Studies and Modeling for Upgrading Units for Heavy Oil Refineries

The primary objective of this study is to predict the optimum operating conditions required to operate an existing atmospheric distillation column to distillate heavier crude oils in same unit which designed mainly to fractionate moderate and lighter crude oils. Detailed simulation model of atmospheric distillation column was made using Aspen Hysys V. 8.4 software. The model built depending on the data of real plant which located in Al-Daura midland refinery company at Baghdad/Iraq. The crude oils were described using the true boiling point (TBP) assays and Peng-Robinson package were used to predict thermodynamic properties. Three different types of crude oils (Kirkuk light, Basrah light and Basrah medium) were selected to identify the model validity. The simulation results agree very well with the industrial plant results. Finally the proposed model was used to predict optimum operating conditions required to distillate a blend of light and heavy Basrah crude oil with different mixing ratios. The objective of the optimization is to predict the maximum profit of the products within the required specifications. At optimum operating conditions the total yearly profit for distillation of a blend of 50% vol. light+50% vol. heavy basrah crude oil was found to be 713.28 M$.


Introduction
Crude oil composed of thousands of hydrocarbons varying from methane to very high molecular weight components, with varying proportions of paraffins, naphthenes, and aromatics [1]. The distillation column is main unit in a petroleum refinery which used to fractionate the crude oil into gases, light and heavy naphtha, gasoline, kerosene, diesel, and residue.
The properties of crude oils are varying making achieving the products specifications in distillation unit is difficult. Petroleum refining in Iraq face several challenges in the last few years, one of these challenges is increasing the demand to light fractions while decreasing the ratio of exploited light crude oils. Also varying the demand for light cuts within a year, the demand for kerosene is high in the winter while the demand for gas oil reach maximum in summer. David et al. (2010) [6] applied Aspen Hysys for analysis of crude oil atmospheric distillation unit. They developed dynamic model combined with a suitable control configuration to study the transient behavior from a one stationary level of operation to anther when the operating conditions or products specifications are changed. Lekan et al. (2012) [7], proposed a model to optimize the atmospheric distillation column of a crude oil. The proposed optimization model is applicable for the change in feed stock, market situations and products prices.
Akba and Umuze (2013) [9], make steady state MESH model to simulate crude oil distillation column. Their model is capable of predicting the concentrations and temperature of any component on the column trays. The model results of the concentrations and temperatures for five components compared with real data given maximum deviations of 8.33% and 6.62% respectively.  [10], simulate a crude oil distillation unit using nonlinear steady-state model embedded in the Aspen Hysys V 7.3. The model optimized using sequential quadratic programming (SQP) algorithm. Profit maximization for the weekly scheduling strategy was carried out to process different crudes blended in specific proportions. By optimizing the scheduling decision at the operational level they obtained 0.89 million American dollars weekly average profit. Yiqing et al. (2013) [11], applied ASPEN Plus to simulate and optimize the crude oil distillation unit. Particle swarm optimization (PSO) algorithm was used to obtain the optimum operating conditions of the existing distillation column with considering energy recovery and product profit.
The optimized operating conditions include product yields, stripping steam flow rate and pumparound duties.
Ahmed and Ala'a (2014) [12], used Aspen Hysys V. 7.1 to simulate and optimize the naphtha stabilizer in Al-Basrah Refinery. Their optimization results proved that, it's possible to increase the C5+ recovery in reformate from 97 % in actual unit to 99.6 % by selection the optimum design variables and operating conditions. Ali et al. (2014) [14], applied hybrid optimization algorithm which combines the Hysys simulation and genetic algorithm (GA) optimization method written by Matlab to optimize the design variables of the atmospheric crude oil distillation column. Dhallia et al. (2014) [13], used Aspen Hysys to study the effect of trays number, feed tray location and reflux ratio on residue and naphtha yield. They observe that the naphtha yield decrease with

