ISSN 09655441, Petroleum Chemistry, 2015, Vol. 55, No. 6, pp. 481–486. © Pleiades Publishing, Ltd., 2015. Original Russian Text © I.M. Gerzeliev, K.I. Dement’ev, S.N. Khadzhiev, 2015, published in Neftekhimiya, 2015, Vol. 55, No. 4, pp. 331–336.
Effect of Catalyst and Feedstock Modification with Ultrafine Molybdenum Disulfide Particles on the Performance Characteristics of Catalytic Cracking I. M. Gerzeliev, K. I. Dement’ev, and S. N. Khadzhiev Topchiev Institute of Petrochemical Synthesis, Russian Academy of Sciences, Moscow, Russia email:
[email protected] Received January 15, 2015
Abstract—The influence of in situ generated ultrafine molybdenum disulfide particles on the dispersion properties of feedstock and the acidity of a microspherical zeolite catalyst during the catalytic cracking of a vacuum distillate has been studied. The character of change in both the yield of catalytic cracking products and the hydrocarbon group composition of the gasoline fraction indicates that molybdenum disulfide exhibits hydrogenating activity under the catalytic cracking conditions. It has been shown that modification with molybdenum disulfide can result in a reduction in the yield of light gas oil, the olefin content in the gasoline fractions, and the yield of hydrogen and, in general, opens the possibility of controlling the qualitative and quantitative compositions of catalytic cracking products. Keywords: catalytic cracking, ultrafine molybdenum disulphide particles, catalyst modification DOI: 10.1134/S0965544115060055
Along with catalyst modification [1], variation of feedstock quality and relevant process parameters [2], change in the structural and mechanical properties of a petroleum disperse system [3], and introduction of passivating additives to reduce metal poisoning [4], the use of multifunctional additives affecting the proper ties of both feedstock and a cracking catalyst is also of interest for enhancement of the catalytic cracking pro cess. In this case, it is desirable that a multifunctional modifying additive should have a positive effect on the petroleum disperse system, exhibit a high hydrogenat ing ability under mild conditions to lower the amount of unsaturated hydrocarbons in the gasoline fractions, decrease the yield of light gas oil, and exert the catalyst passivating effect consisting in reducing the yield of hydrogen. Analysis of the properties of modifying additives for various catalytic processes, in particular, in the hydro conversion shows that these requirements can be met by molybdenum disulfide, which is widely used as a hydrogenating component in catalyst systems [5] and ultrafine hydroconversion catalysts for coal [6, 7] or heavy petroleum residues [8, 9]. From the viewpoint of technology, the option when the modifier is intro duced into feedstock from a precursor, for example, water or oilsoluble salts of molybdenum, is of great interest. Thus, in the hydrogenation of coal and petro leum residues, a microemulsion of aqueous ammo nium paramolybdate solution is used, which reacts with hydrogen sulfide destroying the emulsion to in
situ form molybdenum disulfide with a particle size of 0.02–1 µm [10]. The use of catalysts of this type for the hydroconversion of atmospheric and vacuum residues make it possible to significantly modify these petro leum disperse systems with a high degree of conversion to lighter fractions. The conversion conditions (450– 490°С, 60–70 atm) are milder as compared to con ventional hydrocracking of heavy residues (100– 300 atm) [8, 9]. The successful application of ultrafine molybde num catalyst systems in heavy feedstock hydrogena tion processes under mild conditions suggests that they may exhibit activity in the hydrogenation of hydrocarbons during the catalytic cracking of vacuum distillate. EXPERIMENTAL Molybdenum compounds were introduced into the feedstock in the form of precursor, which was a mix ture of ammonium sulfide and an aqueous solution of ammonium paramolybdate (NH4)6Mo7O24, so as to have a sulfur : molybdenum molar ratio 3 : 1 in the solution. Under these conditions, molybdenum exists as the ammonium salt of thiomolybdic acid in solu tion. Mixtures of molybdenumcontaining additive with hydrotreated vacuum distillate were prepared in a laboratory mixer at a rotating speed of 5000 rpm for a mixing time of 5 min. The size of resulting molybde num disulfide particles was measured by dynamic light
481
482
GERZELIEV et al.
heavy gas oil (320–FBP). The group composition of the gasoline fraction was determined according to PONA analysis by chromatographing on the Kristally uks4000M instrument with FID using a Supelco Pet rocol DH150 column (length 150 m, diameter 0.53 mm, film thickness 1 µm).
