Modeling of the catalytic cracking: Catalyst deactivation by coke and heavy metals

Fuel Processing Technology 200 (2020) 106318

Contents lists available at ScienceDirect

Fuel Processing Technology journal homepage: www.elsevier.com/locate/fuproc

Research article

Modeling of the catalytic cracking: Catalyst deactivation by coke and heavy metals

T

⁎

Galina Nazarova , Elena Ivashkina, Emiliya Ivanchina, Alexandra Oreshina, Irena Dolganova, Mariya Pasyukova National Research Tomsk Polytechnic University, Division for Chemical Engineering, 30 Lenin av., 634050 Tomsk, Russia

A R T I C LE I N FO

A B S T R A C T

Keywords: Catalytic cracking Mathematical modeling Coke Heavy metals Deactivation

This paper proposes a model of the cracking process considering the catalyst deactivation by Ni, V and coke. The developed model is sensitive to the feedstock composition and describes the kinetics of cracking reactions leading to coke formation, the structural and selective properties of the catalyst. It also reflects the main technological parameters. The forecast calculations showed that when the resins and Ni contents in the feedstock increase by 4.2 wt% and 0.6 ppm, the coke contents on the catalyst increase by 0.75 and 0.32 wt% wt. under the other equal conditions. The catalyst activity decreases by 4.4% relative to initial value along with increasing the V content in the feedstock by 1.9 ppm due to its dealumination. If the Ni with V co-presence in the catalytic cracking feedstock and the Ni content increases by 0.6 ppm, the V destructive effect reduces by 1.2% due to reaction of Ni with the vanadic acid, also Ni on the catalyst increases the catalyst dehydrogenation activity. According to the calculations performed, the yield of the gasoline fraction changes by 4.43 wt%, depending on the feedstock composition (CSH/CAH = 1.6–1.8 units), other things being equal.

1. Introduction Currently, the development and modernization of new and existing catalysts for the heavy petroleum fractions and residual charge stock refining [1–12] is considered to be significant. Developing of catalytic processes, including hydrocracking and catalytic cracking, stems from the need to increase the depth of oil refining, the production of motor fuel components and petrochemical feedstock. A considerable increase in the light petroleum product and light olefins yields was achieved in the formative period of catalytic cracking by passing from tableted catalysts based on natural clays to microspheric zeolite-containing [13] and, also, by the process improvement (fixed-bed catalyst - moving-bed of bead catalyst- fluid bed of microspheric catalyst - riser). Nowadays, the various types of feedstock are involved into catalytic cracking, for example, vacuum and atmospheric gas oils, heavy residues of secondary refining processes, fuel oils, etc. Generally, the content of aromatics, resins and heavy metals increases along with the fractions boiling point and effecting the catalyst deactivation and disrupting the catalyst performance [14–21]. Herewith, the catalytic cracking efficiency is determined primarily by the activity, selectivity and stability of microspheric zeolite-

⁎

containing catalysts [22]. Catalysts deactivation by coking [23] is a reversible process, excepting for the graphitized coke formation [24], as the catalytic properties are partially restored during regeneration using the oxygen-containing mixtures. An irreversible loss of the catalyst activity and selectivity stems from the catalyst poisoning by heavy metals, in particular, by Ni and V. This is due to their deposition on the active surface, blocking of the pore space and zeolite structure destruction. This tends to be a significant problem for refineries [25–27]. Catalyst deactivation by heavy metals significantly influences the economic efficiency of the technology. When the catalyst activity decreases, the yield of the target products also drops down, leading to an increase in the fresh catalyst loading and considerable economic losses. Therefore, at present, forecasting the catalytic performance with due consideration of the catalysts activity change by means of mathematical models is relevant. There are various approaches to catalytic cracking modeling [28–34] mainly by lumps formation with a wide range of boiling point, such as “gas oil”, “gasoline”, “gases” and “coke”. Although this approach allows determining the lump weight fractions, it does not consider the hydrocarbons reactivity within lumps and the group characteristics of vacuum gas oil [35] affecting significantly the coke yield,

Corresponding author. E-mail addresses: [email protected] (G. Nazarova), [email protected] (E. Ivashkina), [email protected] (E. Ivanchina), [email protected] (I. Dolganova).

https://doi.org/10.1016/j.fuproc.2019.106318 Received 20 August 2019; Received in revised form 5 December 2019; Accepted 16 December 2019 0378-3820/ © 2019 Published by Elsevier B.V.

Fuel Processing Technology 200 (2020) 106318

G. Nazarova, et al.

The laboratory studies show that co-precipitation of V and Na is synergistically harmful, while the simultaneous presence of V and Ni on the cracking catalyst gives mutual benefit to catalytic and physicochemical properties of the catalyst. The interaction of Ni and V, as well as the positive effect of nickel is discussed in [51,63–68]. In [53], a reaction network was proposed for cracking of nickel and vanadium porphyrins. According to this network, at the first stage, they are thermally decomposed into metal and porphyrin, then porphyrin is cracked thermally to form primary products and coke, and then catalytic transformations of the primary thermocracking products occur. Moreover, the destructive effect of V on the crystalline structure of zeolite is associated with the formation of fusible eutectics with Na or rare earth elements. In [54], the mechanisms of Ni and V deactivation and the mechanism of Ni inhibition of the deactivating effect of V by Ni are discussed. Experimental studies showed that samples containing V and Ni had higher activity than those with only V, which indicates that Ni reduces the destruction of the catalyst structure caused by V, reducing the degradation of physicochemical properties and dealumination. In accordance with the deactivation mechanism, a decrease in the catalyst activity under the action of V occurs upon the interaction of the catalyst with vanadium acid (Eq. (2)), which is formed due to the contact of vanadium oxide with water vapor (Eq. (1)). The studies performed revealed an interaction between V and Ni, which suggests the formation of a refractory nickel vanadate, which can be best described by the reaction in accordance with Eq. (3).

the catalyst activity, the products yield and composition. In addition, presented hydrocarbons reaction networks mostly have rigid formalization, and do not consider the feedstock composition and the ratio of direct and reverse reactions depending on current concentrations and cracking temperature. Most modern models of catalytic cracking consider the catalyst deactivation by coke [36–40] using a negative exponential function, which allows predicting changes in the catalyst activity depending on the catalyst residence time in reactor (Time-on-Stream Theory) [41] or coke content on the catalyst (Coke-on-Catalyst) [42]. Several studies [43,44] focus on detailed description of chemical conversion for catalytic cracking considering the catalyst composition and forecasting the coke content on the catalyst surface. In the research [45], the authors describe the complexity of the interrelationship between catalyst deactivation by aging, poisoning and pollution, and the effect of these factors on the heat balance and catalyst circulation rate. A catalytic cracking model was developed by authors [39] with due consideration of the instantaneous adsorption theory, diffusion effects, chemical reaction and deactivation theory. A significant step forward in the deactivation modeling is using selective models that consider the different effects of catalyst deactivation on the reactions of the reaction network or reaction group, affecting the different distribution of the products [46]. Herewith, the coke accumulation and the amorphous level of coke depend significantly not only on the feedstock characteristics, process conditions of catalyst exploitation in the “riser–regenerator” cycle, but also on the type and chemical composition of the catalyst (acidity, pore size and zeolite pore structure) [24]. At the same time, the authors do not pay due attention to description of catalyst deactivation under the heavy metals action and accounting for the feedstock composition. Nevertheless, the proposed approach for lumps formation provides the forecasting the coke amount and allows studying how these factors influence the catalyst deactivation. The article [47] describes calculation of the mobil metal index (MMI) with due consideration of the total content of metals. The change in the cracking catalyst activity was estimated by the change in the specific surface considering the fresh catalyst loading into the system, V migration to new particles and the rate of nickel accumulation [48]. At the same time, they do not consider the influence of the composition of raw materials on coke formation. Importantly, that studies by different authors show that the mechanisms of catalyst deactivation by coke, Ni and V are different (active sites poisoning, blocking, selectivity shift, fusible eutectics formation with sodium or rare earth elements, vanadic acid formation, etc.) [49–56]. The most discussed mechanism of the FCC catalyst poisoning by vanadium involves the following reactions:

2NiO + 2Н3 VO4 = Ni2 V2O7 + 3Н2 O

The authors of [68] found that the presence of metals significantly reduces the activity of the catalyst in cyclohexane conversion reactions (from 193 (Cat2) to 107 (VCat2)), while the activity of the catalyst containing Ni and V is higher (113) than the activity of the catalyst containing only Ni or V (120 and 107 respectively). Considering the catalysts deactivation is an important item at the development of an adequate mathematical model. As path forward of work [69], the model was expanded by the catalyst deactivation description. That ensures the model sensitivity to deactivating factors when the forecasting the catalyst activity and its effect on the target products yield considering the feedstock characteristics and process conditions of catalytic cracking. Therefore, the study of thermodynamic and kinetic patterns of catalytic cracking reactions when forecasting the catalyst activity considering the feedstock composition, heavy metals content and the structural and selective properties of the catalyst is topical today. The aim of the work is to develop the mathematical model of catalytic cracking that considers the deactivation patterns of the zeolitecontaining catalysts by coke and heavy metals.

