In Situ Upgrading of Light Fluid Catalytic Cracking Naphtha for

In China, because of the lack of hydrotreating, catalytic reforming, alkylation, ... (20) Hence, a riser reactor is better equipment for studying the ...
3 downloads 0 Views 2MB Size
Article pubs.acs.org/IECR

In Situ Upgrading of Light Fluid Catalytic Cracking Naphtha for Minimum Loss Jinhong Zhang, Honghong Shan, Xiaobo Chen, Chunyi Li, and Chaohe Yang* State Key Laboratory of Heavy Oil Processing, China University of Petroleum, Qingdao, 266580, People’s Republic of China S Supporting Information *

ABSTRACT: The key to reducing the olefin content in fluid catalytic cracking (FCC) gasoline is to upgrade the olefin-rich light FCC naphtha (LCN). To minimize the naphtha loss, several parameters were investigated in a pilot-scale riser FCC apparatus. The results indicate that, besides the reaction temperature, the catalyst-to-oil ratio, and the catalyst type, the boiling range and the olefin content of LCNs also have significant influence on the upgrading effect. Moreover, a relatively short residence time is beneficial for efficiently upgrading LCNs. In addition, the influence of the reactor structure should be brought to our attention. When a novel structurally changed reactor with a multinozzle feed system was used, significantly increased olefin conversion and decreased naphtha loss can be achieved. The calculation of hydrogen balance indicates that, because of the decrease of dry gas and coke yields, more hydrogen in the feed can be distributed into the desired products.

1. INTRODUCTION Fluid catalytic cracking (FCC) units play a pivotal role in converting heavy oil into motor fuels, especially in China, where ∼75% of the gasoline comes from FCC gasoline.1 However, the olefin content of FCC gasoline is usually as high as 40−65 vol %,2,3 which cannot meet the increasingly strict environmental regulations. In China, because of the lack of hydrotreating, catalytic reforming, alkylation, and MTBE or ETBE units, the refineries face a great challenge for improving gasoline quality. Because of the high operating flexibility and low investment and operating cost, upgrading gasoline through the FCC process has raised wide attention. Some studies put attention on developing new FCC catalysts,4,5 but the effects in commercial application are limited. More research focused on developing novel FCC processes, such as the maximizing isoparaffins process (MIP),6 the flexible dual-fluid catalytic cracking process (FDFCC),7 the subsidiary riser FCC process (SRFCC), 8,9 and the two-stage riser FCC process (TSRFCC).10 Verstraete et al.11 used an adiabatic circulating fluidized bed (CFB) pilot plant to study the naphtha recycling operations and found that cracking light feeds in an additional riser have an effect similar to that of a catalyst cooler. Hence, the catalyst-to-oil ratio (CTO) can be increased, leading to a higher conversion of heavy oil in the primary riser and to a better yield structure. Recycling naphtha in the FCC process is relatively inexpensive; however, the loss of gasoline yield cannot be neglected. Research found that olefin-rich light FCC naphthas (LCN) crack more efficiently than full-range naphthas11−15 but have lower conversions.11,16,17 In these studies, the main purpose is to crack LCNs into propene production. In order to increase the conversion of LCNs, they are usually injected from the bottom of the riser before the gas oil feed injection point or a parallel riser under higher severity, as well as the ZSM-5 zeolite based additives are added into the catalyst system.18 However, these measures lead to not only higher LPG and propene yields, but also a significant increase in dry gas yield. Hence, the © 2013 American Chemical Society

pyrolysis process will cause more gasoline loss. Because of the higher price of the propene product, previous research usually focused on the pyrolysis of LCNs. However, gasoline is still the main product of FCC units; moreover, the price gap between gasoline and propene is closing recently. Therefore, it is necessary to investigate how to upgrade LCNs for minimum loss, especially in China. Corma et al.12,13 have studied the influences of temperature, coke-on-catalyst, and ZSM-5 based additive and found that low temperatures and the presence of coke on catalyst would promote hydrogen transfer reactions rather than cracking reactions, reducing olefin content of naphtha with less loss, while ZSM-5 zeolite would promote protolytic cracking, increasing the dry gas yield. Nevertheless, their experiments were carried out in MicroDowner or MAT unit and mainly focused on the propene product. Ouyang et al.19 investigated the effect of operating conditions on the process of upgrading FCC gasoline and found that a low reaction temperature and a large CTO are beneficial for decreasing the content of olefins in gasoline. However, their attention was mainly put on the variation of gasoline compositions without considering the naphtha loss. Moreover, they pointed out the fact that, because of the longer residence time and more intense backmixing of catalyst and oil vapor, the olefin content of gasoline obtained from a fixed fluidized bed is often lower than that from a riser reactor, as in the case of the fixed-bed MAT-type reactor.20 Hence, a riser reactor is better equipment for studying the process of gasoline reformation. Wang et al.9 studied the process of upgrading gasoline in a subsidiary riser at moderate severity, and they found that decreasing the temperature difference between gasoline and regenerated catalysts could reduce the dry gas and coke yields, as well as improve the Received: Revised: Accepted: Published: 6366

February 20, 2013 April 19, 2013 April 22, 2013 May 6, 2013 dx.doi.org/10.1021/ie4005496 | Ind. Eng. Chem. Res. 2013, 52, 6366−6376

Industrial & Engineering Chemistry Research

Article

Table 1. Properties of Light FCC Naphthas item density at 20 °C (kg/m3) composition of recovered naphtha (wt %) n-paraffins i-paraffins olefins naphthenes aromatics i-P/n-P ASTM D2887 distillation (°C) IBP 10%/30% 50%/70% FBP