100 2 1
Mass, %
80 60
3
40 20 0 0
100
200
300 400 500 600 Temperature, °C
700 800
Fig. 1. Mass loss curves for the vacuum distillate and emul sion phases: (1) vacuum distillate, (2) hydrocarbon layer, and (3) aqueous layer.
scattering using a Beckman Coulter N5 submicron particle size analyzer. The catalytic cracking of a vacuum distillate in the presence of the precursor was studied in a laboratory flow reactor according to the procedure described in [11]. A largescale study involving admixing of the molybdenum disulfide precursor to the feedstock was carried out in a pilot vacuum distillate catalytic crack ing unit including a riser reactor [12] under the follow ing conditions: temperature, 500°С; space time, 2 to 3 s; catalyst : feedstock recycle ratio, 8 kg/kg; and feed mass flow rate, 500 g/h. Hydrotreated vacuum distil late obtained from a blend of West Siberian oils and a REDUXION DMS PRO microspherical zeolite cata lyst in the equilibrium form (BASF) were used in the experiments. The vacuum distillate and the catalyst were provided by the Moscow Refinery. The gaseous products of the reaction were deter mined with a Kristallyuks4000m chromatograph on two columns. One column, packed with molecular sieves CaX (3 m in length, 5 mm in diameter), was used to determine methane and nonhydrocarbon components of the gas (H2, O2, N2, CO). The other, a Varian HPPLOT/Q capillary column (30 m length, 0.32 mm diameter, and liquidphase film thickness 20 µm), was used to determine the hydrocarbon com position of the reaction gas. A thermal conductivity detector was used in both cases. The carrier gases were argon and helium for the packed and capillary col umns, respectively. Liquid products were determined on the Kristally uks4000M instrument with a flameionization detec tor according to the ASTM2887 procedure. An Agi lent DB2887 capillary column (length 10 m, diame ter 0.53 mm, film thickness 3 mm) was used for this purpose. The liquid products were fractionated using an ARNS1E apparatus to obtain three fractions: gas oline (IBP–200°C); light gas oil (2200–320°С); and
RESULTS AND DISCUSSION For the samples prepared, the introduction of the precursor into the vacuum distillate led to a certain reduction in the average particle size of the dispersed phase from 330 nm to 150–200 nm as compared with the particle size of the dispersed phase in the starting distillate. This change suggests that molybdenum dis ulfide reacts with the complex structural unit of the petroleum disperse system and precursorcontaining microemulsion droplets are uniformly distributed in the feedstock. Accordingly, the MoS2 molybdenum sulfide particles formed during the decomposition of the precursor under catalytic cracking conditions are also uniformly distributed in the bulk. To confirm the fact of molybdenum disulfide inter action with the complex structural unit, the upper (hydrocarbon) and the lower (aqueous) layers after breakdown of the emulsion were investigated by ther mogravimetry (TGA, Fig. 1), as well as the initial vac uum distillate. The presence of two steps in the mass loss curve of the aqueous layer shows the joint separa tion of the aqueous component of the precursor solu tion and a portion of the lighter components of the original vacuum distillate, which, most likely, have been produced via the degradation of the outer layer of the complex structural unit. The same conclusion fol lows from the fact that the hydrocarbon layer became heavier than the initial distillate according to the true boiling point (TBP) data obtained by simulating the distillation of the fractions. The influence of the precursor content in the feed stock on the characteristics of catalytic cracking was studied in a laboratory unit at a temperature of 500°С and a weight hourly space velocity (WHSV) of h–1. The precursor content was varied in the range of 0.015– 0.05% (hereinafter, wt % unless otherwise stated) on a molybdenum basis. By increasing the precursor con tent to 0.03%, the gas yield was increased from 22.8 to 24.9% (Fig. 2). A further increase in the precursor content did not result in a noticeable change in the gas yield. The yield of light gas oil decreased from 16.3 to 15.5% with an increase in the precursor concentration to 0.05%. The gasoline yield under these conditions passed through a minimum (40.6%) at a precursor concen tration of 0.03% (Fig. 3). The further increase in the precursor concentration to 0.05% increased the yield of gasoline to its original value (42.2%). The conver sion of feedstock remained almost unchanged with the precursor concentration varied from 0 to 0.05%, mak ing 91.0–92.5%. PETROLEUM CHEMISTRY
Vol. 55
No. 6
2015
28
46
26
45
24
44
22
43 Yield, %
Yield, %
EFFECT OF CATALYST AND FEEDSTOCK MODIFICATION
20 18 16
483
42 41 40
14
39
12 0
0.01 0.02 0.03 0.04 Precursor concentration, % Gas
38
0.05
0
0.01 0.02 0.03 0.04 Precursor concentration, %
0.05
Light gas oil
Fig. 2. Dependence of the gas and light gas oil yields on the precursor content.
