REACTOR MODELLING METHOD FOR FLUIDIZED-BED CATALYTIC CRACKING

UNIT

Field of the Invention

The present invention relates to a method used for modeling reactors of a fluidized- bed catalytic cracking unit.

Background Art

In petroleum refineries, conversion of heavy vacuum gas oil (HVGO) with a high boiling point into hydrocarbons such as fuel gas, gasoline, LPG, coke, light cycle oil (LCO) with a lower boiling point is performed in fluidized-bed catalytic cracking units ((FCCU). In the said units, separate catalysts are used for conversion processes. One of the products obtained in the fluidized-bed catalytic cracking units is coke, which is generated on the catalyst. In order to separate the catalyst from the coke and provide energy for conversion processes, coke is burned in the fluidized-bed catalytic cracking units (in the regenerator section of the unit).

Since the fluidized-bed catalytic cracking units are multi-input and multi-output units, it is not possible to identically model the reactors of these units. Therefore, in the prior art, such applications are employed that model the reactors of a fluidized-bed catalytic cracking unit based on prediction. In the prior art, the said prediction-based applications are called as lumping methods. In lumping methods, a modeling is made based on a conversion of a lump into another one. However, given that at least 15 reaction constants are required even in an application including only 6 products, the lumping methods are complicated.

The most detailed method among the lumping methods is the Structure Oriented Lumping method disclosed by R.J. Quann, S.B. Jaffe, in Structure-Oriented Lumping - Describing the Chemistry of Complex Hydrocarbon Mixtures (Vol 31 , Pg2491 , 1992), Ind Eng Chem Res, 32 (1993) 1800-1800. In the Structure- Oriented Lumping method, the lumps are defined by a vector, wherein each element represents a given chemical structure (nitrogen group, cycle structure etc.). In the Structure- Oriented Lumping method, if the number of the defined structures is high, then the vector is increased and the method provides more detailed results. However, an increase in the vector also causes a complicated calculation. Therefore, any applications employing the Structure Oriented Lumping method have a high processor requirement.

In order to simplify the lumping methods, applications are used in the prior art, wherein the parameter number is reduced. Although a reduction in the parameter number reduces the complexity of the application and thus decreases the number of operations, these applications do not yield adequate correct results.

The state-of-art document US2012193269A1 discloses a method for data-based modeling of a FCC unit (reactor and regenerator) using the method of artificial neural networks. This modeling work is combined with optimization and control work on maximizing light products. However, the method disclosed in this document fails to provide a complete description of reactions in the fluidized-bed catalytic cracking unit, and consequently does not provide an adequate quick and precise result. Brief Description of the Invention

With the present invention, there is provided a method for modeling reactors of a fluidized- bed catalytic cracking unit in petroleum refineries, wherein the heavy vacuum gas oil (HVGO) with a high boiling point is converted into hydrocarbons such as fuel gas, gasoline, LPG, coke, light cycle oil (LCO) with a lower boiling point. The said method comprises the steps of receiving as input the characterization of the charge inputted into the reactors of the fluidized-bed catalytic cracking unit, the process variables of the said unit, the geometry of the reactor used in the said unit and the model parameters; determining a common temperature value of the charge and the catalyst at the reactor inlet assuming that the charge immediately evaporates after the charge and the catalyst contact each other; determining the rate of the hydrocarbons in gaseous form resulting from the evaporation of the charge; creating a mass balance, an energy balance and a pressure balance by using the information received as input, the common temperature value and the rate of the hydrocarbons in gaseous phase; and calculating the temperature profile, the rate profile, the pressure profile throughout the reactor and the total amount of coke resulting from the product distribution at the end of the reactor by co-solving of the created mass balance, energy balance and pressure balance.

In the modeling method of the present invention, the created mass balance, energy balance and pressure balance is solved with the first-order differential equation solving algorithms. In this way, the modeling method is operated in a quick and reliable manner. In addition, since the data received as input can be changed in the modeling method, when necessary; the method can be updated in different charges or in different environmental conditions so that real-time results may be obtained.

Object of the Invention

An object of the present invention is to provide a method for modeling reactors of a fluidized- bed catalytic cracking unit.