LCN1

LCN2

LCN3

LCN4

637.5

650.1

632.5

727.0

5.77 47.19 46.89 0.15 0 8.18

4.37 21.51 73.67 0.45 0 4.92

6.51 35.54 57.77 0.18 0 5.46

5.28 14.28 63.99 3.09 13.36 2.70

22 31/35 41/47 70

25 34/39 45/51 75

23 32/36 41/46 70

20 25/33 49/76 146

product condensing and measurement system, a pneumatic control system, and a computer control system (which shows the main operation parameters and can be used to adjust these parameters conveniently). This unit can be operated continuously, similar to the commercial units; moreover, the experimental results are consistent with the industrial production, and it has afforded the design data for more than five commercial plants. During all tests, the mass losses were all 110 μm

value 72 153 0.16 980 0.1 11.9 45.8 23.0 19.2 6367

dx.doi.org/10.1021/ie4005496 | Ind. Eng. Chem. Res. 2013, 52, 6366−6376

Industrial & Engineering Chemistry Research

Article

Figure 1. Schematic of the pilot-scale riser FCC apparatus.

that the protolytic cracking is more favorable in the mediumpore zeolite (ZSM-5) than in the large-pore zeolite (Y). Wielers et al.26 also found that, with decreasing pore dimensions, the relative contribution of the protolytic cracking route increases, compared to the β-scission route. In addition, because of the bimolecular mechanism involving a hydride-transfer step, it is favored by higher acid site density. Thus, decreasing the aluminum content of zeolite will enhance the protolytic cracking.26 Overall, in order to minimize the dry gas yield, in the LCN upgrading process, the favorable operating conditions are low reaction temperature, low regenerated catalyst temperature, and the use of a Y-based catalyst. In addition, a feedstock with higher olefin concentration (LCN) is beneficial. When conditions are chosen to minimize thermal and monomolecular reactions, the bimolecular reactions will be relatively enhanced. The bimolecular mechanism will give less dry gas, more gasoline, and lower olefins, but probably more catalytic coke.25 There are five main types of coke identified in catalytic cracking:27 additive coke, contaminant coke, and catalyst-to-oil coke are negligible in our studies, while thermal coke is less important than catalytic coke, because of the lower extent of thermal cracking in the moderate operating conditions. According to the studies of den Hollander et al.28 and Dupain et al.,29 the majority of coke are formed during the first contact time of hot catalyst and oil vapor; therefore, decreasing the temperature difference between oil and regenerated catalysts can reduce the yields of coke obviously.9 After the initial time, the coke can be generated by bimolecular hydrogen transfer reactions.30 On the other hand, hydrogen transfer reactions are desirable for saturating olefins in the LCN

Figure 2. Formation of dry gas via protolytic cracking.

be taken to restrain the protolytic cracking for reducing the dry gas yield. Haag and Dessau23 found that the activation energy for the protolytic cracking was higher than that for the β-scission; thus, the monomolecular cracking is favored at higher temperatures. Moreover, they showed that protolytic cracking becomes more prominent with decreasing reactant concentrations, which is due to the monomolecular character of the protolytic and the bimolecular nature of β-scission.25 Furthermore, since the bimolecular mechanism involves a larger transition state, the pore size of the zeolite will influence the relative ratio of the protolytic and bimolecular cracking. Haag and Dessau23 found 6368

dx.doi.org/10.1021/ie4005496 | Ind. Eng. Chem. Res. 2013, 52, 6366−6376

Industrial & Engineering Chemistry Research

Article

can be obtained. Moreover, the increased catalyst density can provide more acid sites to compensate for the losses caused by coke deposition. In addition, a certain degree, gas backmixing can enhance the conversion of olefins in LCNs, which are more difficult to react than gas oils. Therefore, changing the structure of the reactor is conducive to gasoline reformation. The gas−solid mixing behavior in fluid beds has been studied for several decades.40−42 In FCC units, the contact behavior of oil vapor and hot catalyst will significantly influence the cracking reactions and the product distribution. Research found that increasing the number of nozzles can improve the product distribution and decrease the ratio of thermal cracking.43 Thus, in the commercial FCC units, the multinozzle feed system has been widely used. However, because of the scale of experimental units and the atomization of heavy oil, the feed system is difficult to be carried out, especially in MicroDowner or MAT unit. In our work, since the experiments were carried out in a pilot unit, and the light FCC gasoline can be vaporized easier, we can try to solve this problem by designing a novel reactor. Hence, in order to enhance the contact of oil vapor with catalyst further and improve the product distribution, a novel structurally changed reactor with a multinozzle feed system was designed and machined. On the basis of the above analysis, in the experiments, the reaction temperature was 480−500 °C (much lower than the pyrolysis temperature) and the temperature of regenerated catalyst was set at a relatively low temperature (590 °C); therefore, the catalyst-to-oil ratio (CTO) increased to ∼16 (the calculation method of CTO has been described in the literature22). A large CTO is favorable for hydrogen transfer reactions; moreover, it can keep the catalyst with a low coke coverage and retain the activity, so the influence of catalyst decay can be neglected. 3.2. Influence of Boiling Range. The boiling range of LCN can be seen to vary from one refinery to another.1,11,44 Different boiling ranges will lead to varying compositions of LCNs and further influence the reactions. However, little attention has been focused on this. In LCN2, the C5 hydrocarbons account for 92%, and the C5= constitute ∼69%, while only 1.3% of C6= can be found and no aromatics were detected. In comparison, LCN4, with a higher boiling range, contains more olefins, aromatics, and naphthenes with a carbon number of 6 and higher, and the olefin content still as high as 64%. From Table S1 (see the Supporting Information), it can be seen that the LCN2 should react under longer residence time to reach a close conversion. This is because the reactivity of C5 olefins is considerably lower than that of olefins with longer chains.16,17,45 Indeed, cracking a C5 olefin through classical βscission route necessarily involves the formation of a primary carbenium ion, which significantly reduces its overall reaction rate.11 Because of the higher olefin concentration of LCN2, the monomolecular cracking reactions are restrained, which can be inferred from the significant decrease in the cracking mechanism ratio (CMR), as proposed by Wielers et al.26 [CMR is defined as the volume ratio of the sum of C1 and C2 yields to isobutane yield: CMR = (C1 + C2)/i-C04.] Thus, lower yield and selectivity of dry gas can be obtained. In the liquefied petroleum gas (LPG), the selectivity of propene in the LCN4 cracking process was much higher than that of LCN2, because of the higher hexene content in LCN4. Buchanan et al.45 studied the cracking behavior of pentene, hexene, heptene, and octene over HZSM-5 and found that,