Fig. 3. Dependence of the gasoline yield on the precursor content.
The yield and composition of the gas also signifi cantly depend on the precursor content in the feed stock. An increase in the precursor concentration to 0.05% increases the yields of the propane–propylene fraction (PPF) from 7.2 to 7.9% and the butane–buty lene fraction (BBF) from 12 to 13.0%. The both frac tions were enriched in saturated hydrocarbons: the propylene/propane ratio decreased from 1.72 to 1.32 and the butylenes/butanes ratio decreased from 0.60 to 0.43 (Fig. 4). These results indicate that molybdenum disulfide exhibits hydrogenating properties under catalytic cracking conditions, a fact that is further confirmed by the reduction in hydrogen yield from 0.53 to 0.38% by introducing the precursor into the cracking feed feed stock (Fig. 4). Thus, the precursor concentration of 0.05% was used in further experiments, since the highest incre ment in the gas yield was observed for this value almost without a decrease in yield of the gasoline fraction. The effect of temperature on the cracking process with the precursorcontaining feedstock was investi gated in the laboratory unit in the temperature range of 480–520°С. The precursor content was 0.05%, and a WHSV was 2 h–1. It was found that the cracking parameters of the pure vacuum distillate and the feed stock doped with the precursor symbatically vary with temperature. As the temperature increased from 480 to 520°С, the gasoline yield decreased in both cases (Fig. 5). At temperatures of 480 and 500°С, the yield of gasoline by cracking with or without the precursor was almost the same, whereas that at 520°С in the presence of the additive was significantly lower (38.7%) than for the pure feedstock (40.6%). Note that the feedstock con version was the same in both cases and practically did not depend on temperature in the investigated range,
making 91.4–92.7%. Probably, the cracking reactions of acyclic hydrocarbons in the presence of precursor are facilitated at an elevated temperature. The yield of light gas oil slightly increased with temperature, being lower in the cracking of the feedstock with the additive than in the case of the pure feedstock over the entire the temperature range (Fig. 6). An increase in temper ature from 480 to 520°С in the cracking of the pure distillate increased the yield of light gas oil from 16.1 to 16.8%. The cracking of the feedstock with the additive resulted in an increase in the gas oil yield from 15.3 to 15.8% with the cracking temperature under the same conditions. The yields of gaseous cracking products increased with increasing temperature in both cases. For exam ple, the gas yield during the cracking of the pure feed
PETROLEUM CHEMISTRY
Vol. 55
No. 6
2015
2.0 1.6
Ratio
1.2 0.8 0.4 0 0
0.01 0.02 0.03 0.04 0.05 Precursor concentration, % C3 =/C3
C4 =/C4
Fig. 4. Dependence of the butylene/butane and propy lene/propane ratios on the precursor content.
GERZELIEV et al. 44 43 42 41 40 39 38 37 36
17.0 16.5 Yield, %
Yield, %
484
16.0 15.5 15.0 14.5
480
500 Temperature, °C
without additive
520
with additive
14.0 480
500 520 Temperature, °C
without additive
with additive
Fig. 5. Temperature dependence of the gasoline yield.
Fig. 6. Temperature dependence of the yield of light gas oil.