- reaction of V2O5 with steam to form the mobile vanadium acid (H3VO4 or H4V2O7) under regenerator conditions; - subsequent exposure of this acid to the aluminum of the catalyst frame to form a solid solution of aluminum vanadate (AlVO4). This leads to dealumination, structural destruction and loss of the acid sites [54,57–59].

V2O5 + 3Н2 O = 2Н3 VO4

(1)

2Н3 VO4 + Al2O3 = 2AlVO4 + 3Н2 O

(2)

(3)

2. Object and Methods 2.1. Brief description of the catalytic cracking technology The research object is the catalytic cracking unit for the vacuum gasoil refining. The catalyst is synthetic microspheric zeolite-containing catalyst (Y and ZSM-5) with ZSM-5:Y ratio of 0.11 unit and rare-earth elements content of 1.5–2.0 wt%; the bulk density is 700–840 kg/m3. Chemical conversions of the feedstock are performed in the riser at the regenerated catalyst upstream. The cracking temperature is 495–542 °C, the pressure in the reactor settling zone is 0.078 ÷ 0.16 MPa, the temperature and consumption of the feedstock are 240–350 °C and 160–425 m3/h. The slops consumption is supplied to the riser (not > 14 m3/h) in order to stabilize the regenerator coke load (fr. > 420°С with catalyst fines). After the riser, the coked catalyst enters the regenerator, where coke is oxidized from catalyst outer surface and inner pores by air. The temperatures is 630–700 °C in fluidized bed and not > 730 °C in the settling zone and the collecting

The dealumination of the catalyst framework caused by Ni is negligible compared to V, but the main negative effect of Ni deposited on the cracking catalyst is an increase in the dehydrogenation ability of the catalyst and a selectivity shift. Ni acts as a catalyst for the dehydrogenation reaction, thereby promoting coking and shifting the process selectivity [26,50,52,53]. In addition, co-deposition on metal catalysts can have both negative and positive effects [60] depending on the specific deposited metals, the method and amount of deposition [61,62]. 2

Fuel Processing Technology 200 (2020) 106318

G. Nazarova, et al.

Fig. 1. Brief description of the catalytic cracking technology.

4. BET method [74] to measure the surface area of catalyst using BETanalyzer МЕТА Sorbi–М (the range of surface area is 0.01–2000 m2/g, maximum permissible relative error is ± 6%); 5. Thermo-gravimetric analysis [7] to determine the coke structures on coked and regenerated catalysts using thermal analyzer with mass-spectrometer SDT Q600 V20.9 Build 20 (the temperature range is up to 1500 °C, reproducibility ± 2%). 6. The pycnometric method [75] to determine the density of the feedstock, light gas oil, heavy gas oil and gasoline fraction of the catalytic cracking unit (GOST 3900-85). 7. The methods of the structural-group analysis, n-d-m-method and Hazelvud method to determine the structural-group composition of saturated and aromatic hydrocarbons of feedstock and product respectively [76]. The method precision of determining the carbon distribution is 1.5%, the number of rings is 0.1 units. 8. Gas Chromatography according to ASTM D 2887-08 to determine the fractional composition of the catalytic cracking feedstock; 9. Conradson method according to GOST 19932-99 to determine the mass fraction of the coke residue in the range from 0.01 to 30.0% formed after evaporation and pyrolysis of relatively non-volatile petroleum products; 10. X-ray fluorescence analysis methods according to GOST R 519472002 to determine the sulfur content in the catalytic cracking hydrocarbons fractions using the SPECTROSCAN SL (the range of sulfur concentrations is 0.0007–5.0%, lower detection limit of sulfur 0.0005%);

chamber, the pressure is 0.05 ÷ 0.21 MPa. Thus, the catalysts operation becomes cyclical during the “operation-regeneration” cycle (Fig. 1). The cracking products are sent to the rectification unit to be separated into rich gas, gasoline fraction, light gas oil (fr.195–310°С) and heavy gas oil (fr.310–420°С). The yield and the quality of the catalytic cracking products depend on the catalyst performance changing in “riser–regenerator” cycle, in particular, its activity. The formation of coke on the catalyst surface and its deactivation are determined by the feedstock composition, especially, by the contents of saturated and aromatic hydrocarbons, resins, heavy metals, by the process conditions of the catalyst operation, and by their structural and selective properties, etc. To mathematically describe the catalyst deactivation, it is necessary to study how these factors influence the catalyst activity. 2.2. Methods of research In order to investigate the characteristics of the catalytic cracking feedstock, product and catalyst, we used the following methods: 1. The liquid-adsorption chromatography according to VNII NP (AllRussia Research Institute for Oil Processing) methodology using ASKG silica gel (grain size 0.2–0.5 mm) for dividing the heavy fractions of feedstock and product into the resins, saturated and aromatic hydrocarbons [70]. Precisions of test method are ± 5, ± 10, and ± 20% – at the content of the determined component at the ranges of (50 ÷ 70), (15 ÷ 50) and < 15 wt%, respectively). 2. The optical method according to GOST 18995.2-73 to measure the refractive index of hydrocarbons fractions using Atago NAR-3 T refractometer [71,72]. The range of the refractive index is 1.3–1.7 nD. Precision of the refractive index is ± 0.0001 nD, temperature is 5–50°С, and precisions of temperature measurement is ± 0.2 °C). 3. Chromatography-mass spectrometry [73] using a Hewlett Packard 6890 Gas Chromatograph System chromatography-mass spectrometer with a 5973 Mass Selective Detector and GC Chemstation and Trace DSQ software “ThemoElectron” (temperature range is 20–350 °С, unit mass resolution is 1–1050 amu);

3. Results and discussion 3.1. Experimental studding the cracking catalysts Catalyst performance, in particular, the high BET surface area, activity, selectivity, etc. are determined by the composition and structural characteristics of the catalyst. In addition, these indicators and the catalyst service life decrease during the operation (Table 1). The experimental studies of the cracking catalysts have shown that a significant amount of coke is formed on the catalyst surface during its operation due to the catalytic cracking reactions occurred in riser. The 3

Fuel Processing Technology 200 (2020) 106318

G. Nazarova, et al.

Table 1 The major catalyst performance of the cracking catalyst. Parameter

Value

The average grain diameter, mсm. BET surface area, m2/g – fresh catalyst – regenerated catalyst Activity, % – fresh catalyst – regenerated catalyst Mass fraction of carbon, wt% – before regeneration – after regeneration

70–90

Table 2 The range of changes in the major physico-chemical characteristics of catalytic cracking feedstock.