Another object of the present invention is to provide a method with reduced complexity, as compared to the prior art.

Another object of the present invention is to provide a method describing the reactions performed in the fluidized-bed catalytic cracking unit in the most accurate and simplest manner.

Still another object of the present invention is to provide a method yielding rapid and reliable results.

Description of the Invention

Fluidized-bed catalytic cracking units (FCCU) are unites where heavy vacuum gas oil (HVGO) with a high boiling point is converted into hydrocarbons such as fuel gas, gasoline, LPG, coke, light cycle oil (LCO) with a lower boiling point. The said conversion process is not performed such that only a product is converted into another. A product is initially converted into a second product and then into a third (or more) product. The complexity of the conversion process causes a difficulty in modeling of reactors in the fluidized-bed catalytic cracking unit. Therefore, with present invention, there is provided a method used for modeling reactors of a fluidized- bed catalytic cracking unit.

In the method of the present invention, at first the characterization of the charge, the process variables, the geometry of the reactor and the model parameters are received as input. Then, the common temperature of the charge and the catalyst at the reactor inlet is determined. Here, it is assumed that the charge evaporates immediately, after the charge and catalyst contact each other. After the common temperature is determined, the running rate of the hydrocarbons in the gaseous phase is determined. The mass balance, energy balance and pressure balance are created using the inputted information, the common temperature and the rate of the hydrocarbons in gaseous phase. The created mass balance, energy balance and pressure balance are co-solved and the temperature profile, the rate profile, the pressure profile throughout the reactor and total amount of coke resulting from the product distribution at the end of the reactor are calculated.

The charge of the fluidized-bed catalytic cracking units is composed of multiple products. Since it is not possible to analyze each product, in a preferred embodiment of the invention, the petroleum fractions contained in the charge are divided into groups with very low boiling ranges, called as pseudocomponent. By using the normal boiling temperature and specific gravity of the said groups, such features as molecular weight, critical temperature, critical pressure, acentric factor are determined for each group. These features are used as the charge characterization data. In the method of the present invention, the charge is preferably divided into 74 pseudocomponents. The characteristics of the some pseudocomponents in the embodiment of the present invention are given in Table 1 .

Table 1

Within the scope of invention, the process variables include such information as the temperature and amount of the charge, the pressure of the reactor, the circulation rate of the catalyst (the amount of the catalyst circulated/flown in the unit per unit time), the temperature of the catalyst received from the regenerator and the amount of coke on the catalyst received from the regenerator. The said information is updated periodically since it is subject to change in the course of operation of the fluidized-bed catalytic cracking unit.

The said reactor geometry includes such information as the length and diameter of the reactor. Since the geometry of the reactor determines the operation performance of the fluidized-bed catalytic cracking unit, it is important for the modeling to input the said information.

The said model parameters include the reaction rate constant, the reaction heat and the product distribution curve (which components and at which proportion a component produces, when cracked). The said information is obtained by prediction-based methods using the real field data. The said model parameters must be further determined based on the charge fed to the fluidized-bed catalytic cracking unit or the catalyst contained in the unit.

In order to determine the said reaction rate constant, the following equation (1 ) is used. (a _l -a ₂ xTB _i )

k^ ia x TB^ x e ^RTmi* (1 ) wherein k _t denotes the reaction constant of the i pseudocomponent; a , b , a _x , a ₂ denote the model parameters; TB _i denotes the boiling temperature (K) of the i ^th pseudocomponent; R denotes the gas constant (8,314 (J/mol.K); T _mix denotes the common temperature (K).

For the product distribution curve, the following equations (2) and (3) are used.

wherein P denotes the normalization function; λ and // denote the model parameters;

TB _t denotes the boiling temperature (K) of the i ^th pseudocomponent; TB _j denotes the boiling temperature (K) of the j ^th pseudocomponent; p{i, j) denotes the distribution function.

