Fundamentals of Hydrotreating Part 1

Chemistry

The type and amount of impurities to be removed by catalytic hydrotreating in a petroleum distillate can vary substantially depending on the type and source of the feed. In general, light feeds (e.g., naphtha) contain very little and few types of impurities, while heavy feeds (e.g., residua) possess most of the heavy compounds present in a crude oil. Apart from having a high concentration of heavy compounds, the impurities in heavy feeds are more complex and refractory (i.e., difficult to react) than those present in light feeds. That is why hydrotreating of light distillates is conducted at lower reaction severity, whereas heavy oils require higher reaction pressures and temperatures.

The reactions occurring during catalytic hydrotreating can be classified in two types: hydrogenolysis and hydrogenation. In hydrogenolysis a carbon-heteroatom single bond undergoes "lysis" by hydrogen. The heteroatom is any atom other than hydrogen or carbon present in petroleum, such as sulfur, nitrogen, oxygen, and metals. In hydrogenation, hydrogen is added to the molecule without cleaving bonds. The principal hydrogenolysis and hydroge-nation reactions in catalytic hydrotreating are described below.

Hydrogenolysis reactions

• Hydrodesulfurization (HDS). Removal of organic sulfur compounds from a petroleum fraction and conversion to hydrogen sulfide (H2S). Sulfur removal difficulty increases in the following order: paraffins < naphthenes < aromatics. The type of sulfur compounds can be classified as mercaptans, sulfides, disulfides, thiophenes, benzothiophenes, dibenzothiophenes, and substituted dibenzothiophenes. The ease of removal of these sulfur compounds is in the same order, the mercaptans being the easiest to remove and dibenzothiophenes the most difficult.

• Hydrodenitrogenation (HDN). Removal of organic nitrogen compounds and conversion to ammonia (NH3). Removal of nitrogen requires more severe reaction conditions than does HDS. The molecular complexity (five- and six-membered aromatic ring structures), the quantity, and the difficulty of nitrogen-containing molecules to be removed increase with increasing boiling range of the distillate. Nitrogen compounds can be basic or nonbasic. Pyridines and saturated heterocyclic ring compounds (indoline, hexahydrocarabazole) are generally basic, whereas pyrroles are nonbasic.

• Hydrodeoxygenation (HDO). Removal of organic oxygen compounds and conversion to water. Similar to HDS and HDN, lower- molecular-weight oxygen compounds are easily converted, while higher-molecular-weight oxygen can be difficult to remove. Phenol is one of the most difficult oxygen compounds to convert.

• Hydrodemetallization (HDM). Removal of organometals and conversion to the respective metal sulfides. Nickel and vanadium being the most common metals present in petroleum, hydrodemetallization is frequently subdivided into hydrodeniquelization (HDNi) and hydrodevanadization (HDV). Once metal sulfides are formed, they are deposited on the catalyst and contribute to irreversible deactivation.

Hydrogenation reactions

• Saturation of olefins. Conversion to their saturated homologs of organic compounds containing double bonds.

• Saturation of aromatics, or hydrodearomatization (HDA). Conversion of aromatic compounds to naphthenes. The aromatic compounds found in petroleum distillates are mono – , di – , tri – , and polynuclear aromatics. Monoaromatics are much more difficult to saturate than the others since their saturation requires more energy.

• Hydrocracking (HYC). During hydrotreatment of light and middle distillates, some hydrocracking can occur, but its extent is normally low. However, when processing heavy feeds it can be very high. Hydrocracking is also a hydrogenolysis reaction in which carbon-carbon bonds are broken.

Asphaltenes can undergo both types of reactions (hydrocracking and hydrogenation) depending on reaction conditions. At relatively low or moderate temperatures, the reaction is more hydrogenation dominated during hydrocracking of heavy residue; however, at high temperatures, hydrocracking is more prominent. The overall conversion of asphaltenes is called hydrodeas-phaltenization (HDAsp). Examples of some typical reactions occurring during catalytic hydrotreating are presented in Figure 3.13.

Figure 3.13. Examples of typical hydrotreating reactions. 3.2.2 Thermodynamics

There are fundamental differences in the removal of various impurities, largely because of the structure of the different molecules. HDS and olefin saturation are the most rapid reactions, and HDN and HDA are the most difficult. In contrast to HDS, for HDN the aromatic must first be saturated, and then the nitrogen is removed. Most of the reactions are irreversible with the exception of HDA, which is equilibrium limited at high temperatures, since at these conditions the reverse reaction of naphthene dehydrogenation becomes favored.

All the hydrotreating reactions are exothermic, causing an increase in the reactor temperature as the feed passes through the catalyst bed. The reactor AT value depends on the concentration of each heteroatom and the extent of each reaction during hydrotreatment. The heat of the reaction varies significantly among the different reactions and from one compound to the other, as can be seen in Table 3.3 (Ali, 2007). As the number of moles of hydrogen required to remove each organocompound increases, the amount of heat released also increases.

Equilibrium constants of different hydrotreating reactions are also reported in Table 3.3. From these values the following observations can be made:

• The values of Keq of HDS and HDN are positive over a wide range of temperatures (within values commonly reported on a commercial scale), which indicates that these reactions are essentially irreversible and can proceed to completion if hydrogen is present in stoichometric quantity.