At least 250 138–150 82–84 74–82 0.4–0.61 0.02–0.1

range of the carbon mass fraction on the catalyst before regeneration is 0.4–0.61 wt% and it significantly decreases after regeneration to 0.02–0.1 wt%. As a result, the catalyst activity is restored. Thus, the range of the regenerated catalyst activity is 74.0–82.0% and this value largely determines the composition and quality of the cracking products. The fresh catalyst activity after the stages of preparation and entry into operation is 2–4 units higher. We determined the coke structures on coked and regenerated catalysts by thermo-gravimetric analysis (Fig. 2). The study was carried out up to temperatures typical for the fluidized bed of the regenerator (630–700°С), as at higher temperatures of the zeolite lattice is destroyed (the initial temperature of the zeolite destruction is observed at temperatures of 710–720°С when the rare-earth elements content is 1.5–2.0% [9]). The results of thermogravimetric analysis allow estimating the coke burning temperature from the coked and regenerated catalysts surface and show that the coke has amorphous structure with C/H ratio is 0.2–1.5 unit. This is because the peak temperatures of exothermic process of coke burning are 383.4 °C – for the coked catalyst and 392.3° С - for regenerated catalyst, which is associated with the presence of more consistent residual coke structures on the catalyst surface after regeneration. The maximum heat flows in the temperature range of the second peak are 3.12 W/g (temperature is 285.5 °C) – for coked and 2.54 W/g (temperature is 294.3 °C) – for regenerated catalysts, and they increase along with the coke content. The total weight loss in the temperature range of 0–700° С is 4.0 and 2.0 wt% – for the coked and regenerated catalysts, respectively. Formation of the amorphous coke is typical when the contact time between feedstock and catalyst is short. In this case, the catalyst deactivation is mainly due to acid-site poisoning. Thus, the contact time at the researchable object is short (about 3–5 s), the catalyst deactivation is mainly due to acid-site poisoning than from blocking the input pore

Quality parameters

Value

Density at 20°С, g/sm3 Fractional composition, °С Conradson carbon rate, % Mass fraction of sulfur, wt% Mass fraction of vanadium, ppm Mass fraction of nickel, ppm Saturated hydrocarbons, wt% Aromatic hydrocarbons, wt% Resins, wt%

0.89–0.92 350–570 0.02–0.16 0.019–0.35 0–0.3 0–0.2 58.2–74.6 23.2–38.6 2.1–6.3

and the canals of zeolite (this typically occurs when contact time is high [24].

3.2. Studying the feedstock characteristics of the catalytic cracking The characteristics and composition of the feedstock, including the saturated and aromatic hydrocarbons content, Conradson carbon rate, heavy metals content, etc., significantly affect the catalyst deactivation degree. We found out that the feedstock characteristics by major quality parameters change considerably (Table 2). The vacuum gas oil is characterized by a broad fractional composition of 350–570 °C with density of 0.88–0.92 g/cm3 (Table 2). Despite the high end boiling point of the fraction, the feedstock meets the requirements for coking ability (not > 0.3%). However, the elements that determined the Conradson carbon rate (resins, asphaltenes, polycyclic aromatic hydrocarbons) are almost completely converted to coke under catalytic cracking conditions and deactivate catalyst. The content of heavy metals in the catalytic cracking feedstock, causing the deactivating effect on the cracking catalyst, is up to 0.2 and 0.3 ppm for Ni and V, respectively. The feedstock group composition changes significantly, and this influences considerably the coke formation on the catalyst, temperature mode, the yield and quality of products. When the content of aromatic hydrocarbons and resins (up to 38.6 and 6.3 wt%, respectively) in the feedstock is high, the coke content on the catalyst surface and within the pores rises. This leads to an unbalanced temperature mode in the “riser-regenerator” system and reduces its activity and the feedstock conversion degree. To estimate the characteristics of saturated and aromatic hydrocarbons of vacuum distillate, we used the methods of determining the structural-group composition of the petroleum fractions (n-d-m-method and Hazelwood method) and method of chromatography-mass

Fig. 2. The weight and heat flow change of the coked (a) regenerated (b) catalyst during high-temperature oxidation of coke. 4

Fuel Processing Technology 200 (2020) 106318

G. Nazarova, et al.

Table 3 Hydrocarbon groups characteristics of catalytic cracking feedstock. Hydrocarbons group

Characteristics

Paraffins Naphthenes Aromatics

С14–С40+ Mono- and bicyclic naphthenes with long substituents С1 – С25 (the average number of naphthenic rings is 1.7–2.3 units) Mono- and poly- structures with long substituents (the average number of aromatic rings is 1.6–2.8 units, the average number of naphthenic rings is 1.3–1.5 units) High molecular polycyclic naphthenic-aromatic compounds

Resins

deactivation accounting (ψ) by coke and heavy metals at modeling the catalytic cracking.

spectrometry (Table 3). We used the experimental results to develop the reaction network. The discovered by chromatography-mass spectrometry structure of hydrocarbons in the catalytic cracking feedstock were used to analyze the thermodynamic probability of reactions using the quantum chemistry methods [77,78].

3.3.1. Mathematical modeling of reversible deactivation of the catalytic cracking catalyst by coke For the mathematical description of the reversible deactivation of the catalyst due to coke formation, an approach based on considering the change in the catalyst activity depending on the coke content on its surface (Coke-on-Catalyst model) was used, the parameters of the following functional dependence were established:

3.3. The model The content of coke and hydrocarbons groups is calculated by solving of the differential equations according to developed kinetic scheme [69]. The reactions reversibility is checked and corrected at each calculation step by the current chemical attraction values [76]:

A = A0 ⋅exp(−kd⋅Ck )

where А – current catalyst activity; A0 – catalyst activity with a minimum coke content in the cycle (regenerated catalyst); kd – deactivation constant. Predicting the amount of coke in Eq. (6) is carried out by solving the system of differential Eq. (4) in accordance with the hydrocarbon conversion mechanism [69], considering the hydrocarbon composition of the processed feed and the reactions leading to coke formation (k13–k18). Thus, the current activity of the catalyst is determined depending on the initial activity of the regenerated catalyst, coke concentration and deactivation constant. At the same time, the deactivation constants differ, depending on the composition of the catalysts, including the type of zeolites used, affecting the reaction rates. Considering the fact that the catalyst includes two zeolites with different structural and acidic characteristics, a selective approach is implemented to quantitatively evaluate the effect of catalyst deactivation on the reaction rates catalyzed by Y and ZSM-5 zeolites. This approach increases the model parameterization as the deactivation constants for the reactions catalyzed by zeolites ZSM-5 and Y are different. The deactivation constants in Eqs. (7), (8) were determined considering the acid properties of zeolites. It was shown [24,80–82] that the rate of bimolecular reactions promoting coke formation, including hydrogen transfer, extremely depends on the density of acid sites. Also the density of acid sites is determined by the synthesis conditions, including the ion exchange conditions and dealumination zeolite framework, etc. The authors [80] observed close values of coking/cracking ratios for H-ZSM-5 and dealuminated HY zeolite having similar densities of strong acid centers. In [83], it was shown that zeolite H-ZSM-5 is characterized by stronger acidic properties, the density of acid sites in HZSM-5 zeolite is 1.48 times higher than that of the HeY zeolite (Table 4). The calculation of hydrocarbon groups was performed with considering the catalyst deactivation by coke at each calculation step for reactions catalyzed by Y and ZSM-5 zeolites. This is discovered and described by an exponential dependence that is determined using the experimental data considering the acidic and structural properties of zeolites catalyst (Eqs. (7), (8)):

20

⎧ ⎪ dCi/ dτ = ∑ (± ψ⋅(Wj − W−j )) ⎪ j=1 20 ⎨ ⎪ dT / dτ = ∑ (± ψ⋅((Δr HT° )j⋅Wj ) − ((Δr HT° )−j⋅W−j ))/ρf cf ⎪ j=1 ⎩

(4)

Initial conditions τ = 0, Ci = Ci0 , Т0 = Ti.t Here Сi – concentration of i-th hydrocarbons group, mol/m3; τ – residence time, s; j – the number of reaction; ψ – parameter of catalyst deactivation by coke, nickel and vanadium: ψ = AZSM – for reactions catalyzed by ZSM-5, ψ = AY – for reactions catalyzed by Y, ψ = AH2 – for dehydrogenation, aromatization, condensation and coke formation reactions; Wj – rate of j-th reaction in forward direction, mol/(s·m3); W-j – rate of j-th reaction in inverse direction mol/(s·m3); Т – flow temperature, К; ρf – density of the flow, kg/m3; cf – heat capacity of flow, J/ kgK; (ΔrHoT)j – heat effect of the direct chemical reaction, J/mol; (ΔrHoT)-j – heat effect of the inverse chemical reaction, J/mol. The initial temperature of flow (Tit) is the thermal equilibrium temperature between feedstock and the catalyst (the initial reaction temperature). It is determined with due consideration of the flowrate, temperature and heat capacity of the feedstock and the catalyst according to the Eq. (5)

Gcat ccat (Tcat − (Tit + 1)) = Gf cf (Tit + Tf )

(6)

(5)

Here Gcat – catalyst flowrate, kg/s; ccat – heat capacity of the catalyst, J/kgK; Tcat – catalyst temperature after regeneration, K; Тit – thermal equilibrium temperature between feedstock and catalyst, K; Gf – feedstock flowrate, kg/s; cf – heat capacity of the flow, J/kgK; Tf – feedstock temperature, K. Thus, in order to describe the deactivation of the cracking catalyst by coke and heavy metals, it is important to comprehensively consider the following factors: – the regenerated catalyst activity; – the feedstock group composition and characteristics, including the content of heavy metals in the feedstock; – thermodynamics and kinetics of catalytic cracking reactions, including reactions caused to the coke formation considering the structural and selective properties of the catalysts; – technological modes of catalysts operation.