In this step, the following equation (4) is used in order to determine the common temperature of the charge and catalyst at the reactor inlet assuming that after the charge and catalyst are mixed, the charge immediately evaporates; J f <? -c p.co _t idT = Y / ί J f ί. -c p, l ,iqui .d,, iiff + Y / j να >,ϊ + / j [ i. - C p,gas,i i r

(4)

wherein M _c denotes the flow rate (kg/h) of the catalyst; C _ca; denotes the heat capacity (kj/kg.K) of the catalyst; T _mix denotes the common temperature (K); N denotes the number of components; T _feed denotes the temperature (K) of the charge; T _B ,i denotes the boiling temperature (K) of the component i; M _i denotes the flow rate (kg/h) of the component i; Cp iquid denotes the heat capacity (kj/kg.K) of the component i in liquid phase; AH _vap denotes the evaporation heat (kJ/kg) of the component i; T _cat denotes the inlet temperature (K) of the catalyst into the reactor; c _{gas )} denotes the heat capacity (kj/kg.K) of the component i in gaseous phase.

Here, the heat capacities of the hydrocarbons are calculated using equations (5-8) for the liquid phase and equations (9-15) for the gaseous phase. For the liquid hydrocarbons;

A _t = -1,17126 + (0, 023722 + 0, 024907 x SG) x K _w + ^{( 14982 "} °' ^{046535 X K» }} ₍₅ )

SG = le (1 + 0, 82463 x K x (1, 12172 - ' ^ ) ₍₆₎

A ₃ = - \e (1 + 0, 82463 x K _w ) x (2, 9027 - ° ^{, 7} ° ⁹⁵⁸ ) ₍₇₎

SG

Cp _{li< jM} = (4 + A ₂ x T _mix + 4 x 7\ ² ) x 4, i 868

(8) wherein A _{l j} A _{i an(} j _l A ₃ denote the constants used in the heat capacity calculations; K„ denotes the Watson K factor; SG denotes the specific weight of the liquid mixture; T _mi denotes the common temperature (K) of the mixture; Cp _quid denotes the heat capacity (KJ/Kg.K).

For the hydrocarbons in gaseous phase;

B _l = -0, 35644 + 0, 02972 xK _w +B ₄ x (0, 29502 - °' ²⁴⁸⁴⁶ ) ₍₉₎

SG

50694

B ₂ = - \e x (2, 9247 - (1, 5524 - 0, 05543 xK xK _w +B ₄ x (6, 0283 - ) (1 o)

SG

B ₃ = -le ⁷ x (1,6946 + 0,0844 xB ₄ ) ₍₁₁₎

P _cor =— χ((0,66-0,92χίο)χΓ/ ² +(0,831 + 3χίο)χΓ/ ³ +(0,145 + 1,16χίο)χΓ; ⁴ + 0,526χίοχΓ/ ⁹ ) (14)

C _Pgas =(B _{1 +} B ₂ xT _mix +B,xT -^ x _cw ,)x4,1868

(15) wherein B _{x 5} B ₂₅ B _{i 5} B ₄ denote the constants used in heat capacity calculations; K _w denotes the Watson K factor; SG denotes the specific weight of the liquid mixture; T _mix denotes the common temperature (K) of the mixture; Cp _gas denotes the heat capacity (KJ/Kg.K); T _c denotes the critical temperature (K); T _T denotes the reduced temperature; P _cor denotes the corrected pressure; P denotes the reactor pressure (bar); P _c denotes the critical pressure (bar); MW denotes the molecular weight; ω denotes the acentric factor.

Within the scope of invention, in the step of determining the rate of the hydrocarbons in gaseous phase, it is assumed that the gases act as ideal gas. Furthermore, in this step, it is also assumed that the rate of the hydrocarbon and catalyst is equal. In order to determine the rate of the hydrocarbons in gaseous phase, the following equation (16) is used.

M R -T _mix

S MW - P _inla - A < ¹⁶ > wherein v _g denotes the rate (m/h) of the hydrocarbons in gaseous phase; M denotes the flow rate (kg/h) of the hydrocarbons; R denotes the gas constant (8,314 (J/mol.K); T _mix denotes the common temperature (K) of the mixture; MW denotes the molecular weight; P _inlet denotes the inlet pressure (bar); A denotes the cross-sectional area (m ² ) of the reactor.