upgrading process. Since the hydrogen transfer is an exothermic reaction with a slower reaction rate, it will benefit from a relatively lower reaction temperature31 and longer residence time.32 From the theoretical analysis, it is difficult to enhance the hydrogen transfer reactions while avoiding the generation of catalytic coke. Considering that prolonging the residence time will cause the increase of dry gas and coke yields, a relatively short residence time should be chosen, considering both the olefin conversion and the naphtha loss, which will be analyzed in the following section. In our work, the main purpose is to upgrade LCN with minimum loss, while preserving the octane value. The best choice is to convert olefins to isoparaffins and aromatics, because the loss of gasoline octane number caused by olefin reduction can be compensated by the increase of isoparaffins and aromatics that have the higher octane number.1−3 In the upgrading process, olefins in LCNs have a significantly higher reactivity than other hydrocarbons, but how does one convert olefins to isoparaffins and aromatics, as many as possible, instead of being cracked? Corma et al.13 found that, despite a coke-on-catalyst content of 1.2 wt %, the Y zeolite still has enough activity for transforming olefins into paraffins through hydrogen transfer reaction. Zhang et al.33 and Yuan et al.34 found that a small coke-on-catalyst content (no more than 0.5 wt %) could increase the gasoline yield as well as reduce the dry gas and coke yields during the process of FCC gasoline upgrading, while only little influence on olefin reduction was observed. Research found that reactions proceed under the bimolecular mechanism will give less dry gas, more gasoline, and lower olefins (compare to the protolytic cracking).25 Therefore, using the partially coked catalyst to upgrade LCNs could effectively enhance the bimolecular reactions and reduce the naphtha loss. This may be because the coke deposition changed the acid strength distribution and the strong acid sites associated with dry gas and coke formation were weakened.9 However, the coke-on-catalyst content may reduce the site density, which will decrease the reaction rate for bimolecular hydrogen transfer reactions;31 therefore, the coke content on catalyst should not be too high. In our previous work,35 in the development of MFT FCC process, the semispent catalyst after heavy oil catalytic cracking was used to upgrade LCNs, but it cannot be applied in the single process of gasoline upgrading. Wang et al.9 found that the spent catalyst after gasoline reformation still has abundant acid sites; therefore, it can be further used to upgrade heavy oils. UOP developed an RxCat technology that mixes spent and regenerated catalyst in a blending vessel located at the bottom of the riser.36,37 However, additional equipment is essential. Here, we want to afford a simpler way by changing the structure of the riser reactor. Wang et al.38 studied the gas−solids flow patterns of the MIP reactor in a cold-flow circulating fluidized bed (CFB) pilot unit and found that, with an enlarged section, the solid density would have a significant increase, which may be due to the decreased gas−solid velocity and the anabatic particle slip.39 In the heavy oil catalytic cracking process, the gas−solid backmixing will aggravate the product distribution, because of the decreased catalytic activity of the spent catalyst for heavy oil and the overcracking of light oil. However, it is beneficial for upgrading LCNs. First, the slip of coked catalysts makes the enlarged section have a function of catalyst blending, where the regenerated catalyst and the spent catalyst after gasoline reformation are blended; thus, catalysts with better properties 6369

dx.doi.org/10.1021/ie4005496 | Ind. Eng. Chem. Res. 2013, 52, 6366−6376

Industrial & Engineering Chemistry Research

Article

Figure 3. Influence of boiling range on the composition of upgraded LCNs.

Figure 4. Formation of aromatics.

aromatics can only be seen in LCN4-upgraded naphtha. For LCN2, more olefins were converted to isoparaffins, whereas for LCN4, more of the olefins were converted to aromatics (except the cracking part). The initial composition of LCNs will markedly influence the relative ratios of the isomerization, hydrogen transfer, and aromatization reactions. Taking a closer look at the carbon number distribution, we can see that no naphthenes (Figure 3d) with a carbon number higher than 6 exist in LCN2-upgraded naphtha, but the carbon number of aromatics (Figure 3e) can be as high as 11. Figure 4 shows two formation routes of aromatics.46,47 It can be seen

when cracking hexene, the volume selectivity of propene in the products was much higher than that observed when cracking pentene, and both of them were higher than the selectivity of butene. By contrast, from the cracking of octene, more butene can be obtained, and equal amounts of propene and butene were generated in the process of heptene cracking. The conclusion seems to be consistent with our experimental results, although the Y-type zeolite was used. Figure 3a shows that, after reformation, the isoparaffin content significantly increased both in two LCNs, while the olefin content dramatically decreased. The sharp increase in 6370

dx.doi.org/10.1021/ie4005496 | Ind. Eng. Chem. Res. 2013, 52, 6366−6376

Industrial & Engineering Chemistry Research

Article

Figure 5. Influence of olefin content on the composition of upgraded light cracked naphthas (LCNs).