AY = A0 ⋅exp( −0.077⋅Ck )

(7)

AZSM = A0 ⋅exp(−0.113⋅Ck )

(8)

Here 0.113 and 0.077 are experimentally determined deactivation constants; АY – current relative catalyst activity to the reactions catalyzed by Y zeolite, %; AZSM – current relative catalyst activity to the

Fig. 3 shows the scheme for the reversible and irreversible catalyst 5

Fuel Processing Technology 200 (2020) 106318

G. Nazarova, et al.

Fig. 3. Scheme for reversible and irreversible catalyst deactivation.

contact of the catalyst elements with H3VO4 (Eq. (2)) that is formed during the contact of the vanadium oxide with water stream (Eq. (1)). According to overview, Ni co-presence with V reduces the V destructive effect, as a result the zeolite framework dealumination is decreased, the catalyst surface area, acidic sites and catalyst activity are keeping. All listed above is considered when describing the catalyst deactivation (Eqs. (1)–(3)). Also the formation of the metal oxides occurs in the air during the regeneration from the Ni and V metalorganic compounds adsorbed on the catalyst.

Table 4 Data on thermoprogrammed desorption of ammonia on zeolites [83]. Zeolite

Y ZSM-5

Amount of desorbed ammonia (acidity), mmol/g (%) Strong

Medium

Weak

– 0.61 (47)

0.56 (64) –

0.2 (36) 0.69 (53)

reactions catalyzed by ZSM-5 zeolite, %; A0 – activity of regenerated catalyst taking into account the effects of vanadium by Ni co-presence with V, %;, Сk – coke content on the catalyst wt%. Moreover, the nickel effect shifting the catalyst selectivity with the increase in the catalyst dehydrogenation ability [84] is considered while calculating the hydrocarbons groups concentrations: The change in the dehydrogenation activity of the catalyst (AH2) was determined based on the established experimental regularity of the hydrogen changing in the cracking gas from the nickel content in the process feed relative to the dehydrogenation ability of the catalyst at zero nickel concentration:

AH2 = e1,1554·СNi

2Ni + O2 = 2NiО 4 V + 5O2 = 2V2O5

(10)

It is assumed that the Ni and V contained in the feedstock are deposited on the catalyst completely. In order to demonstrate the combined effect of Ni and V on irreversible catalyst deactivation due to dealumination, predictive calculations were performed using the model in accordance with Eqs. (6)–(9). Predictive calculations were carried out for different contents of V and Ni in the cracking feedstock - 0.3 and 0.0 ppm and 0.3 and 0.2 ppm respectively, and only the volume of processed feedstock changed from 6500 tons to 2.4 million tons taking into account the amount of catalyst located in the reactor-regenerative unit system. Such calculations show how Ni and V affect the irreversible deactivation of the catalyst. At this stage, irreversible deactivation due to the deposition of graphitized coke on the surface of the catalyst is not considered and is planned for the next stages of the study. The patterns of the catalyst activity decreasing depending on the heavy metals content under the other equal conditions are visually shown in Fig. 4. The performed calculations indicate that the catalyst activity decreases by 0.68% due to dealumination if the processed feedstock volume is 2.4 mln.tones and contains 0.3 ppm of V content. In addition, the V destructive effect decreases simultaneously with an increase of the Ni content in the catalytic cracking feedstock by 0.2 ppm. As a result, the regenerated catalyst activity is restored by 0.4%. Thus, the catalyst activity with Ni and V co-presence is higher (79.1–78.87%) than that of the catalyst containing only V (79.1–78.56%). This is due to the mechanism of Ni inhibition of the destructive action of V (Eqs. (6)–(9)) [54]. In accordance with the deactivation mechanism, a decrease in the catalyst activity under the action of V occurs upon the interaction of the catalyst with vanadium

(9)

where СNi – the nickel content in the feedstock, ppm. The revealed dependence characterizes the intensity of hydrogen formation in dehydrogenation, aromatization, condensation and coke formation reactions. This indicates the enrichment of cracking products by polycyclic aromatic hydrocarbons and coke. Thus, with increasing the nickel content up to 0.2 ppm, the dehydrogenation ability of the catalyst increases by 1.23 times, the coke content on the catalyst increases from 0.44 to 0.53 wt%. 3.3.2. Mathematical description of the combined effect of nickel and vanadium on irreversible catalyst deactivation The regenerated catalyst activity significantly impacts the feedstock conversion level, the products yield and composition. To describe the irreversible deactivation and calculate the regenerated catalyst activity, the combined effect of Ni co-presence with V in catalytic cracking feedstock is to be considered. According to [54], the catalyst activity reducing by vanadium is caused due to dealumination occurring at the 6

Fuel Processing Technology 200 (2020) 106318

G. Nazarova, et al.

change. A new mathematical model of the catalytic cracking process comprehensively considers factors that significantly affect the reversible and irreversible deactivation of the catalyst: the regenerated catalyst activity, the feedstock group composition, the content of heavy metals in the feedstock, thermodynamics and kinetics of catalytic cracking reactions, including reactions caused to the coke formation considering the properties of zeolites, technological modes of catalysts operation. A comprehensive account of the thermodynamic and kinetic laws of the process, as well as the deactivation of the catalysts, provides the universality and high adequacy of the mathematical description without the need for calibration of kinetic parameters. The mathematical model considering the stated above has he following assumptions: 1) the chemical conversions are carried out in accordance with the formalized mechanism; the reactions reversibility is corrected at each calculation step by reactions chemical attraction values; if the reaction chemical attraction is much greater than the RT value, the reaction is assumed as non-reversible [79]; 2) the statement about the kinetic non-reversibility is equivalent to the statement that the inverse reaction rate is negligible compared to the forward reaction rate (it is assumed that k-j = 0 in kinetic equations); 3) there is no heat exchange with the environment; 4) the thermal equilibrium occurs before the reaction starts; 5) nickel and vanadium contained in the feedstock are deposited on the catalyst volume completely; 6) the reactor mode is plug-flow; 7) the kinetic model is formalized and quasi-homogeneous.

Fig. 4. The catalyst activity depending on content of Ni and V with increasing of the feedstock volume.

acid (2), which is formed due to the contact of vanadium oxide with water vapor (1). Ni, by chemical interaction with vanadium acid (3), partially inhibits the destructive effect of V, which leads to a decrease in dealumination of the zeolite framework, preservation of the catalyst surface area, acid sites and activity of the catalyst. However, it is important to consider the fact that the catalyst dehydrogenation ability rises along with the Ni content. The degree of catalyst recovery by 0.4% relative to the initial activity of the catalyst provides an increase in the consumption of the target component of the gasoline fraction by 2.62 tons/day. Complex considering of the items above is important from the point of forecasting the coke amount formed on the catalyst and the catalyst deactivation degree by the coke and heavy metals. This will ensure the continuous optimization of new and existing catalytic cracking units depending on the feedstock composition and the catalyst activity

3.3.3. Verification of the models without and taking into account the catalyst deactivation To verify the models, we compared the calculated and experimental data on the products yield, on and coke content on the catalytic

Fig. 5. Verification of the models without and taking into account the catalyst deactivation. 7

Fuel Processing Technology 200 (2020) 106318

G. Nazarova, et al.

Fig. 6. Coke content depending on content of resin (a) and nickel (b) in feedstock (model calculation).

cracking catalyst, and on the PPF and BBF in the rich gas (Fig. 5). Due to considering the catalyst deactivation, the average relative calculation error for the model was reduced by 11.3%. The average relative error of the calculations does not exceed 2.58%, thus, the model is adequate and can be used to forecast and to optimize the catalytic cracking unit. 3.4. Application of mathematical model Using the mathematical model of the catalytic cracking process allows us to estimate how the process conditions and the feedstock composition influence the catalyst activity, products yield and composition. The catalyst activity change affects the yield and composition of cracking products and strongly depends on the coke content on the catalyst surface and on the catalyst dealumination. These factors are determined by the content of the aromatic hydrocarbons, resins, heavy metals. Figs. 6, 7 show the effect of the feedstock composition, specifically the content of resins and nickel in the catalytic cracking feedstock, on the coke content on the catalyst and on its activity. Thus, when the resins content in the feedstock increases from 2.1% to 6.3%, the coke content on the catalyst increases from 0.52 to 1.27% wt. This is equal to 3.1–7.0 wt% of coke per feedstock under the equal process conditions and activity of regenerated catalyst. An increase of the catalyst dehydrogenation ability also leads to the enrichment of the cracking products by polycyclic aromatic hydrocarbons and coke. This is due to intensification of reactions with hydrogen formation (dehydrogenation, aromatization, condensation and coke formation). Consequently, the relative catalyst activity decreases. With an increase of the Ni content in the cracking raw material to 0.6 ppm [85] at a constant content of V (0.7 ppm), the coke content on the catalyst increases from 0.77 to 1.09 wt% (Fig. 7). This decreases the catalyst activity and target products yield. The coke content increasing on the catalyst from 0.55 to 1.27% wt.