In creating a mass balance in the inventive modeling method, both the mass consumed due to the cracking for each product and the mass of the product obtained from a heavier product are taken into account. Here, prediction-based methods are used to determine at which amount each product may be generated. The mass balance (17) used in this step is as follows;

cat r P ccaatt - ® - A

(17) wherein M _t denotes the flow rate (kg/h) of the component i; z denotes the length (m) of the reactor; k _i denotes the reaction rate constant (m ³ /kgcat.h) of the component i; A denotes the cross-sectional area (m ² ) of the reactor; v _g denotes the rate (m/h) of the hydrocarbons in gaseous phase; s _cat denotes the proportion of the catalyst to the reactor volume; p _cat denotes the catalyst intensity (kg/m ³ ); Φ denotes the deactivation function of the catalyst; N denotes the number of the components; p(i,n) denotes the possibility of generating the component i from the component n; φ denotes the ratio of hydrocarbon resulting from the cracking (the total is 1 with a coke generating inclination); k _n denotes the reaction rate constant (m ³ /kgcat.h) of the component n; M _n denotes the flow rate (kg/h) of the component n. For the said deactivation function of the catalyst, the following equation (18) is used; m _ -500(M _coke /M _cat )

vt>— e (is) wherein M _coke denotes the amount of coke generated (kg); M _cat denotes the amount of catalyst used (kg); Φ denotes the deactivation function.

In the inventive modeling method, in creating an energy balance, the energy inputted to and outputted from the fluidized-bed catalytic cracking unit is equalized. The energy balance (19) used in this step is as follows;

wherein M denotes the flow rate (kg/h) of the hydrocarbons; C _Piavg denotes the average heat capacity (kj/kg.K) (which is calculated using Cp,,^ calculated in equation 8 and Cp _gas calculated in equation 15); T denotes the temperature (K); M _cat denotes the flow rate (kg/h) of the catalyst; C _ca; denotes the heat capacity (kj/kg.K) of the catalyst; AH, denotes the enthalpy (kJ/kg) resulting from the reaction of the component i; ^ denotes the reaction rate constant (m ³ /kgcat.h) of the component i; M _i denotes the flow rate (kg/h) of the component i; v _g denotes the rate (m/h) of the hydrocarbons in gaseous phase; s _cat denotes the proportion of the catalyst to the volume of the reactor; p _cat denotes the catalyst intensity (kg/m ³ ); Φ denotes the deactivation function of the catalyst (which is calculated in equation 18). The enthalpy of the pseudocomponents is calculated using equations (20) and (21 ); = H _c , _coke ' (Χ-Φ) +∑[ρϋ, ί) · φ · ii _{c j} ] - H _{c i} (21 )

7=1 wherein AH _i denotes the enthalpy (kJ/kg) resulting from the reaction of the i pseudocomponent; H _{c coke} denotes the combustion enthalpy (kJ/kg) of the coke generated; p(j, i) denotes the distribution function; H _{c i} denotes the combustion enthalpy of the i ^th component whereas H _{c j} denotes the combustion enthalpy of the j ^th component (kJ/kg); c , d and Φ denote the model parameters; TB _t denotes the boiling point (K) the i ^th component.

Finally, in creating a pressure balance within the scope of the invention, the pressure profile of the hydrocarbons in gaseous phase is created with the below simplified pressure equation (22). d(pv) d (pvv) dP

dt dz dz ( _c cVc ² + p _g ) (22)

wherein P denotes the pressure of the reactor (bar); z denotes the length of the reactor (m); s _c denotes the volume ratio of the catalyst; p _c denotes the catalyst intensity (kg/m ³ ); g denotes gravity (gravitational acceleration); v _c denotes the rate of the catalyst (m/h); v _g denotes the rate of the hydrocarbons in gaseous phase (m/h). In the modeling method according to the present invention, a mass balance, an energy balance and a pressure balance are created and the balances created are solved with the first-order differential equation solving algorithms. In this way, the modeling method is operated in a quick and reliable manner. In addition, since the data received as input can be changed in the modeling method, when necessary; the method can be modified in different charges or in different environmental conditions so that real-time results may be obtained.