stock increased from 20.7 to 26.3% with an increase in temperature from 480 to 520°С under the same condi tions and the PPF and BBF yields increased from 6.2 and 11.7% to 8.5 and 12.4%, respectively. In the crack ing of the precursorcontaining feedstock, the gas yield increased from 22.1 to 28.7%; the yield of PPF, from 6.6 to 9.5%; and the yield of BBF, from 12.7 to 14.0%. An irregular increase in the BBF yield with increasing temperature was also observed; the incre ment was 8.5 rel. % at 480°С and 12.9 rel. % at 520°C. This behavior is presumably due to the enhancement of the cracking reactions of acyclic hydrocarbons with increasing temperature to 520°С. The yield of hydrogen increases with an increase in temperature in both cases: from 0.40 to 0.68 rel. % in the cracking of the pure distillate and from 0.29 to 0.43% in the cracking of the distillate with the precur sor. The reduction in the yield of hydrogen in the case of cracking in the presence of the additive confirms the data obtained in the other sets of experiments and may indicate the enhancement of the hydrogenation reac tions causing hydrogen consumption. The decrement in the hydrogen yield during the cracking with the additive was also found to depend on temperature. The hydrogen yield decrements at 480, 500, and 520°С were 28, 32, and 35 rel. %, respectively. The optimal cracking temperature on the basis of the total gas and gasoline yield was taken to be 500°С. To assess the performance of the precursor under conditions close to industrial conditions, a set of experiments was performed on the pilot catalytic cracking unit described in [12]. Vacuum distillate con taining 0.05% precursor (on a Mo basis) was fed for 4 h. The results are shown in Table 1. In the case of the precursorcontaining feedstock, neither the conversion nor the yield of coke and light gas oil noticeably changed. The yield of the gasoline fraction decreased insignificantly, from 50.7 to 49.0%. At the same time, the gas yield increased from 16.0 to 17.0%. With a long time on stream (4 h) of the feed stock containing the precursor, the material balance
changed slightly. When the precursor supply stopped, the characteristics of the process took their initial val ues. Note that the product gas is enriched in saturated hydrocarbons in the case of distillate cracking with the precursor, as has been shown above in the experiments using the laboratory unit. The propylene/propane ratio obtained by cracking decreases from 6.6 to 5.2 in the presence of the additive, and the buty lenes/butanes ratio decreases from 1.6 to 1.4. The catalyst was sampled for determination of molybdenum content and acidity, which were mea sured by atomic absorption spectroscopy (AAS) and ammonia temperatureprogrammed desorption (TPD), respectively. The measurement results are shown in Table 2. The mass of molybdenum deposited on the catalyst corresponds within the limits of measurement error to that of molybdenum introduced with the feedstock. This coincidence indicates the complete transfer of molybdenum from the feedstock onto the catalyst sur face. Supplying the precursor together with the feed stock to contact the catalyst increased the total acidity of the catalyst from 24.6 to 48.1 mmol NH3/g within the first hour. Longer feeding of molybdenum (to 4 h) reduced the acidity from 48.1 to 35.2 mmol NH3/g. This changed the ratio between acid sites of different strengths: the proportion of strong acid sites slightly increased and that of acid sites of moderate strength decreased. The increase in the catalyst acidity during the cracking of the precursorcontaining feedstock and the increase in the proportion of strong acid sites explain the increase in the gas yield and the decrease in the yield of the gasoline fraction. The presence of the precursor in the feedstock affects the group composition of the product gasoline fraction (Table 3). The character of the effect is not straightforward and depends on the length of the pre cursor supply. After a 1h period of supplying the feed PETROLEUM CHEMISTRY
Vol. 55
No. 6
2015
EFFECT OF CATALYST AND FEEDSTOCK MODIFICATION
485
Table 1. Product yields and characteristics of the catalytic cracking process of vacuum distillate with the precursor on a pilot unit Product withdrawal time point before precursor supply
Product
Gas, including, dry gas PPF including propylene BBF including butylenes Gasoline (IBP–200°C fraction) Light gas oil (200–320°C fraction) Residue (fraction >320°C) Coke Loss