that aromatics can be formed from olefins that undergo route (1) (without naphthenes), and naphthene is the intermediate in the second route. Thus, higher concentrations of naphthenes in LCN4 can enhance the formation of aromatics. From the increase of naphthenes in LCN4-upgraded naphtha (Figure 3d), it can be inferred, in the LCN2 upgrading process, C6+ naphthenes would be generated, but further cracked or formed aromatics, even though they cannot be detected in the product. Moreover, the higher contents of hexene, heptene, and octene in LCN4 would increase the generation rate of C6+ naphthenes. Figure 3f shows that olefins with a carbon number of 7 and higher were almost completely converted, which was also observed by Corma et al.13 and Verstraete et al.11 To evaluate the different upgrading processes systematically, five indexes were defined. The first one is the olefin conversion (OC), which was defined as the ratio of the converted olefin content to the olefin content before reformation multiplied by 100:15 OC (%) =

converted olefin content × 100 olefin content before reformation

Two additional indexes were defined to quantitatively describe how many olefins were converted to isoparaffins and aromatics, which are mainly the increased components in the upgraded naphthas. The selectivity of isoparaffins (Si‑P) is the ratio of the increased isoparaffins to the converted olefin content multiplied by 100. Similarly, the selectivity of aromatics (SAr) is the ratio of the increased aromatics to the converted olefin content multiplied by 100. Si‐P (%) =

increased isoparafffin content × 100 converted olefin content

(2)

SAr (%) =

increased aromatics content × 100 converted olefin content

(3)

Thus, the cracking ratio of olefins can be estimated. The loss rate (RL) was defined as the ratio of the sum of dry gas and coke yields to the olefin conversion multiplied by 100, which can be used to compare the loss at a certain olefin conversion under different tests.

(1) 6371

dx.doi.org/10.1021/ie4005496 | Ind. Eng. Chem. Res. 2013, 52, 6366−6376

Industrial & Engineering Chemistry Research RL (%) =

sum of dry gas and coke yields × 100 olefin conversion

Article

expected, higher olefin concentration would restrain the protolytic cracking,26 which can be inferred from the lower value of the CMR index. Moreover, a higher concentration of olefins would lead to more-intense polymerization reactions and form more C7 and C8 oligomers, and then the higher selectivity of butene can be reached.45 The detailed composition of feed and upgraded LCNs can be seen in Figure 5. In both two LCNs, the main olefin is pentene, which mainly converted to isopentane through isomerization and hydrogen transfer reactions (except the cracking part). Because of the higher olefin concentration, more oligomerizations happened and more diesel and aromatics with higher carbon numbers were generated. However, the quantity is limited. The calculated results (in Table S1 in the Supporting Information) show that the higher olefin concentration would reduce the olefin conversion, which may be due to the competing effect. Moreover, it will lead to a higher cracking ratio of olefins and a lower reforming efficiency, because of the higher reactivity of olefins than paraffins. Therefore, in the industrial production, measures should be taken to enhance the conversion of olefins, when a LCN with higher olefin content is recycled. It should be noted that the lower aromatic selectivity of LCN2 is due to more olefins in LCN2 being converted. Indeed, more aromatics were generated in the upgrading process of LCN2, which can be seen in Figure 5e. Compared to the study of Wang et al.,9 the influence of olefin content on fullrange gasoline shows different laws. When upgrading the higholefin-content (50 wt %) Shengli FCC gasoline (compared to the Fushun FCC gasoline with 39 wt % olefin content), a higher olefin conversion and reforming efficiency can be achieved, but the loss rate also increased. 3.4. Influence of Residence Time. The residence time in a fixed fluidized bed reactor is much longer, and the gas−solid contacting efficiency in a small-scale fixed bed is much higher, than in a riser reactor, which is beneficial to hydrogen transfer reactions for olefin reduction. Therefore, in this experimental equipment, the olefin content of naphtha is often lower than that in the riser reactor.19 To avoid the interference, this work studied the influence of residence time from 1.7 s to 3.3 s in a pilot-scale riser reactor. As can be seen from Table S2 in the Supporting Information, the dry gas and coke yields both increased as the residence time was prolonged, but their selectivities show different changing tendency. Because C3 and C4, especially isobutane and isobutene, are the characteristic products of bimolecular catalytic cracking, while methane, ethane, and ethene are the typical products of thermal cracking,48 the degree of thermal cracking can be described by the thermal cracking index (TCI), which was defined as the weight ratio of the sum of C1 and C2 yields to the sum of isobutane and isobutene yields:22,49

(4)