Fig. 8. Regenerated catalyst activity depending on the content of Ni and V (model calculation).

reduces the catalyst activity from 75.8 to 71.8% – for reactions catalyzed by Y zeolite, and from 74.3 to 68.6% – for reactions catalyzed by ZSM-5 zeolite. This equal to 9.23 and 13.27% of the regenerated catalyst activity (Fig. 8). Using the mathematical model of catalytic cracking allows us to study how the Ni and V content influence the regenerated catalyst activity considering the feedstock volume (annual volume is 2.4 million tons). In accordance with [83], the range of the heavy metals content in a vacuum distillate can vary significantly: up to 1.9 ppm – for vanadium and up to 0.6 ppm – for nickel. The forecast calculations showed that the catalyst activity decreases by 4.4% (from 79.1 to 75.62%) with increasing of the vanadium concentration in the feedstock by 1.9 ppm due to its dealumination (Fig. 8). When the Ni with V co-presence in the catalytic cracking feedstock and the metals are fully adsorbed on the catalyst, the V destructive effect reduces by 1.2% with increasing the Ni content in the feedstock up to 0.6 ppm. This is due to the Ni contact with the vanadic acid. Thus, the catalyst activity containing both Ni and V is higher (79.1–76.57%) than that of the catalyst containing only V (79.1–75.6%). The catalyst activity decreasing along the reactor height due to coke accumulation affects the ratio and quality of catalytic cracking products (Fig. 9–11). According to the calculations performed, the yield of the gasoline fraction changes by 4.43 wt%, depending on the feedstock composition (CSH/CAH = 1.6–1.8 units), other things being equal. The calculations show that if the feedstock contains a lot of high molecular paraffins and naphthenes (CSH/CAH = 2.4–3.2 units), the yield of gasoline with 90.0–91.5 octane number by the research method is 58.1–59.58 wt%. If the feedstock contains a lot of aromatic hydrocarbons (CSH/CAH = 1.6–1.8 units), the gas yield is 17.0–17.7 wt% and it is higher than that for the feedstock with a high content of aromatic hydrocarbons (CSH/CAH = 1.6–1.8 units) – 15.84 and 16.63%. The reduction of the gasoline yield by 0.34% for the feedstock with CSH/CAH of 2.7 units relative to the feedstock with CSH/CAH of

Fig. 7. Relative activity for reactions catalyzed by Y and ZSM-5 zolites depending on the coke content (model calculation). 8

Fuel Processing Technology 200 (2020) 106318

G. Nazarova, et al.

Fig. 9. The yield (a) and the octane number (b) of gasoline fraction depending on the feedstock composition (model calculations).

concentration of deactivating substances (coke and heavy metals (it was assumed that the metals contained are deposited on the catalyst completely). The forecast calculations showed that when the resins content in the feedstock increases by 4.2 wt% (from 2.1 to 6.3 wt%), the coke content on the catalyst increases by 0.75 wt% (from 0.52 to 1.27 wt%) under the other equal conditions. The catalyst activity decreases by 4.4% (75.62%) relative to initial value (79.1) with increasing the V content in the feedstock by 1.9 ppm due to its dealumination. If the Ni with V co-presence in the catalytic cracking feedstock and the Ni content increases by 0.6 ppm, the V destructive effect reduces by 1.2% (from 75.62 to 76.57) due to reaction of Ni with the vanadic acid. Also it leads to increase the catalyst dehydrogenation activity, when the Ni content in the feedstock increases by 0.6 ppm, the coke content on the catalyst increases by 0.32 wt% (from 0.77 to 1.09 wt%) under the other constant. The model application allows us to predict the yield of catalytic cracking products, the gasoline octane number, the concentration of propane-propylene and butane-butylene fractions of fat gas, and the coke content on the catalyst at its continuous activity change. The average relative error of the calculations is reduced by 11.3% due to introduction the catalyst deactivation laws.

2.4 units is connected with the catalyst deactivation due to the condensation reactions. As a result, for this feedstock, the coke yield is higher (4.52% wt.) than for the feedstock with CSH/CAH of 2.4 units (3.65 wt%). For high concentration (35.8 and 33.9 wt%) of aromatic hydrocarbons in the feedstock (CSH/CAH = 1.6–1.8 units), the yield of gasoline is lower (55.15 and 57.0 wt%). The gasoline, has a high octane points, compared to other types of raw materials. In addition, when processing such raw materials, the yield of light (12.4 and 12.2%), heavy (9.78 and 9.4%) gas oil is higher than octane number (93.6 and 92.9) when processing the vacuum distillate with a high content of saturated hydrocarbons (8.4–8.9 wt%). The calculations indicate that the feedstock composition and the regenerated catalyst activity affect the distribution of the catalytic cracking products of and coke. At the same time, it is possible to increase the consumption and octane characteristics of the catalytic cracking gasoline fraction by adjusting the process conditions considering these factors. Along with this, the cracking temperature is one of the main parameters affecting the yield and composition of the cracking products, as well as the degree of the catalyst deactivation.

4. Conclusions CRediT authorship contribution statement

The developed mathematical model of the catalytic cracking considers the thermodynamics and kinetics of the process reactions, including those resulting in coke formation according to the developed reaction network. The model also describes the influence of heavy metals on the catalyst activity. This allows us to predict the catalyst activity depending on the concentration of heavy metals in the feedstock, on the raw material composition, on the intensity of coke formation when the catalyst dehydrogenating ability changes, and on the structural and selective catalyst properties. We have established the laws governing the cracking catalyst deactivation in an up-flow lift reactor depending on the nature and

Galina Nazarova: Project administration, Methodology, Conceptualization, Software, Writing - original draft, Visualization, Data curation, Resources, Investigation. Elena Ivashkina: Supervision, Methodology, Conceptualization, Writing - review & editing, Data curation. Emiliya Ivanchina: Methodology, Writing - review & editing. Alexandra Oreshina: Resources, Validation. Irena Dolganova: Funding acquisition. Mariya Pasyukova: Formal analysis.

Fig. 10. Rich gas and coke yields depending on the feedstock composition (model calculations). 9

Fuel Processing Technology 200 (2020) 106318

G. Nazarova, et al.

Fig. 11. Light and heavy gasoil yields depending on the feedstock composition (model calculations).

Acknowledgements The research was supported by Tomsk Polytechnic University within the framework of Tomsk Polytechnic University Competitiveness Enhancement Program and Russian State Project “Science” 10.13268.2018/8.9.

