چكيده به لاتين
Petroleum refining is one the biggest and most important industries in the world and it is main producer of transportation fuels and chemicals. Oil refineries are very energy-intensive and consume between 7 to 15 % of the crude oil input for direct or indirect heating. Natural gas condensate refining starts with condensate distillation unit (CDU) in which the gas condensate mixtures come in for preliminary separation. Persian Gulf star crude oil distillation unit is one of the largest energy consumers in the refinery. a small change in the operation and design of the unit could lead to considerable economic improvements. Many researchers have thus concentrated on decreasing energy consumption of CDU by offering different optimization methods. The main goal of this study is finding a more efficient sequence, both economically and operationally, that enables the separation of pentane and hexane as value-added products instead of the conventional process in the atmospheric distillation columns.
In previous studies, researchers used two different modules or sofwares for simulation and optimization tasks. They linked these two modules together by an interface that was time consuming, especially in sequencing studies in which many configurations must be optimized and simulated simultaneously. In order to avoid time consumption on linkage tasks, a distillation simulation code based on the inside-out algorithm has been developed. Therefor simulation and optimization take place in one single environment. In this article an automated process design procedure is applied that can generate, simulate and compare basic zeotropic distillation configurations. At first the search space generated by separation matrix that covers all probable separation scenarios depending on the number of products and column types. Then, the related separation matrix must be translated so that computational operations would be possible and distillation columns would connect to each other. The result of shortcut simulation is exploited as initial estimates for the rigorous simulations. Rigorous simulations are carried out by the new set of these parameters generated by the optimizer. In order to calculate the TAC of a sequence, simulation results are collected and returned to the optimizer. The new individuals are then generated by the genetic algorithm operators to target optimal conditions of satisfying product specifications through the lowest amount of TAC.
In this article, 17 simple and complex sequences have been investigated for the separation of condensate feedstock into CDU conventional products. The results show that direct configuration demonstrates the best performances over all evaluation parts of TAC, operating costs, capital costs and exergy losses. Also, two approaches of hydrogenation and extractive distillation suggested to be implemented on hexane unit in order to remove aromatic compounds from hexane.