To describe the efficiency of gasoline reformation, the reforming efficiency (e) was defined as the ratio of the olefin reduction (between feed and product) to the conversion; higher values of e mean that the gasoline can be upgraded at less expense of gasoline yield (in this work, the main purpose is to upgrade LCNs for minimum loss; thus, the LPG product is not included in the index). The calculated results are listed in Table S1 in the Supporting Information. It is clear that the olefin conversion of LCN4 is much higher than that of LCN2, because of the higher reaction activity of olefins with larger carbon numbers.45 Moreover, because of the enhanced aromatization reactions, the aromatic selectivity of LCN4 is higher than that of LCN2, while a lower selectivity of isoparaffin (no more than 10%) can be seen. For LCN2, the majority of the converted olefins (∼65%) were cracked, which is more than that observed for LCN4 (∼10% higher than that of LCN4). This may due to the higher content of aromatics in LCN4, which are more difficult to crack than olefins and isoparaffins, and the enhanced aromatization reactions, which lead to the formation of more aromatics. Although in the LCN2 upgrading process, less dry gas and coke were generated, because of the lower olefin conversion, a higher loss rate was finally obtained. Thus, from the viewpoint of olefin reduction, lower yields and selectivities of dry gas and coke cannot be used to evaluate the process directly. The reforming efficiency of LCN2 is only 1.12, which means that a reduction of 1.12 wt % olefins in the upgraded LCN2 will cause a 1 wt % yield loss of gasoline. By contrast, LCN4 has a higher reforming efficiency (1.56). The experimental results indicate that, in the LCN upgrading process, the interaction of hydrocarbons seems interesting; therefore, it will be further studied in our future work. In the LCN4 upgrading process, the selectivity of aromatics is higher than that in the full-range gasoline upgrading process studied by Wang et al.,9 which was 23%−28%. This may be due to the following three reasons: (1) The microactivity of the catalyst used in our study is higher than that in the literature; thus, the aromatization reactions were enhanced. (2) LCN4 has more olefins, which are the main reactant of hydrogen transfer and aromatization reactions. (3) Lower aromatic content in LCN4 is beneficial to the formation of aromatics, considering the reaction equilibrium. 3.3. Influence of Olefin Content. It is well-known that different feeds, catalysts, and FCC processes will create significantly different compositions of gasolines (or LCNs), even though they have close boiling ranges. Different group compositions, especially the olefin content, will directly influence the cracking behavior of LCNs and the upgrading effect. However, little information about the influence of olefin content on the upgrading behavior of LCNs can be obtained from the open literature. In this work, LCN1 and LCN2, which have similar boiling ranges but the different olefin contents, were investigated. According to Table S1 in the Supporting Information, the olefin content of LCNs significantly influences the feed conversion. Compared to LCN1, LCN2 (with higher olefin content) had a higher conversion, as well as higher dry gas and coke yield, but lower selectivities of dry gas and coke. As

TCI =

C1 + C2 yield i‐C04 + i‐C=4 yield

(5)

A value of TCI < 0.6 means that catalytic cracking is the main reaction, while a value of TCI > 1.2 means that thermal cracking is serious.49 When the residence time increased from 1.7 s to 3.3 s, the values of TCI increased by ∼10%, to 0.304. Thus, it can be inferred, in the LCN upgrading process, the ratio of thermal cracking is limited. Moreover, the thermal cracking reactions can take place when oil vapor travels through the riser. The significant increase of dry gas selectivity indicates that except for thermal cracking, a fair amount of dry gas is 6372

dx.doi.org/10.1021/ie4005496 | Ind. Eng. Chem. Res. 2013, 52, 6366−6376

Industrial & Engineering Chemistry Research

Article

formed via protolytic cracking (the value of CMR increased by ∼20%). Therefore, shortening the residence time is beneficial to reduce the dry gas yield by reducing the monomolecular cracking, as well as the thermal cracking. In contrast, the selectivity of coke showed a downward trend, which indicates that the majority of cokes were generated at the initial time when oil vapor comes into contact with the hot regenerated catalyst.28,29 However, the coke yield still can be reduced by shortening the residence time. The diesel yield shows a downward trend; this is because the oligomeric cracking rate is related to the alkene partial pressure.50 At the first reaction time, because of the high alkene partial pressure, the generation rate of oligomers was higher than its cracking rate. As the alkene partial pressure decreased, more oligomers were cracked, underwent β-scission, or formed aromatic coke through hydrogen-transfer reactions. Figure 6 shows that, when prolonging the residence time from 1.7 s to 3.3 s, only a slight further reduction of olefins can

Figure 7. Schematic of the conventional riser reactor and the novel riser reactor.

vaporization) and a relatively lower reaction temperature was chosen (480 °C). The results in Table S3 in the Supporting Information show that the novel riser reactor has better performance than the conventional riser reactor. The increase of catalyst density in the riser reactor and the enhancement of contact efficiency of oil and catalyst significantly enhanced the bimolecular reactions and restrained the thermal cracking (the value of TCI decreased from 0.245 to 0.162) and monomolecular reactions (the value of CMR decreased from 1.555 to 0.925). Thus, less dry gas was generated and more gasoline can be obtained, which in consistent with the viewpoint of Corma and Orchilles.25 The hydrogen transfer coefficient (HTC), which is defined as the volume ratio of butanes to butenes, can be used to describe the degree of hydrogen-transfer reaction,49,51 which is helpful for reducing the olefin content in naphthas.

Figure 6. Influence of residence time on the composition of upgraded LCN1.

be seen, and the isoparaffins also showed a downward trend. According to Table S2 in the Supporting Information, the olefin conversion only increased by 3% in the additional 1.6 s of reaction time, while more than 8% of the olefins were cracked, and the loss rate also increased by 0.8%; thus, the reforming efficiency decreased from 1.67 to 1.41. The results imply that prolonging the residence time is not a good way to enhance the conversion of olefins. 3.5. Influence of Reactor Structure. 3.5.1. Experimental Results. On a conventional riser reactor, several measures have been taken to minimize the loss during the LCN upgrading process (for example, choosing a relative lower reaction temperature and catalyst temperature, using the catalysts with larger pore diameters, and operating under a large catalyst-to-oil ratio and short residence time conditions). However, the reforming efficiency and olefin conversion are still very low, especially for LCNs with higher olefin content and lower boiling range. With the objective of further improving the reforming effect of LCNs, a novel riser reactor with a multinozzle feed system was designed and machined, which is shown in Figure 7. The diameter of the enlarged section is twice as large as the prelifting section, and there are six LCN feed nozzles distributed along the circumference of the novel riser, while the conventional riser reactor is equipped with only one nozzle. In this work, LCN3 was used as the feed (to avoid the influence of vaporization rates, the LCN was fed after

HTC =

volume of butanes volume of butenes

(6)

