Processes with and without self-regulation

Basically, the controlled process can be a process with self-regulation or a process without self-regulation.


Processes with self-regulation

On Picture 1 is shown the controlled process as example for a process with self-regulation. In this example the controlled variable is temperature. If we set an arbitrary output level for the controller in manual mode, and wait until the process value has stabilized, the controlled variable (the process variable) will always be proportional to the output level. We can record the static characteristic of a process - this means that we need to express the process value as a function of the output level. Then, we can see that this function will be non-linear in most cases. For that purpose, the output level is increased in 10% steps, and held until the temperature has stabilized. We will find that the temperature increments for each step are larger for lower temperatures than for higher temperatures. The characteristic is non-linear! The non-linearity is one of the reasons why the controller parameters may have to be altered for different setpoints, in order to keep on getting a good control-loop response, because for different ranges of the process variable we have different characteristic (different slope).

Picture 1: Process with self-regulation (non-linear characteristic)


Processes without self-regulation

A process that does not have self-regulation typically responds to an output level with a continual change of the process value. The process value deviation depends on the process characteristics and is proportional to the output level and the time. On Picture 2 are shown the response of a process without self-regulation, which does not have delay or dead time elements, and a block diagram for this process. If the output level for the process is 0%, then the process value remains unchanged. If, for instance, the output level makes a step change, then the process value also starts to change. This change is faster if the output level is higher. Because of the integrating effect of such a response, such processes are known as integral processes or I processes. If an output level is applied to a process that does not have self-regulation, the process value will typically keep on changing until it reaches some kind of limit. For a constant output level we can write:

Δx = KIS * Δy * t

where the parameter KIS is the transfer coefficient for the process without self-regulation.

Picture 2: Process without self-regulation, step response and block-diagram (symbol)


The best-known example of a process without self-regulation is the filling of a tank with an inlet and an outlet valve. The filling level process is shown on Picture 3. The outlet valve, which represents the disturbance, is closed. If the inlet valve is now opened and left at the same setting, then the level in the tank, which is actually the process variable, will gradually rise, steadily and continuously. The level in the tank will rise faster if the flow rate is higher. The level will rise until the tank overflows. So, we can state that there is no self-regulation in this case. Even after a disturbance, e.g. if the outlet is used, there will be no stable state, as there would be for a process with self-regulation. However, there is exception if the input flow is equal to the output flow.

Picture 3: Filling level process