[18]

[19]

References [20] [1] A.P. Glotov, S.V. Kardashev, S.V. Egazar’yants, S.V. Lysenko, V.A. Vinokurov, E.A. Karakhanov, Catalytic cracking of petroleum feedstock in the presence of additives derived from cross–linked mesoporous oxides for reduction of the sulfur content in liquid products, Chem. Technol. Fuels Oils 52 (2016) 171–174, https:// doi.org/10.1007/s10553-016-0687-0. [2] P. Bai, M. Xie, U.J. Etim, W. Xing, P. Wu, Y. Zhang, B. Liu, Y. Wang, K. Qiao, Z. Yan, Zeolite Y mother liquor modified γ-Al2O3 with enhanced brönsted acidity as active matrix to improve the performance of fluid catalytic cracking catalyst, Ind. Eng. Chem. Res. 57 (5) (2018) 1389–1398, https://doi.org/10.1021/acs.iecr.7b04243. [3] V. Blay, B. Louis, R. Miravalles, T. Yokoi, K.A. Peccatiello, M. Clough, B. Yilmaz, Engineering zeolites for catalytic cracking to light olefins, ACS Catal. 7 (10) (2017) 6542–6566, https://doi.org/10.1021/acscatal.7b02011. [4] A. Akah, Application of rare earths in fluid catalytic cracking: a review, J. Rare Earths 35 (2017) 941–956, https://doi.org/10.1016/S1002-0721(17)60998-0. [5] Y. Ji, H. Yang, W. Yan, Strategies to enhance the catalytic performance of ZSM-5 zeolite in hydrocarbon cracking: a review, Catalysts 7 (12) (2017) 1–31, https:// doi.org/10.3390/catal7120367. [6] A. Usman, M.A.B. Siddiqui, A. Hussain, A. Aitani, S. Al-Khattaf, Catalytic cracking of crude oil to light olefins and naphtha: experimental and kinetic modeling, Chem. Eng. Res. Des. 120 (2017) 121–137, https://doi.org/10.1016/j.cherd.2017.01.027. [7] S. Oruji, R. Khoshbin, R. Karimzadeh, Preparation of hierarchical structure of Y zeolite with ultrasonic-assisted alkaline treatment method used in catalytic cracking of middle distillate cut: the effect of irradiation time, Fuel Process. Technol. 176 (2018) 283–295, https://doi.org/10.1016/j.fuproc.2018.03.035. [8] E.O. Altynkovich, O.V. Potapenko, T.P. Sorokina, V.P. Doronin, T.I. Gulyaeva, V.P. Talzi, Butane–butylene fraction cracking over modified ZSM-5 zeolite, Pet. Chem. 57 (2017) 215–221, https://doi.org/10.1134/S0965544117010030. [9] V.P. Doronin, T.P. Sorokina, P.V. Lipin, O.V. Potapenko, N.V. Korotkova, V.I. Gordenko, Development and introduction of zeolite containing catalysts for cracking with controlled contents of rare earth elements, Catal. Ind. 7 (1) (2015) 12–16, https://doi.org/10.1134/S2070050415010043. [10] I.M. Gerzeliev, R.M. Arslanov, V.M. Kapustin, G.N. Bondarenko, S.N. Khadzhiev, Catalytic cracking of hydrotreated vacuum gas oil over a zeolite-containing catalyst modified with nickel and cobalt nanoparticles, Pet. Chem. 56 (2016) 51–55, https://doi.org/10.1134/S0965544116010023. [11] V. Abbasov, T. Mammadova, N. Andrushenko, N. Hasankhanova, Y. Lvov, E. Abdullayev, Halloysite-Y-zeolite blends as novel mesoporous catalysts for the cracking of waste vegetable oils with vacuum gasoil, Fuel 117 (2014) 552–555, https://doi.org/10.1016/j.fuel.2013.09.013. [12] T. Kaminski, M.M. Husein, Kinetic modelling of thermal cracking of Arabian atmospheric and vacuum residue, Fuel Process. Technol. 15 (2019) 89–97, https:// doi.org/10.1016/j.fuproc.2019.03.007. [13] O.P. Taran, A.N. Zagoruiko, S.A. Yashnik, A.B. Ayusheev, A.V. Pestunov, I.P. Prosvirin, R.V. Prihod’Ko, V.V. Goncharuk, V.N. Parmon, Wet peroxide oxidation of phenol over carbon/zeolite catalysts. Kinetics and diffusion study in batch and flow reactors, J. Environ. Chem. Eng. 6 (2) (2018) 2551–2560, https://doi.org/ 10.1016/j.jece.2018.03.017. [14] P.B. Venuto, E.T.J. Habib, M. Dekker, Fluid Catalytic Cracking with Zeolite Catalysts, Inc., New York (1979), p. 176. [15] B. Delmon, G.F. Froment, Catalyst Deactivation, (1999). [16] C.H. Bartholomew, G.A. Fuentes, Industrial evaluation of selective hydrogenation catalyst poisoning, Catal. Deactiv. 1 (1997) 447–454, https://doi.org/10.1016/ S0167-2991(97)80187-5. [17] A. Corma, P.J. Miguel, A.V. Orchillés, Influence of hydrocarbon chain length and

[21]

[22]

[23]

[24] [25]

[26]

[27]

[28]

[29]

[30]

[31] [32]

[33]

[34]

[35]

10

zeolite structure on the catalyst activity and deactivation for n-alkanes cracking, Appl. Catal. A Gen. 117 (1994) 29–40, https://doi.org/10.1016/0926-860X(94) 80156-8. G. Caeiro, P. Magnoux, P. Ayrault, J.M. Lopes, F.R. Ribeiro, Deactivating effect of coke and basic nitrogen compounds during the methylcyclohexane transformation over H-MFI zeolite, Chem. Eng. J. 120 (2006) 43–54, https://doi.org/10.1016/j.cej. 2006.03.036. A.A. Lappas, L. Nalbandian, D.K. Iatridis, S.S. Voutetakis, I.A. Vasalos, Effect of metals poisoning on FCC products yields: Studies in an FCC short contact time pilot plant unit, Catal. Today 65 (2001) 233–240, https://doi.org/10.1016/S09205861(00)00588-5. M.C. Silaghi, C. Chizallet, J. Sauer, P. Raybaud, Dealumination mechanisms of zeolites and extra-framework aluminum confinement, J. Catal. 339 (2016) 242–255, https://doi.org/10.1016/j.jcat.2016.04.021. L.R. Aramburo, L. Karwacki, P. Cubillas, S. Asahina, D.A.M. De Winter, M.R. Drury, I.L.C. Buurmans, E. Stavitski, D. Mores, M. Daturi, P. Bazin, P. Dumas, F. ThibaultStarzyk, J.A. Post, M.W. Anderson, O. Terasaki, B.M. Weckhuysen, The porosity, acidity, and reactivity of dealuminated zeolite ZSM-5 at the single particle level: the influence of the zeolite architecture, Chem. Eur. J. 17 (2011) 13773–13781, https://doi.org/10.1002/chem.201101361. A. Ishihara, Preparation and reactivity of hierarchical catalysts in catalytic cracking, Fuel Process. Technol. 194 (2019) 106116, , https://doi.org/10.1016/j.fuproc. 2019.05.039. E. Rodríguez, G. Elordi, J. Valecillos, S. Izaddoust, J. Bilbao, J.M. Arandes, P. Castaño, Coke deposition and product distribution in the co-cracking of waste polyolefin derived streams and vacuum gas oil under FCC unit conditions, Fuel Process. Technol. 192 (2019) 130–139, https://doi.org/10.1016/j.fuproc.2019.04. 012. C.H. Bartholomew, Mechanisms of catalyst deactivation, Appl. Catal. A Gen. 212 (2001) 17–60, https://doi.org/10.1016/S0926-860X(00)00843-7. A.S. Escobar, M.M. Pereira, H.S. Cerqueira, Effect of iron and calcium over USY coke formation, Appl. Catal. A Gen. 339 (2008) 61–67, https://doi.org/10.1016/j. apcata.2008.01.008. R.N. Magomedov, A.Z. Popova, T.A. Maryutina, K.M. Kadiev, S.N. Khadzhiev, Current status and prospects of demetallization of heavy petroleum feedstock (Review), Pet. Chem. 55 (2015) 423–443, https://doi.org/10.1134/ S0965544115060092. A. Corma, J.M. Serra, Heterogeneous combinatorial catalysis applied to oil refining, petrochemistry and fine chemistry, Catal. Today (2005) 3–11, https://doi.org/10. 1016/j.cattod.2005.07.117. Q. Yang, A.S. Berrouk, Y. Du, H. Zhao, C. Yang, M.A. Rakib, A. Mohamed, A. Taher, CFD investigation of hydrodynamics, heat transfer and cracking reactions in a largescale fluidized catalytic cracking riser, Appl. Math. Model. 40 (2016) 9378–9397, https://doi.org/10.1016/j.apm.2016.06.016. Y. Li, J. Chu, J. Zhang, A new generic reaction for modeling fluid catalytic cracking risers, Chin. J. Chem. Eng. 25 (2017) 1449–1460, https://doi.org/10.1016/j.cjche. 2017.01.006. E. Baudrez, G.J. Heynderickx, G.B. Marin, Steady-state simulation of Fluid Catalytic cracking riser reactors using a decoupled solution method with feedback of the cracking reactions on the flow, Chem. Eng. Res. Des. 88 (2010) 290–303, https:// doi.org/10.1016/j.cherd.2009.05.003. K.K. Dagde, Y.T. Puyate, Six-lump kinetic modelling of adiabatic plug-flow riser reactor in an industrial FCC unit, Int. J. Eng. Res. Appl. 2 (2012) 557–568. K. Xiong, C. Lu, Z. Wang, X. Gao, Kinetic study of catalytic cracking of heavy oil over an in-situ crystallized FCC catalyst, Fuel 142 (2015) 65–72, https://doi.org/ 10.1016/j.fuel.2014.10.072. X. Dupain, E.D. Gamas, R. Madon, C.P. Kelkar, M. Makkee, J.A. Moulijn, Aromatic gas oil cracking under realistic FCC conditions in a microriser reactor, Fuel 82 (2003) 1559–1569, https://doi.org/10.1016/S0016-2361(03)00077-2. Y.M. John, R. Patel, I.M. Mujtaba, Modelling and simulation of an industrial riser in fluid catalytic cracking process, Comput. Chem. Eng. 106 (2017) 730–743, https:// doi.org/10.1016/j.compchemeng.2017.01.013. D.S. Stratiev, I.K. Shishkova, I.M. Nikolaychuk, E. Sharafutdinov, I.S. Chavdarov, R. Nikolova, M. Mitkova, D. Yordanov, N. Rudnev, Impact of feed properties on gasoline olefin content in the fluid catalytic cracking, Pet. Sci. Technol. 34 (2016) 652–658, https://doi.org/10.1080/10916466.2016.1156127.