As can be seen in Table S3 in the Supporting Information, the value of HTC significantly increased from 0.258 to 0.348, which indicates that the novel reactor can dramatically enhance hydrogen-transfer reactions. However, it will also lead to a higher selectivity of coke. However, because of the lower conversion, the actual coke yield was lower than that of the conventional riser reactor. Figure 8 presents the composition of feed and upgraded naphthas reformed in the conventional and novel riser reactors. Similar to LCN1, in LCN3, the main olefin is pentene. After reformation, only the isoparaffins had a significant increase. Compared to a conventional riser reactor, even though under a lower conversion, the novel reactor still increased the olefin conversion by more than 5%, while more than 25% of the olefins were converted to isoparaffins and aromatics, rather than being cracked (see Table S3 in the Supporting Information). Therefore, the reforming efficiency increased by ∼70%, as well as the loss rate decreased from 5.3% to 3.3%. Moreover, the novel reactor also enhanced the isomerization reactions, thus less n-paraffins were generated (see Figure 8b). 6373

dx.doi.org/10.1021/ie4005496 | Ind. Eng. Chem. Res. 2013, 52, 6366−6376

Industrial & Engineering Chemistry Research

Article

Figure 8. Influence of reactor structure on the composition of upgraded LCN3.

Because of the enhanced bimolecular hydrogen transfer reactions, more aromatics with higher carbon numbers were generated, as can be seen in Figure 8e. 3.5.2. Hydrogen Balance. In the FCC process, all the elements are in balance. Thus, it is desired to distribute more hydrogen into desired products (LPG, gasoline, and diesel) and less hydrogen into low value byproducts (dry gas and coke). From hydrogen balance calculation, we can obtain the effective utilization of hydrogen.52 Table 3 shows the hydrogen distribution in products. The calculation method has been described in the literature in detail.22 As can be seen, when the novel reactor was used, more hydrogen is distributed into the desired products, especially into the gasoline. Thus, a better quality of upgraded naphtha can be expected, and a higher effective utilization ratio of hydrogen (EH) can be obtained. It can be seen that the increase of EH is mainly due to the decrease of dry gas yield, which is the highest-hydrogen-content product in the FCC process.

Table 3. Hydrogen Balance of Different Processes item hydrogen in feedstock (wt %) hydrogen in products (wt %) dry gas LPG gasoline diesel coke sum relative error (%) EH (%)a

reactor 1

reactor 2

15.29

15.29

0.25 2.43 12.35 0.25 0.11 15.39 0.65 97.66

0.13 1.61 13.41 0.28 0.09 15.52 1.50 98.58

a

EH (%) = (H in LPG + H in gasoline + H in diesel)/(H in feedstock).22

4. CONCLUSIONS When upgrading light naphthas in the fluid catalytic cracking (FCC) unit with the commercial Y zeolite-based catalyst, having a certain amount of fractions with the higher boiling range will lead to higher dry gas and coke yields, but it can 6374

dx.doi.org/10.1021/ie4005496 | Ind. Eng. Chem. Res. 2013, 52, 6366−6376

Industrial & Engineering Chemistry Research

Article

(3) Fan, Y.; Bao, X. J.; Lei, D.; Shi, G.; Wei, W. S.; Xu, J. A novel catalyst system based on quadruple silicoaluminophosphate and aluminosilicate zeolites for FCC gasoline upgrading. Fuel 2005, 84 (4), 435−442. (4) Katoh, S.; Nakamura, M.; Skocpol, B. Reduction of olefins in FCC gasoline. In Studies in Surface Science and Catalysts; GehrkOccellie, M. L.; O’Connor, P., Eds.; Elsevier B.V.: Amsterdam, 2001; Vol. 134, pp 141−152. (5) Harding, R. H.; Peters, A. W.; Nee, J. R. D. New developments in FCC catalyst technology. Appl. Catal., A 2001, 221 (1−2), 389−396. (6) Xu, Y. H.; Zhang, J. S.; Long, J.; He, M. Y.; Xu, H.; Hao, X. R. Development and commercial application of FCC process for maximizing iso-paraffins (MIP) in cracked naphtha. Eng. Sci. 2003, 5 (5), 55−58. (7) Meng, F. D.; Wang, L. Y.; Hao, X. R. Technology for reducing olefin in cracked naphtha-FDFCC process. Pet. Process. Petrochem. 2004, 35, 6−10. (8) Bai, Y. H.; Gao, J. S.; Xu, C. M. Study on reaction rules of different processed for decreasing FCC gasoline olefin content. Pet. Refin. Eng. 2004, 34, 7−10. (9) Wang, G.; Yang, G. F.; Xu, C. M.; Gao, J. S. A novel conceptional process for residue catalytic cracking and gasoline reformation dualreactions mutual control. Appl. Catal., A 2008, 341 (1−2), 98−105. (10) Yang, C. H.; Shan, H. H.; Zhang, J. F. Two-stage riser FCC technologies. Pet. Refin. Eng. 2005, 35 (3), 28−33. (11) Verstraete, J.; Coupard, V.; Thomazeau, C.; Etienne, P. Study of direct and indirect naphtha recycling to a resid FCC unit for maximum propylene production. Catal. Today 2005, 106 (1−4), 62−71. (12) Corma, A.; Melo, F.; Sauvanaud, L.; Ortega, F. J. Different process schemes for converting light straight run and fluid catalytic cracking naphthas in a FCC unit for maximum propylene production. Appl. Catal., A 2004, 265 (2), 195−206. (13) Corma, A.; Melo, F. V.; Sauvanaud, L.; Ortega, F. Light cracked naphtha processing: Controlling chemistry for maximum propylene production. Catal. Today 2005, 107−108 (0), 699−706. (14) Li, C. Y.; Yang, C. H.; Shan, H. H. Maximizing propylene yield by two-stage riser catalytic cracking of heavy oil. Ind. Eng. Chem. Res. 2007, 46 (14), 4914−4920. (15) Yang, G. F.; Wang, G.; Gao, J. S.; Xu, C. M. Coke formation and olefins conversion in FCC naphthaolefin reformulation at low reaction temperature. J. Fuel Chem. Technol. 2007, 35 (5), 572−577. (16) den Hollander, M. A.; Wissink, M.; Makkee, M.; Moulijn, J. A. Gasoline conversion: Reactivity towards cracking with equilibrated FCC and ZSM-5 catalysts. Appl. Catal., A 2002, 223 (1−2), 85−102. (17) Buchanan, J. S. Gasoline selective ZSM-5 FCC additives: Model reactions of C6−C10 olefins over steamed 55:1 and 450:1 ZSM-5. Appl. Catal., A 1998, 171 (1), 57−64. (18) Wang, G.; Xu, C. M.; Gao, J. S. Study of cracking FCC naphtha in a secondary riser of the FCC unit for maximum propylene production. Fuel Process. Technol. 2008, 89 (9), 864−873. (19) Ouyang, F. S.; Pei, X.; Zhao, X. H.; Weng, H. X. Effect of operation conditions on the composition and octane number of gasoline in the process of reducing the content of olefins in fluid catalytic cracking (FCC) gasoline. Energy Fuels 2009, 24 (1), 475− 482. (20) Passamonti, F. J.; de la Puente, G.; Sedran, U. Comparison between MAT Flow Fixed Bed and Batch Fluidized Bed Reactors in the Evaluation of FCC Catalysts. 2. Naphtha Composition. Energy Fuels 2009, 23 (7), 3510−3516. (21) Kossiakoff, A.; Rice, F. O. Thermal Decomposition of Hydrocarbons, Resonance Stabilization and Isomerization of Free Radicals. J. Am. Chem. Soc. 1943, 65 (4), 590−595. (22) Wang, G.; Lan, X. Y.; Xu, C. M.; Gao, J. S. Study of optimal reaction conditions and a modified residue catalytic cracking process for maximizing liquid products. Ind. Eng. Chem. Res. 2009, 48 (7), 3308−3316. (23) Haag, W. O.; Dessau, R. M. Duality of Mechanism in Acid Catalyzed Paraffin Cracking. In The Eighth International Congress on

