OPTIMIZATION OF HEAT EXCHANGER NETWORKS FOR A CRUDE DISTILLATION UNIT
ABSTRACT
A common problem that occurs in crude oil refining, and other chemicals manufacturing industries is the following: The production process generates numerous streams of fluids, each stream at a certain temperature in 0C known as its starting temperature. The temperature of each stream has to be changed to a different level known as its target temperature, for it to enter the next processing stage. If the target temperature of a stream is smaller than its starting temperature, then this stream has to shed some of the heat energy in it (i.e., it needs to be cooled) before entering the next processing stage; that’s why such streams are known as hot streams. On the other hand, if the target temperature of a stream is greater than its starting temperature, then this stream needs to be heated before entering the next processing stage; that’s why such streams are known as cold streams. This contribution provides a new methodology for optimizing crude oil distillation systems. The proposed approach determines the optimum operating conditions for the crude oil distillation unit, where the objective is maximum net profit, while proposing retrofit modifications for the heat exchanger network (HEN) that allow a feasible operation which gives a minimum utilities cost. To improve product profit, the yields of the most valuable products are increased, while considering product specifications, heat recovery and equipment constraints. An artificial neural network model (ANN) is generated to simulate the distillation unit, while the HEN model consists of a mass and energy balance formulated using principles of graph theory. The newness of this research lies in the simultaneous consideration of the distillation column and HEN models in the optimization algorithm, with the focus on profitability. Results show that an important economic improvements can be achieved.
Keywords: Crude Distillation, Product, Optimization, Heat Exchanger Network, Neural Network