Case Study
The crude oil distillation unit described in this study was originally designed for fractionation of 70000 BPD of Kirkuk light, Basrah light and Basrah medium crude oils. The obtained products are LPG, Light and heavy Naphtha, Kerosene, Light Gas Oil, Heavy Gas Oil and Atmospheric Residue.
The atmospheric distillation column consists of 46 tray with a total condenser, three side-strippers and two pump-arounds. The tray numbers arranged from top 1 to bottom 46.
The furnace is used to heat the crude oil to desired temperature (341-345 o C), then the crude oil is pumped to the atmospheric distillation column at the tray 42. The column operates with pressure 1-1.6 kg/cm 2 (g) and total condensation. The side-strippers which used to separate kerosene, gasoline and light gas oil contains 4, 5 and 8 trays respectively. Two pump-arounds are used to provide internal reflux at various sections of the column. Overflash 5.0 vol. % is used in order to reach sufficient fractionation efficiency. Naphtha leaves column to naphtha stabilizer to separate the gases from whole naphtha which then separated to light and heavy naphtha in the second distillation column of 35 tray [2].

Process simulation
Process modelling and simulation enables the designer to explore the process behaviour and select the optimum operating conditions to operate the process with maximum products at low cost.
Simulation can save a lot of money and time also it is cheaper and much faster than making a series of experiments [3]. Aspen Hysys is a powerful tool for chemical processes modelling, simulation and optimization. Aspen Hysys can be use for both steady state and dynamic simulation of complex crude oil distillation system [4].

Crude Oil Properties
Crude oils classified as light, Medium, Heavy and extra Heavy depending on API gravity ranges as shown in table (1).  (2) include the TBP assays of main four Iraqi crude oils used in this study. Table (3) represents typical crude oil cut points [3] .

Fig. (4) Comparison between experimental and predicted products weight fraction for
Basrah light crude oil.

Fig. (5) Comparison between experimental and predicted products weight fraction for
Basrah medium crude oil.

Products weight fraction in Heavy Crude oil
Blending of crude oil was used to produce a crude oil blend that has higher value than the raw heavy crude oil. Figure (6

Optimization of Operating Conditions
Aspen Hysys software was used to study the effect of crude oil specifications and column operating conditions on the products yield. The optimization was done by using Sequential Ten effective variables were selected (four steam flow rates, three columns feed temperatures and three products withdrawal temperatures) as operating conditions variables to maximize total profit.
Optimization was subjected to equality and inequality constraints to achieve products with specific properties by manipulating operating conditions. For kerosene, light gasoil and heavy gasoil, the range of products withdrawal temperatures was measured by the difference between the 95% ASTM D86 distillation temperature of a lighter product and the 5% ASTM D86 distillation temperature of an adjacent heavier product. Table (  According to Iraqi market, the feed, products and utility prices are summarized in table (5). The total profit was calculated for one year of 350 working days. will effect directly on heat consumption and on the total profit. It was shown from table (7), the manipulating the operating conditions increase the products profit within acceptable range without change the annual cost. The total yearly profit are 691.98, 713.28 and 724.4903 M$ for these three blends respectively. Light crude oils produce larger quantities of light valuable cuts. Lighter crude consume more steam, cooling and heating duties. Heavier crude oil consumes more electricity for pumping and cooling fan operation. Crude oil and utilities cost represent about 23% of the products outcome.

Conclusion
For three different crude oils (Kirkuk light, Basrah light and Basrah medium), Hysys simulation results agree very well with the results of actual unit. Depending on simulation model the user could predict the effect of design and operating variable on the petroleum products quality and quantity.
By making optimization for the atmospheric distillation unit operating conditions the products with a specified quality could be achieve without changing the design of the unit equipments.
The simulation results show that the blending 50% Basrah heavy crude oil with the same percent of Basrah light crude oil will increase the light Gas oil weight percent from 14.8 % in Basrah Heavy to 16.34 in the 50% blend. Lighter crude oils consume more steam, cooling and heating energy and less electricity. The total profit of lighter crude is greater than heavier crude oils because it produces light valuable cuts more than that produced by heavier crude oils. Lighter crude consume more steam, cooling and heating duties. Heavier crude oil consumes more electricity for pumping and cooling fan operation. Crude oil and utilities cost represent about 23% of the products outcome.