Comparison of Feedback and Feedforward Control Systems for a Counter-Current Heat Exchanger

The aim of this study is to provide a simplified approach to a process modeling and control framework applied in a shell-and-tube heat exchanger. Initially, 1st order differential equations (ODEs) are developed in order to analyze the dynamic behavior and interactions of cold (coolant water) and hot (industrial flue gases) streams (a lumped-parameter model was used). After performing open-loop simulations that are required in order to verify the linear equivalence and the extraction of process transfer functions, a control analysis is then applied. Specifically, a series of P/PI/PID controllers tuned via known methodologies (Ziegler–Nichols/Z-N and Tyreus–Luyben/T-L) are compared under different scenarios: (i) a constant set-point (reference) signal, (ii) a variable set-point (trajectory) signal and (iii) a constant set-point along with the effect of disturbances (temperature of the inlet hot and cold streams and hot stream flowrate). Based on performance criteria (ISE, IAE, ITAE, ITE) and control features (e.g., overshoot, settling times, efficiency), the best controllers are identified and proceed to a further comparison with (i) an internal model-based feedback and (ii) an internal model-based feedforward controller. Simulations and a comparative analysis take place in order to determine the best among these controllers for the applied heat exchanger. The results drawn from the simulations conclude that the most suitable controller for the heat exchanger is the feedforward as it provides a fast response (settles at nearly 120% lower time as compared to T-L tuned controllers and >400% as compared to Z-N tuned controllers), nearly zero overshoot (whereas other controllers exhibit a 8–10 K temperature overshoot) and reduced effort (the rate of changes is the lower among all controllers) from the manipulated variable (coolant water). The next best controllers are the internal model-based feedback controller and T-L tuned PID.