Professional Documents
Culture Documents
Application of Discrete Lumped Kinetic Modeling On Vacuum Gas Oil Hydrocracking PDF
Application of Discrete Lumped Kinetic Modeling On Vacuum Gas Oil Hydrocracking PDF
Simulation and Optimization 2013, Vol. 15, No. 2, pp 67-73 June 30, 2013
Abstract: The kinetic model of vacuum gas oil (VGO) hydrocracking based on discrete lumped approach was investigated,
and some improvement was put forward at the same time in this article. A parallel reaction scheme to describe the conver-
sion of VGO into products (gases, gasoline, and diesel) proposed by Orochko was used. The different experimental data
were analyzed statistically and then the product distribution and kinetic parameters were simulated by available data. Fur-
thermore, the kinetic parameters were correlated based on the feed property, reaction temperature, and catalyst activity. An
optimization code in Matlab 2011b was written to fine-tune these parameters. The model had a favorable ability to predict
the product distribution and there was a good agreement between the model predictions and experiment data. Hence, the ki-
netic parameters indeed had something to do with feed properties, reaction temperature and catalyst activity.
Key words: hydrocracking; kinetic modeling; vacuum gas oil (VGO); optimization code; parallel reaction scheme
· 67 ·
China Petroleum Processing and Petrochemical Technology 2013,15(2):67-73
steps of cation chemistry, respectively. For a relatively tion of oil refining processes.
large number of pseudo-components the analytical infor- To have a better understanding of the VGO hydrocrack-
mation and experimental data are required, which impose ing, in this work we conducted some experiments in a pi-
restrictions on their applications to hydrocracking of real lot plant equipped with two downflow fixed-bed reactors
feedstocks. and developed a discrete lumped kinetic model, which
The complexity of real feedstocks suggests that lumped will be presented in further papers.
model will continue to be used for the research on VGO
hydrocracking kinetics. However, detailed approaches 2 Experimental Data and Kinetic Model
need to be studied more precisely in order to obtain a bet-
ter understanding of hydrocracking kinetics and provide 2.1 Experimental setup and feedstock
an idea to optimize lumped model for prediction of hy- The experiments were conducted in a single-stage series
drocracking product properties. A good kinetic model is fixed-bed high-pressure pilot plant, which is shown sche-
a useful tool for reactor design, simulation, and optimiza- matically in Figure 1. Hydrocracking was conducted in a
downflow mode of operation. The hydrotreating catalyst VGO, Saudi VGO and mixed oil (Iranian VGO: CGO=8:2
was loaded in a hydrotreating reactor and the hydrocrack- sampled from Zhenhai refinery), were chosen as the feed.
ing catalyst (A) was loaded in a hydrocracking reactor. Main properties of these feedstocks are listed in Table 1.
The reactor temperature was controlled at the desired These feedstocks had high density, EBP, BMCI and sulfur
level by using a three-zone electric furnace providing an content so that severe hydrogenation operating conditions
isothermal temperature condition along the reaction sec- were needed in order to achieve the expected conversion
tion. The temperature inside the reactor was measured by degree and index requirements of product properties. Hy-
a movable axial thermocouple located inside the reactor. drogen gas was provided from an electrolysis hydrogen
Three typical heavy VGO samples, including Iranian source, with the hydrogen purity exceeding 99.9%.
· 68 ·
Han Longnian, et al. Application of Discrete Lumped Kinetic Modeling on Vacuum Gas Oil Hydrocracking
Table 1 Main properties of feedstock late yield increased when the hydrocracking temperature
Items Iranian VGO Saudi VGO Mixed oil increased from 381 ℃ to 390 ℃ and these results were
Density (20 ℃), g/cm3 0.9130 0.9208 0.9163 consistent with the catalytic activity of the hydrocracking
Distillation, ℃ catalyst A.