Fuel Processing Technology 200 (2020) 106318

G. Nazarova, et al.

[36] M. Sedighi, K. Keyvanloo, J. Towfighi, Kinetic study of steam catalytic cracking of naphtha on a Fe/ZSM-5 catalyst, Fuel 109 (2013) 432–438, https://doi.org/10. 1016/j.fuel.2013.02.020. [37] M.T. Shah, R.P. Utikar, V.K. Pareek, G.M. Evans, J.B. Joshi, Computational fluid dynamic modelling of FCC riser: a review, Chem. Eng. Res. Des. 111 (2016) 403–448, https://doi.org/10.1016/j.cherd.2016.04.017. [38] G.F. Froment, On fundamental kinetic equations for chemical reactions and processes, Curr. Opin. Chem. Eng. 5 (2014) 1–6, https://doi.org/10.1016/j.coche. 2014.02.002. [39] F. Lopez-Isunza, Deactivation of fluid catalytic cracking catalysts: modelling approach, Stud. Surf. Sci. Catal. 134 (2001) 187–199. [40] P. Varshney, D. Kunzru, S.K. Gupta, Modelling of the riser reactor in a resid fluidised-bed catalytic cracking unit using a multigrain model for an active matrixzeolite catalyst, Indian Chem. Eng. 57 (2015) 115–135, https://doi.org/10.1080/ 00194506.2014.975754. [41] I. Elizalde, J. Ancheyta, Modeling catalyst deactivation during hydrocracking of atmospheric residue by using the continuous kinetic lumping model, Fuel Process. Technol. 123 (2014) 114–121, https://doi.org/10.1016/j.fuproc.2014.02.006. [42] T.M. Moustafa, J. Corella, G.F. Froment, Kinetic modeling of coke formation and deactivation in the catalytic cracking of vacuum gas oil, Ind. Eng. Chem. Res. 42 (2003) 14–25, https://doi.org/10.1021/ie0204538. [43] N.V. Dewachtere, G.F. Froment, I. Vasalos, N. Markatos, N. Skandalis, Advanced modeling of riser-type catalytic cracking reactors, Appl. Therm. Eng. 17 (1997) 837–844, https://doi.org/10.1002/aic.690390610. [44] J.L. Fernandes, L.H. Domingues, C.I.C. Pinheiro, N.M.C. Oliveira, F.R. Ribeiro, Influence of different catalyst deactivation models in a validated simulator of an industrial UOP FCC unit with high-efficiency regenerator, Fuel 97 (2012) 97–108, https://doi.org/10.1016/j.fuel.2012.03.009. [45] P. O’Connor, A.C. Pouwels, FCC catalyst deactivation: a review and directions for further research, Stud. Surf. Sci. Catal. 88 (1994) 129–144, https://doi.org/10. 1016/S0167-2991(08)62734-2. [46] J. Corella, On the Modeling of the Kinetics of the Selective Deactivation of Catalysts. Application to the Fluidized Catalytic cracking Process, Ind. Eng. Chem. Res. 43 (2004) 4080–4086, https://doi.org/10.1021/ie040033d. [47] A.T. Jarullah, N.A. Awad, I.M. Mujtaba, Optimal design and operation of an industrial fluidized catalytic cracking reactor, Fuel 206 (2017) 657–674, https://doi. org/10.1016/j.fuel.2017.05.092. [48] W.R. Gilbert, Mathematicalmodel of FCC catalyst deactivation, Chem. Eng. Commun. 190 (11) (2003) 1485–1498, https://doi.org/10.1080/714909158. [49] E. Rautiainen, R. Pimenta, M. Ludvig, C. Pouwels, Deactivation of ZSM-5 additives in laboratory for realistic testing, Catal. Today 140 (2009) 179–186, https://doi. org/10.1016/j.cattod.2008.10.010. [50] U.J. Etim, B. Xu, R. Ullah, Z. Yan, Effect of vanadium contamination on the framework and micropore structure of ultra stable Y-zeolite, J. Colloid Interface Sci. 463 (2016) 188–198, https://doi.org/10.1016/j.jcis.2015.10.049. [51] U.J. Etim, P. Bai, X. Liu, F. Subhan, R. Ullah, Z. Yan, Vanadium and nickel deposition on FCC catalyst: influence of residual catalyst acidity on catalytic products, Microporous Mesoporous Mater. 273 (2019) 276–285, https://doi.org/10.1016/j. micromeso.2018.07.011. [52] R.E. Roncolatto, Y.L. Lam, Effect of vanadium on the deactivation of FCC catalysts, Braz. J. Chem. Eng. 15 (1998) 108–112, https://doi.org/10.1590/S010466321998000200002. [53] T.P. Sorokina, L.A. Buluchevskaya, O.V. Potapenko, V.P. Doronin, Conversion of nickel and vanadium porphyrins under catalytic cracking conditions, Pet. Chem. 50 (2010) 51–55, https://doi.org/10.1134/S096554411001007X. [54] U.J. Etim, B. Xu, P. Bai, R. Ullah, F. Subhan, Z. Yan, Role of nickel on vanadium poisoned FCC catalyst: a study of physiochemical properties, J. Energy Chem. 25 (2016) 667–676, https://doi.org/10.1016/j.jechem.2016.04.001. [55] H.S. Cerqueira, G. Caeiro, L. Costa, F. Ramôa Ribeiro, Deactivation of FCC catalysts, J. Mol. Catal. A Chem. 292 (2008) 1–13, https://doi.org/10.1016/j.molcata.2008. 06.014. [56] N.L.A. Souza, I. Tkach, E. Morgado, K. Krambrock, Vanadium poisoning of FCC catalysts: a quantitative analysis of impregnated and real equilibrium catalysts, Appl. Catal. A Gen. 560 (2018) 206–214, https://doi.org/10.1016/j.apcata.2018. 05.003. [57] R. Sadeghbeigi, Fluid Catalytic Cracking Handbook, Third edition, (2012), https:// doi.org/10.1016/C2010-0-67291-9. [58] V.A. Tsiatouras, N.P. Evmiridis, Study of interactions between ion-exchanged chromium and impregnated vanadium in USY zeolite material, Ind. Eng. Chem. Res. 42 (2003) 1137–1144, https://doi.org/10.1021/ie020647n. [59] M.L. Occelli, Vanadium-Zeolite Interactions in Fluidized cracking Catalysts, Catal. Rev. 33 (1991) 241–280. [60] P. Bai, U.J. Etim, Z. Yan, S. Mintova, Z. Zhang, Z. Zhong, X. Gao, Fluid catalytic cracking technology: current status and recent discoveries on catalyst contamination, Catal. Rev. Sci. Eng. 61 (3) (2019) 333–405, https://doi.org/10.1080/ 01614940.2018.1549011. [61] A.S. Escobar, F.V. Pinto, H.S. Cerqueira, M.M. Pereira, Role of nickel and vanadium over USY and RE-USY coke formation, Appl. Catal. A Gen. 315 (2006) 68–73, https://doi.org/10.1016/j.apcata.2006.09.004. [62] A.S. Escobar, M.M. Pereira, R.D.M. Pimenta, L.Y. Lau, H.S. Cerqueira, Interaction between Ni and V with USHY and rare earth HY zeolite during hydrothermal deactivation, Appl. Catal. A Gen. 286 (2) (2005) 196–201. [63] S.-J. Yang, Y.-W. Chen, C. Li, Vanadium-nickel interaction in REY zeolite, Appl. Catal. A Gen. 117 (2) (1994) 109–123, https://doi.org/10.1016/0926-860X(94) 85092-5. [64] P. O’Connor, A.C. Pouwels, FCC catalyst deactivation: a review and directions for

