Predicting the products of crude oil distillation columns

UoM administered thesis: Master of Philosophy

  • Authors:
  • Jing Liu

Abstract

Crude oil distillation systems, consisting of crude oil distillation columns and the associated heat recovery systems, are highly energy intensive. Heat-integrated design of crude oil distillation systems can provide opportunities to find the energy-efficient design solutions. Shortcut distillation models, based on the Fenske-Underwood-Gilliland model, have been applied to model the crude oil distillation columns, taking advantage of their simplicity and robustness in convergence. However, product specifications in the petroleum industry, related to boiling properties (e.g. true boiling point temperatures) and flow rates, have to be translated to those required by shortcut models, namely the key components and their recoveries. However, the two kinds of product specifications are so different from each other that 'translating them' is a very challenging task.In this thesis, an optimization-based methodology for transforming the product specifications used in industry and to those for shortcut modelling is developed. This method is based on the Fenske distillation model; it can automatically identify the most appropriate key components and the associated recoveries that characterize specified separations. The proposed method may be applied to simple columns and atmospheric distillation columns. Case studies demonstrate that the product results predicted by the method, in terms of boiling temperatures and flow rates, are in good agreement with those obtained from the rigorous simulations. Compared to the existing methods (e.g. method of Chen (2008)), the method is simpler, such as column design and energy balances are not required, and much more robust in convergence. Moreover, the method is applicable to the heat-integrated design of crude oil distillation systems, especially in the optimization framework involving shortcut column models, e.g. Suphanit (1999), Chen (2008).The proposed method is applied in two optimization contexts: one optimizes a particular product flow rate in a crude oil distillation column; the other maximizes the total product income of a crude oil distillation column for given product unit values. A stochastic method, Random Optimization, is applied in the maximization of total product income. True boiling temperature constraints are considered in these optimizations. Case studies illustrate the application of the two optimization methods, and the key components and recoveries associated with the optimal solutions can be easily identified.

Details

Original languageEnglish
Awarding Institution
Supervisors/Advisors
Award date1 Aug 2012