IBP/10% 336/379 325/382 327/389 Table 2 Product yield distribution of Iran VGO hydrocracking
30%/50% 409/433 411/435 414/434 Reaction temperature, ℃ 381 385 390
70%/90% 463/506 462/505 462/513 Conversion (>370 ℃), % 65.30 75.80 83.52
· 69 ·
China Petroleum Processing and Petrochemical Technology 2013,15(2):67-73
gasoline lump (light and heavy naphtha), and diesel lump yexp -ymod
ARD = ×100% (5)
(kerosene and diesel) were considered as three of four yexp
lumps. Moreover, the hydrocracker products and yield are 3
shown in Table 2. The experimental data of hydrocrack- OBJ = ∑ ( yexp -ymod ) 2 ×100% (6)
i =1
ing process were used to model the kinetic parameters of in which yexp and ymod are experimental and model yield
hydrocracking reactions. of product, respectively. The ARD and OBJ of model pre-
The hydrocracking process took place according to a dictions are shown in Table 3. It can be seen from Table
retarded first-order reaction, which was the equation de- 3 that the calculation error was small enough to meet the
veloped by Orochko[1]. Based on this method, the yields requirements of the model.
of diesel lump and gasoline lump can be expressed by the
Table 3 Model macrokinetic parameters and standard
total conversion y and macro-kinetic parameters.
'
deviation
(1 − y ) k -(1-y )
Diesel lump yield: Z = (1) Reaction temperature, ℃ 381 385 390
1-k '
Conversion, % 65.30 75.80 83.52
Gasoline lump yield:
'' ' '' Gexp, % 1.81 2.12 3.58
(1 − y ) k -(1-y ) k (1 − y )-(1 − y ) k
X = k' × ' ' ''
+ k' × (2)
(1-k ) × (k -k ) (1-k ' ) × (1-k '' ) Gmod, % 1.809966 2.120228 3.579525
Gases lump yield: G=Y-(Z+X) (3) ARD (G) 1.9029e -5
1.0763e -4
1.3273e-4
in which Z, X, and G are experimental yield of diesel Xexp, % 11.04 16.46 20.83
lump, gasoline lump and gases lump, respectively, and y
Xmod, % 11.03991 16.460169 20.829039
is the conversion rate of VGO. -6 -5
ARD (X) 7.9051e 1.0239e 4.6152e-5
Among them, the macrokinetic parameters (k′ and k″)
Zexp, % 52.45 57.22 59.11
depend on feed properties, process temperature and
Zmod, % 52.450053 57.219603 59.110486
catalyst characteristics. All kinetic model parameters
-6 -6
are determined by experimental data and simulated by ARD (Z) 1.0073e 6.9329e 8.2249e-6
an optimization procedure written in the Matlab 2011b. OBJ 1.1594e-12 2.3784e-11 1.3863e-10
Furthermore, the objective function which determines the k′ 0.3624 0.3324 0.3081
optimized model parameters is defined as follows. k″ 0.3969 0.2268 0.2278
OBJ=min[(Zmod-Zexp)2+(Gmod-Gexp)2+(Xmod-Xexp)2] (4)
Since k′ and k″ are kinetic factors with similar meaning
3 Results and Discussion to the rate constants, the relation between macrokinetic
parameters and reaction temperature based on the Arrhe-
3.1 Calculation of kinetic modeling parameters nius equation should be studied. The correlation between
The operating conditions and experimental data of a pilot- -ln(k′, k″) and 1/T is presented in Figure 3. It can be obvi-
plant scale hydrocracker were used to calculate and vali- ously seen that the macrokinetic parameter k′ could sat-
date the model parameters. The results of fine-tuning are isfy the Arrhenius equation along with high related coef-
shown in Table 3. By using these parameters, the product ficient (R2=0.9905), but there was not a close relationship
distribution can be calculated. A comparison between between macrokinetic parameter k″ and reaction tempera-
the model results for predicting the product yields is also ture because of its relatively low coefficient (R2=0.6899).