16.0 1.8 4.5 3.9 9.7 6.0 50.7 18.5 7.0 3.6 4.0
Conversion (gas + gasoline)
66.7
with precursor supply for (h) 1
2
Product yield, % 17.4 17.3 2.2 2.0 5.1 5.4 4.2 4.5 10.2 10.2 5.8 5.8 49.0 48.6 18.4 18.6 7.5 7.5 3.5 3.7 4.0 4.0 Process characteristics, % 66.4 66.0
3
4
after cessation of precursor supply
17.1 2.2 5.3 4.5 9.6 5.6 49.0 18.0 8.0 3.7 4.0
16.5 1.9 4.9 4.1 9.7 5.8 48.5 18.9 8.4 3.7 4.0
16.1 2.1 4.5 3.9 9.5 6.0 51.5 17.5 6.9 3.8 4.0
67.1
65.0
67.5
Table 2. Physicochemical properties of catalyst samples taken during the experiments in the pilot unit Proportion of weak Proportion and mediumstrength of strong sites, mol % sites, mol %
Catalyst sampling time
Molybdenum content, ppm
Total acidity, µmol NH3/g
Before molybdenum supply
20
24.59
82.2
17.8
after 1 h
170
48.07
81.8
18.2
after 2 h
320
42.17
80.4
19.6
after 3 h
480
37.35
80.4
19.6
after 4 h
630
35.17
81.2
18.8
With molybde num supply
stock containing the precursor, the amount of isopar affinic–naphthenic hydrocarbons increased from 45.7 to 48.3% and that of aromatics content, in contrast, decreased from 28.0 to 25.5%. The amount of unsat urated hydrocarbons remains almost unchanged (23.4–23.6%). By increasing the duration of supply from 1 to 4 h, the amount of aromatic hydrocarbons was increased from 25.5 to 31.3% and that of naph thenes and isoparaffins, conversely, reduced from 48.3 to 43.1%. After cessation of the precursor supply with the feedstock, the aromatics content in gasoline increased from 28 to 30.7% as compared with the gas oline obtained before introduction of the precursor. The amount of isoparaffinic–naphthenic hydrocar bons decreased from 45.7 to 43.8% in this case. PETROLEUM CHEMISTRY
Vol. 55
No. 6
2015
In summary, the data obtained in this study show that modification with molybdenum disulfide affects both the particulate properties of the feedstock and the acidity of the microspherical zeolitecontaining cata lyst. In the presence of the additive, not only the yield of cracking products, but also their hydrocarbon com position changes, indicating that molybdenum disul fide exhibits hydrogenating activity under catalytic cracking conditions. Depending on the range of pro cesses in refineries and seasonal demand for particular products, modifying with molybdenum disulfide dur ing the catalytic cracking process can be directed to the reduction in the yield of light gas oil, the olefin content in gasoline fractions, the yield of hydrogen and, hence, the wet gas compressor load.
486
GERZELIEV et al.
Table 3. Hydrocarbon group composition of gasoline fractions obtained by cracking, % Fraction sampling time Hydrocarbon group
nParaffins Isoparaffins + naphthenes Aromatic Unsaturated
before precursor supply
with precursor supply for (h) 1
2
3
4
2.7 45.7 28.0 23.6
2.8 48.3 25.5 23.4
2.6 46.8 27.3 23.3
2.7 43.7 29.0 24.6
2.5 43.1 31.3 23.1
ACKNOWLEDGMENTS The authors are indebted to G.A. Shandryuk and Z.D. Voronina, senior researchers at the Topchiev Institute, for assistance in thermogravimetric and AAS analyses, respectively. REFERENCES 1. L. Zhongqing, Fu Jun, He Mingyuan, and Li Ming gang, Prepr. Pap.Am. Chem. Soc. Div. Fuel Chem. 48, 712 (2003). 2. Cracking of Petroleum Fractions on Zeolite Catalysts, Ed. by S. N. Khadzhiev (Khimiya, Moscow, 1982) [in Rus sian]. 3. A. G. Alikin, N. K. Matveeva, and Z. I. Syunyaev, Khim. Tekhnol. Topl. Masel, No. 5, 9 (1989). 4. L. C. Caero, L. C. Oróñez, J. Ramírez, and F. Pedraza, Catal. Today 107, 657 (2005).
after cessation of precursor supply 2.4 43.8 30.7 23.2
5. H. Topsøe, B. S. Clausen, and F. E. Massoth, Hydrotreating Catalysis: Science and Technology (Springer, Berlin, 1996). 6. A. S. Maloletnev and M. Ya. Shpirt, Ross. Khim. Zh. 52 (6), 44 (2008). 7. A. S. Maloletnev, G. S. Golovin, N. V. Krasnobaeva, et al., RU Patent No. 2 324 655 (2008). 8. Y. P. Suvorov, S. N. Khadzhiev, and A. A. Krichko, Pet. Chem. 40, 169 (2000). 9. S. N. Khadzhiev, I. M. Gerzeliev, V. M. Kapustin, et al., Pet. Chem. 51, 32 (2011). 10. M. Ya. Shpirt and L. A. Zekel’, http://www.ptechnol ogy.ru/Science/Science29.html 11. OST (Branch Standard) 380116178: Cracking Cata lysts, Microspherical and Ground: Test Methods (1978) [in Russian]. 12. I. M. Gerzeliev, K. I. Dement’ev, A. Yu. Popov, et al., in Proceedings of IX School–Conference of Young Petro chemical Scientists (Zvenigorod, 2008) [in Russian]. Translated by S. Zatonsky
PETROLEUM CHEMISTRY
Vol. 55
No. 6
2015