enhance the hydrogen transfer and aromatization reactions, which leads to a higher olefin conversion. Thus, a reduced loss rate and an increased reforming efficiency can be achieved. The higher olefin content in light FCC naphthas (LCNs) will cause higher gasoline conversion and lower dry gas and coke selectivities. However, because of the decreased olefin conversion, the higher loss rate and lower reforming efficiency can be seen. Although prolonging the residence time can increase olefin conversion, more gasoline will be converted to dry gas and coke. Thus, a relatively short residence time is beneficial to efficiently upgrade LCNs. Based on the analysis of the reaction mechanisms and the contact behavior of oil vapor and catalysts in the riser, a novel structurally changed reactor with a multinozzle feed system was designed. Experimental results show that significantly increased olefin conversion and reforming efficiency, as well as improved hydrogen utilization, can be achieved. Therefore, besides optimizing the operating conditions, paying more attention to the research and development of novel reactors is also very important. The research can be combined with the selective hydrodesulfurization (HDS) process to efficiently upgrade FCC gasoline with high olefin content, where the LCNs with high olefin content and low sulfur content are upgraded in the FCC units, while the heavy cracked naphthas (HCNs) with low olefin content and high sulfur content are upgraded by selective HDS. Therefore, the octane loss caused by the olefin reduction in the direct hydrorefining process can be restrained.



ASSOCIATED CONTENT

S Supporting Information *

The influences of boiling range and olefin content on the upgrading process of LCNs, the influence of residence time on the upgrading process of LCN1, and the influence of reactor structure on the upgrading process of LCN3, including operating conditions, product distribution, and the evaluation indexes. This material is available free of charge via the Internet at http://pubs.acs.org.



AUTHOR INFORMATION

Corresponding Author

*Tel.:+86-532-86981169. Fax: +86-532-86981718. E-mail: [email protected]. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS The authors acknowledge the financial support provided by the National Basic Research Program of China (Grant No. 2012CB215006), the National Natural Science Foundation for Young Scholars (Grant No. 21206198), and the Graduate Student Innovation Project of China University of Petroleum (Grant No. CX2013029). We also thank the reviewers for some valuable comments that led to a better manuscript.



REFERENCES

(1) Fan, Y.; Shi, G.; Bao, X. J. A process for producing ultraclean gasoline by coupling efficient hydrodesulfurization and directional olefin conversion. AlChE J. 2012, 59 (2), 571−581. (2) Li, D. D.; Li, M. F.; Chu, Y.; Nie, H.; Shi, Y. H. Skeletal isomerization of light FCC naphtha. Catal. Today 2003, 81 (1), 65− 73. 6375