shown in Table 3. It can be seen from Table 3 that there It might occur because the model parameter (k″) was
was an extraordinarily good agreement between model not only a function of reaction temperature, but was also
prediction results and experimental data. The absolute related with other process conditions. This experiment
relative deviation (ARD) and mean square error (OBJ) of studied only the process conditions of different reaction
the model for predicting the product yields are defined as temperatures so that the complete correlation could not be
follows: fully reflected. Perhaps, a further research work needs to
· 70 ·
Han Longnian, et al. Application of Discrete Lumped Kinetic Modeling on Vacuum Gas Oil Hydrocracking
with similar meaning to the rate constants. Gas yield, % 2.12 3.26 4.27
So a correlation function was put forward to calculate Gasoline yield, % 16.46 21.47 18.19
the macrokinetic parameters. Among them, the charac-
Middle distillate yield, % 57.22 58.60 59.96
teristic factor (Kf) and the volume-mean boiling point
(Tv) functioned as factors which could mainly represent k′ 0.3324 0.3164 0.2956
the feedstock property. The weight percent conversion of k″ 0.2268 0.2035 0.3204
VGO (Xf), reaction temperature (T), and zeolite content of OBJ 2.3784e -11
5.5592e -12
7.3715e-11
catalyst are also included in this empirical correlation. Kf
and TV are calculated by the following formulas and the And then the model parameters describing the relation-
calculated results are listed in Table 1. ship between macrokinetic parameters and feedstock
d15.6=0.99417d20 +0.009181 (7) properties, catalyst activity, and reaction temperature were
Tv=(T10+T30+T50+T70+T90)/5 (8) simulated by an optimization procedure written in the
Tm=TV+5.7715-6.7083×(T90-T10)/80-19.32×100/TV + Matlab 2011b. The objective function which determines
14.913×(T90-T10)/80×100/TV+3.6694×[(T90-T10)/80]× the optimized model parameters is defined as mentioned
(100/TV)2-6.2278×[(T90-T10)/80]2×100/TV (9) previously. Judging from the optimization results listed in
1/3
Kf=1.216×(Tm +273.15 ) /d15.6 (10) Table 5, it has good prediction accuracy which is appro-
· 71 ·
China Petroleum Processing and Petrochemical Technology 2013,15(2):67-73
OBJ 5.3897e-14
Model parameters B1 B2 B3
· 72 ·
Han Longnian, et al. Application of Discrete Lumped Kinetic Modeling on Vacuum Gas Oil Hydrocracking
ternational Journal of Chemical Reactor Engineering, 2010, microkinetic model for long-chain paraffin hydrocracking
8(A1): 1-24 and hydroisomerization on an amorphous Pt/SiO2·Al2O3
[5] Sadighi S, Ahmad A, Rashidzadeh M. 4-Lump kinetic
catalyst[J]. Ind Eng Chem Res, 2009, 48(7): 3284-3292
model for vacuum gas oil hydrocracker involving hydrogen [14] Guillaume D, Valery E, Verstraete J J, et al. Single
consumption[J]. Korean J Chem Eng, 2010, 27(4): 1099- event kinetic modeling without explicit generation of
1108 large networks: Application to hydrocracking of long
[6] Moghadassi A R, Amini N, Fadavi O, et al. The application
paraffins[J]. Oil & Gas Science and Technology, 2011,
of the discrete lumped kinetic approach for the modeling of 66(3): 399-422
a vacuum gas oil hydrocracker unit[J]. Petroleum Science [15] Orochko D I, Perezhigina I Y, Rogov S P, et al. Applied
and Technology, 2011, 29(23): 2416-2424 over-all kinetics of hydrocracking of heavy petroleum
[7] Kumar A, Sinha S. Steady state modeling and simulation of
distillates[J]. Khimiya I Tekhnologiya Toplivi Masel,
hydrocracking reactor[J]. Petroleum & Coal, 2012, 54(1): 1970, 8(6): 561-565 (in Russian)
· 73 ·