[65]

[66]

[67]

[68]

[69]

[70]

[71]

[72] [73]

[74] [75]

[76]

[77]

[78]

[79] [80]

[81]

[82]

[83]

[84]

[85]

further research, Studies in Surface Science and Catalysis, 1994, pp. 129–144, , https://doi.org/10.1016/s0167-2991(08)62734-2. L.G.J. De Haart, G. Blasse, The application of nickel molybdate and nickel vanadates as photoanodes in a photoelectrochemical cell, Mater. Chem. Phys. 12 (6) (1985) 545–550. N.D. Salvia, Formation of Ni and Mg vanadates during the flameless oxy-combustion of heavy fuels/N. D. Salvia, M. Malavasi, G. D. Salvia, A. Vaccari, Fuel Process. Technol. 138 (2015) 534–539. M. Xue, J. Ge, H. Zhang, J. Shen, Surface acidic and redox properties of V-Ag-Ni-O catalysts for the selective oxidation of toluene to benzaldehyde, Appl. Catal. A Gen. 330 (2007) 117–126, https://doi.org/10.1016/j.apcata.2007.07.014. F.V. Pinto, A.S. Escobar, de B.G. Oliveira, Y.L. Lam, H.S. Cerqueira, B. Louis, J.P. Tessonnier, D.S. Su, M.M. Pereira, The effect of alumina on FCC catalyst in the presence of nickel and vanadium, Appl. Catal. A Gen. 388 (2010) 15–21, https:// doi.org/10.1016/j.apcata.2010.07.055. V. Chuzlov, G. Nazarova, E. Ivanchina, E.Ivashkina, I. Dolganova, A. Solopova, Increasing the economic efficiency of gasoline production: reducing the quality giveaway and simulation of catalytic cracking and compounding, Fuel Process. Technol. 106139. H. Gika, G. Kaklamanos, P. Manesiotis, G. Theodoridis, Chromatography: highperformance liquid chromatography, Encycl. Food Heal. (2015) 93–99, https://doi. org/10.1016/B978-0-12-384947-2.00159-8. B.V.K. Naidu, N.N. Mallikarjuna, T.M. Aminabhavi, Blend compatibility studies of polystyrene/poly(methyl methacrylate) and polystyrene/styrene-acrylomtrile by densitometry, viscometry, refractometry, ultraviolet absorbance, and fluorescence techniques at 30°C, J. Appl. Polym. Sci. 94 (2004) 2548–2550, https://doi.org/10. 1002/app.20973. J. Rheims, J. Köser, T. Wriedt, Refractive-index measurements in the near-IR using an Abbe refractometer, Meas. Sci. Technol. 8 (1997) 601–605. S. Chiaberge, T. Fiorani, P. Cesti, Methyldibenzothiophene isomer ratio in crude oils: Gas chromatography tandem mass spectrometry analysis, Fuel Process. Technol. 92 (11) (2011) 2196–2201, https://doi.org/10.1016/j.fuproc.2011.07. 011. S. Lowell, J.E. Shields, The single point BET method, Powder Surface Area and Porosity, 1984, pp. 30–35, , https://doi.org/10.1007/978-94-009-5562-2_5. Y.R. Katsobashvili, G.A. Teplyakova, Production of light fuels by hydrocracking high-sulfur asphaltic crudes, Chem. Technol. Fuels Oils 20 (7–8) (1984) 370–371, https://doi.org/10.1007/BF00731651. S.M. Simo, S.A. Naman, K.R. Ahmed, Investigate the carbon distribution and structural group composition of two kurdistan crude oils (T-21A PF2) and their fractions, ICOASE 2018 - International Conference on Advanced Science and Engineering, 8548925 2018, pp. 409–414, , https://doi.org/10.1109/ICOASE. 2018.8548925. E.D. Ivanchina, E.N. Ivashkina, G.Y. Nazarova, V.I. Stebeneva, T.A. Shafran, S.V. Kiseleva, KDV, N.V. Korotkova, R.V. Esipenko, Development of the Kinetic Model of Catalytic cracking, Catal. Ind. (2017) 477–486, https://doi.org/10.18412/ 1816-0387-2017-6-477-486. E. Ivanchina, E. Ivashkina, G. Nazarova, Mathematical modelling of catalytic cracking riser reactor, Chem. Eng. J. 329 (2017) 262–274, https://doi.org/10. 1016/j.cej.2017.04.098. V. Parmon, Thermodynamics of Non-Equilibrium Processes for Chemists with a Particular Application to Catalysis, (2010). M. Guisnet, P. Magnoux, Coking and deactivation of zeolites: influence of the pore structure, Appl. Catal. 54 (1) (1989) 1–27, https://doi.org/10.1016/S01669834(00)82350-7. K. Moljord, P. Magnoux, M. Guisnet, Coking, aging and regeneration of zeolites XV. Influence of the composition of HY zeolites on the mode of formation of coke from propene at 450°C, Appl. Catal. A Gen. 122 (1) (1995) 21–32, https://doi.org/10. 1016/0926-860X(94)00210-X. N.S. Gnep, Ph. Roger, P. Cartraud, M. Guisnet, B. Juquin, Ch. Hamon, Creation de mesopores dans des zeolithesmonodimensionnelles. Une methode pour ameliorer leur stabilite catalytique, C. R. Acad. Sci. 309 (1989) 1743. P.V. Lipin, V.P. Doronin, T.I. Gulyaeva, Conversion of higher n-alkanes under deep catalytic cracking conditions, Pet. Chem. 50 (2010) 362–367, https://doi.org/10. 1134/S0965544110050075. C. Vincz, R. Rath, G.M. Smith, B. Yilmaz, R. McGuire, Dendritic nickel porphyrin for mimicking deposition of contaminant nickel on FCC catalysts, Appl. Catal. A Gen. 495 (2015) 39–44, https://doi.org/10.1016/j.apcata.2015.01.043. S.N. Khadzhiev, Cracking of Petroleum Fractions on Zeolite Catalysts, (1982).

Nomenclature Abbreviations BBF: butane-butylene fractions CAH: concentration of aromatic hydrocarbons CSH: concentration of saturated hydrocarbons PPF: propane-propylene fraction

Latin letters AZSM: deactivation factor for reactions catalyzed by ZSM-5

11

Fuel Processing Technology 200 (2020) 106318

G. Nazarova, et al.

Wj: rate of j-th reaction in forward direction (mol/(s·m3)) W-j: rate of j-th reaction in inverse direction (mol/(s·m3)) CSH/CAH: saturated to aromatic hydrocarbons ratio

AY: deactivation factor for reactions catalyzed by Y AH2: deactivation factor of dehydrogenation, aromatization, condensation and coke formation reactions ccat: heat capacity of the catalyst (J/kg·K) cf: heat capacity of the flow (J/kgK) Сi: concentration of the i-th hydrocarbons group (mol/m3) Gcat: catalyst flowrate (kg/s) Gf: feedstock flowrate (kg/s) СNi: the nickel content in the feedstock (ppm) RT: the gas constant (J·mol/К) Т: flow temperature (К) Tf: feedstock temperature (K) Tcat: catalyst temperature after regeneration (K) Тit: thermal equilibrium temperature between feedstock and catalyst (K)

Greek letters (ΔrHoT)j: heat effect of the direct chemical reaction (J/mol) (ΔrHoT)-j: heat effect of the inverse chemical reaction (J/mol) ρf: density of the flow (kg/m3) τ: residence time (s) ψ: parameter of catalyst deactivation by coke, nickel and vanadium

12