Optimization and control of heat exchanger network under uncertainty

Abstract

The dynamic control of heat exchanger network is important for developing energy efficient and safe industrial processes. In Chapter 3 of this work, the uncertainty is considered in inlet temperature of hot stream. The cold stream is bypassed around the heat exchanger. This work aims to track the setpoint temperature of the mixed stream by manipulating the bypass fraction of the cold stream around the heat exchanger. The implemented control is in Nonlinear Model Predictive Control (NMPC) framework. The uncertain optimal control problem (OCP) is dealt by using scenario tree based approximation along with an affine policy based method. The first principles model of shell and tube heat exchanger is used. The orthogonal collocation technique is applied to discretize the first principles model into the system of algebraic equations. The results show that for possible scenarios of uncertainty, the controlled variable efficiently tracks setpoint. In comparison, considering the same scenarios of uncertainty used, the deterministic optimization approach shows significant deviation of the controlled variable from the setpoint as time passes. Fouling is a concerning problem for heat exchange in industries. It is the deposition of unwanted materials on heat exchanger surfaces which offers extra resistance for heat transfer. Generally, chemical cleaning and flowrate distribution are used to mitigate fouling. In Chapter 4 of this work, the optimal cleaning scheduling and bypass control problems are formulated simultaneously considering disturbances in inlet temperature of cold stream. This integrated problem is formulated as a MINLP problem. The cleaning scheduling variables are binary decision variables, and the bypass fractions are continuous decision variables. A axially lumped and radially distributed model of HEN is considered in this chapter. The uncertain OCP is made tractable using an affine policy based methods and scenario tree based approximations. The performance of the proposed uncertain optimization problem is demonstrated using various case studies which include PHT of crude. This is because fouling is a relevant and most evident in PHT due to presence of impurities in crude. In this uncertain optimization, the value of objective function which represents the additional cost occurred because of fouling as compared to ideal clean conditions is reduced by 44% as compared to deterministic optimization.