measurement is more or less stable between 59.5 -61 degrees.
Even when temperature reading goes above setpoint 60 I still see the PID Loop output jump from 7400 to 32000 momentarily and then back to 7400 and then 6400 and so on until temperature reading goes to 60.8 where it stabilises to 6400 for sometime.
Now when this behavior starts it goes on for the whole day during the production.
Next day when production starts it may behave perfectly all day long

Step 7 Microwin has a feature called Autotune PID. If you open this feature when online to the PLC without necessarily starting auto tune, you can view in a graph the actual temperature as read from the analogue PLC input and you can also see the graph of the PID output. The measured temperature is almost a straight line so I don't think the problem has to do with measurement