Override Control for Heat Exchanger Protection

control schemes are designed around the control objectives, process fluids in question, and equipment limitations. But let’s go into an overview of the big control schemes. If you’re a engineer, regardless of your interest in control theory, KNOW THESE SCHEMES . They will recur over and over again and are very important to understand:

The first is feedback control . ( Summary: Process Objective measurement directly sets inlet variable )

Temperature%20Control%20Loop

This is a steam driven heat exchanger (not that the pressure of the steam is very important here as well) with feedback control. This would likely be driven by a PID (proportional integral derivative) control scheme where the outlet temperature automatically manipulates the steam valve.

Obviously in almost all heat exchanger applications the OUTLET process temperature will be controlled and in most the inlet cooling/heating medium will be adjusted (albeit perhaps not directly in front of the changer) if it’s a process/process exchanger.

This is a popular form of control as no model is necessary to implement this scheme - deviations from setpoint in the outlet temp will yield corrections at the inlet steam.

Another form is cascade control ( Summary: Process objective measurement sets a secondary flow controller, which both controls disturbances in the objective measurement and external factors ).

Cascade control adjusts the PID of a secondary flow controller loop. The benefit here is that the additional PID loop manages secondary steam flow issues such as steam pressure changes or valve problems.

Override%20Control%20for%20Heat%20Exchanger%20Loop

Ok, now let’s get to Feedforward control . ( Summary: Process objective is set indirectly by the inlet variables, which are adjusted according to an understanding of the dynamic responses ). This is where things get more confusing:

Override%20Control%20for%20Heat%20Exchanger%20Protection

Essentially with feedforward control, the outlet process variable is not used (though it is common to use a feedforward-feedback control scheme, though I won’t talk about this here)

Feedfoward control responds to dynamic disturbances in the inlet process fluid. However, you will need a process model to effectively integrate feedforward action (you need to know how changes in the disturbances affect the control objective).

Let me get a little philosophical as well (maybe this will help with understanding how these work with life analogies). Feedforward control is what the business world strives for. The general idea is that we make changes BEFORE we see a problem - it’s the more proactive approach. But that’s also what makes it challenging - we need to build the domain expertise into the control scheme.

In essence, most business models and life processes should incorporate a form of feedforward control - you should make changes before you get screwed. In a heat exchanger set-up, we are trying to accomplish the same thing.

Author - Casey McNamara