dx.doi.org/10.1021/ie4005496 | Ind. Eng. Chem. Res. 2013, 52, 6366−6376

Industrial & Engineering Chemistry Research

Article

Catalysis, Berlin, Germany, 1984; Verlag Chemie: Berlin, Germany, 1984; pp 305−315. (24) Ye, Z. J.; Xu, Y. H.; Wang, X. Q. Study on the Compositions of Products from Catalytic Cracking and Thermal Cracking for Heavy Fractions of FCC Gasoline. Acta Pet. Sin. (Pet. Process. Sect.) 2006, 22 (3), 46−53. (25) Corma, A.; Orchillés, A. V. Current views on the mechanism of catalytic cracking. Microporous Mesoporous Mater. 2000, 35−36 (0), 21−30. (26) Wielers, A. F. H.; Vaarkamp, M.; Post, M. F. M. Relation between properties and performance of zeolites in paraffin cracking. J. Catal. 1991, 127 (1), 51−66. (27) Cerqueira, H. S.; Caeiro, G.; Costa, L.; Ramôa Ribeiro, F. Deactivation of FCC catalysts. J. Mol. Catal A: Chem. 2008, 292 (1− 2), 1−13. (28) den Hollander, M. A.; Makkee, M.; Moulijn, J. A. Fluid catalytic cracking (FCC): activity in the (milli)seconds range in an entrained flow reactor. Appl. Catal., A 1999, 187 (1), 3−12. (29) Dupain, X.; Makkee, M.; Moulijn, J. Optimal conditions in fluid catalytic cracking: A mechanistic approach. Appl. Catal., A 2006, 297 (2), 198−219. (30) Guisnet, M.; Magnoux, P. Organic chemistry of coke formation. Appl. Catal., A 2001, 212 (1−2), 83−96. (31) Corma, A.; Miguel, P. J.; Orchilles, A. V. The role of reaction temperature and cracking catalyst characteristics in determining the relative rates of protolytic cracking, chain propagation, and hydrogen transfer. J. Catal. 1994, 145 (1), 171−180. (32) Gao, Y. C.; Zhang, J. S. Study on hydrogen transfer reaction in catalytic cracking. Pet. Refin. Eng. 2000, 30 (11), 34−38. (33) Zhang, X.; Chen, X. B.; Zhang, J. F.; Shan, H. H.; Yang, C. H. Study on Process Conditions for Decreasing FCC Naphtha Olefin Content in Riser. Pet. Refin. Eng. 2005, 35 (4), 8−12. (34) Yuan, Y. X.; Yang, C. H.; Shan, H. H. Study on olefins reduction of gasoline on FCC semi-regenerated catalysts with different coke constant. J. China Univ. Pet. (Ed. Nat. Sci.) 2006, 30, 109−112. (35) Zhang, J. H.; Shan, H. H.; Chen, X. B.; Li, C. Y.; Yang, C. H. Multifunctional Two-Stage Riser Catalytic Cracking of Heavy Oil. Ind. Eng. Chem. Res. 2013, 52 (2), 658−668. (36) Wegerer, D. A.; Lomas, D. A. FCC feed contacting with catalyst recycle reactor. U.S. Patent 5,451,313, 1995. (37) Spretz, R.; Sedran, U. Operation of FCC with mixtures of regenerated and deactivated catalyst. Appl. Catal., A 2001, 215 (1−2), 199−209. (38) Wang, X.; Gao, S.; Xu, Y.; Zhang, J. Gas−solids flow patterns in a novel dual-loop FCC riser. Powder Technol. 2005, 152 (1−3), 90−99. (39) Lu, B.; Wang, W.; Li, J. H.; Wang, X. H.; Gao, S. Q.; Lu, W. M.; Xu, Y. H.; Long, J. Multi-scale CFD simulation of gas−solid flow in MIP reactors with a structure-dependent drag model. Chem. Eng. Sci. 2007, 62 (18−20), 5487−5494. (40) Gilliland, E. R.; Mason, E. A. Gas and Solid Mixing in Fluidized Beds. Ind. Eng. Chem. 1949, 41 (6), 1191−1196. (41) van Deemter, J. J. Mixing and contacting in gas-solid fluidized beds. Chem. Eng. Sci. 1961, 13 (3), 143−154. (42) Li, T. W.; Zhang, Y. M.; Grace, J. R.; Bi, X. T. Numerical investigation of gas mixing in gas-solid fluidized beds. AlChE J. 2010, 56 (9), 2280−2296. (43) Chen, J. W.; Cao, H. C. Catalytic Cracking Technology and Engineering, 2nd Edition; SINOPEC Press: Beijing, China, 2005; p 1164. (44) Passamonti, F. J.; de la Puente, G.; Sedran, U. Reconversion of Olefinic Cuts from Fluidized Catalytic Cracking Naphthas. Ind. Eng. Chem. Res. 2004, 43 (6), 1405−1410. (45) Buchanan, J. S.; Santiesteban, J. G.; Haag, W. O. Mechanistic considerations in acid-catalyzed cracking of olefins. J. Catal. 1996, 158 (1), 279−287. (46) Chen, J. W.; Cao, H. C. Catalytic Cracking Technology and Engineering, 2nd Edition; SINOPEC Press: Beijing, 2005; p 133. (47) Guisnet, M.; Gnep, N. S.; Aittaleb, D.; Doyemet, Y. J. Conversion of light alkanes into aromatic hydrocarbons: VI.

Aromatization of C2−C4 alkanes on H-ZSM-5Reaction mechanisms. Appl. Catal., A 1992, 87 (2), 255−270. (48) Greensfelder, B. S.; Voge, H. H.; Good, G. M. Catalytic and Thermal Cracking of Pure Hydrocarbons: Mechanisms of Reaction. Ind. Eng. Chem 1949, 41 (11), 2573−2584. (49) Chen, J. W.; Cao, H. C. Catalytic Cracking Technology and Engineering. 2nd ed.; SINOPEC Press: Beijing, 2005; p 154−155. (50) Williams, B. A.; Babitz, S. M.; Miller, J. T.; Snurr, R. Q.; Kung, H. H. The roles of acid strength and pore diffusion in the enhanced cracking activity of steamed Y zeolites. Appl. Catal., A 1999, 177 (2), 161−175. (51) de Jong, J. I. Hydrogen transfer in catalytic cracking. Presented at the Ketjen Catalyst Symposium, Scheveningen, The Netherlands, 1986; Paper F-2. (52) Meng, X. H.; Ren, J. D.; Xu, C. M.; Gao, J. S.; Liu, Z. C. Hydrogen balance for catalytic pyrolysis of atmospheric residue. Fuel Process. Technol. 2009, 90 (4), 616−620.

6376

dx.doi.org/10.1021/ie4005496 | Ind. Eng. Chem. Res. 2013, 52, 6366−6376