TABLE 3.3. Equilibrium Constants and Standard Enthalpies of Various Hydrotreating Reactions

		logio Keq	at Temperature (°C):
Reaction	25	100	200	300	400	AH°a
Hydrodesulfurization C3H7 — SH + H2 0- C3H8 + H2S	10.57	8.57	6.92	5.87	5.15	-57
Thiophene + 3H2 0 «C4H10 + H2S	30.84	21.68	14.13	9.33	6.04	- 262
Benzothiophene + H2 0 ethylbenzene + H2S	29.68	22.56	16.65	12.85	10.20	-203
Dibenzothiophene + 2H2 0 biphenyl + H2S	24.70	19.52	15.23	12.50	10.61	-148
Hydrodenitrogenation Indole + 3H2 0 ethylbenzene + NH3	—	—	—	7.8	5.0	-49
Carbazole + 2H2 0 biphenyl + NH3	—	—	—	6.8	5.1	- 126
Pyridine + 5H2 0 n-pentane + NH3	—	—	—	8.9	4.4	- 362
Quinoline + 4H2 0 propylbenzene + NH3	—	—	—	7.0	3.3	-272
Hydrogenation of aromatics Naphthalene + 2H2 0 tetralin	—	—	1.26	- 1.13	-2.80	-140
Tetralin + 3H2 0 trans-decalin	—	—	0.74	- 2.95	- 5.56	- 193
Cyclohexylbenzene + 3H2 0 cyclohexylhexane	—	—	2.47	- 1.86	-4.91	- 295
Phenanthrene + 4H2 0 octahydrophenanthrene	—	—	1.16	-3.64	-7.12	-251

• In general, as the temperature increases, the values of Keq decrease, which is in agreement with the exothermicity of the reactions.

Most HDS reactions are straightforward except those of aromatic sulfur species, which must start with ring opening and sulfur removal, followed by saturation of the resulting olefin. In the case of the HDS of dibenzothiophenes, there are two major pathways: direct hydrodesulfurization, in which the sulfur atom is removed from the structure and replaced by hydrogen without hydro-genation of any of the other carbon-carbon double bonds; and the hydrogena-tion route, which assumes that at least one aromatic ring adjacent to the sulfur-containing ring is first hydrogenated before the removal of sulfur. Also, an aromatic ring may be hydrogenated after sulfur removal. The hydrogena-tion pathways are subject to thermodynamic equilibrium constraints. Thus, the partially hydrogenated intermediates have lower equilibrium concentrations at higher temperatures, and HDS via the hydrogenation route becomes limited at low pressures and high temperatures.

In the case of the hydrogenation of aromatic ring compounds, it is also an exothermic reaction, and equilibrium yields are favored by low temperatures. The maximum aromatic reduction (i.e., the optimum reaction temperature) is a function of the types and amount of aromatic compounds in the feed, space velocity, hydrogen partial pressure, and catalyst type. Complete hydrogenation of aromatics is not possible, owing to equilibrium limitations under typical hydrotreating conditions.

For the hydrodearomatization reaction to proceed, the polynuclear aromatics are first hydrogenated to three- ring to two- ring to one- ring and to the end products (naphthene rings). Saturation of the final aromatic ring is difficult because of the resonance stabilization of the monoaromatic ring.

Olefin saturation is very rapid and highly exothermic. For example, HDN shows a heat of reaction of 1Btu/lb of feed for each 100 ft3 of hydrogen consumed, HDS generates 1 Btu/lb of feed for each 10 ft3 of hydrogen consumed, and the olefin saturation generates 1 Btu/lb of feed for each 2 ft3 of hydrogen consumed. Diolefins are readily hydrogenated to olefins at low temperatures (Gary and Handwerk, 2001).

Heats of hydrotreating reactions have been reported in the literature for reactor modeling purposes. For example, Tarhan (1983) used the following values for hydrotreating of straight-run gas oil:

Hydrodesulfurization	-251,000 kJ/kmol
Hydrodeoxygenation	-68,200 kJ/kmol
Hydrodenitrogenation	-64,850 kJ/kmol
Hydrocracking	-41,000 kJ/kmol
Hydrogenation	-125,520 kJ/kmol

Others prefer the use of the overall heat of reaction, as in the case of the HDS of atmospheric residue (AHr = -7820kJ/kg sulfur = -250,748 kJ/kmol). Overall heats of reaction are mean values derived from heat balances of several similar HDT processes and include the contribution of all the reactions (Shah and Paraskos, 1975; Dohler and Rupp, 1987).

Kinetics

Most of the kinetic studies reported in the literature of the various hydrotreat-ing reactions have been conducted using pure compounds (i.e., model compounds), as well as binary and multicomponent mixtures of them (Girgis and Gates, 1991). The available kinetics data with model compounds are usually represented with pseudo-first-order rate equations or with Langmuir-Hinshelwood rate equations. However, the complexity of the individual reactions occurring in an extremely complex mixture and the interference of the products with those from other components of the mixture is unpredictable. Or the interference of secondary and tertiary products with the course of a reaction, and hence with the formation of primary products, may also be a cause for concern. Hence, caution is advised when applying the data from model compound studies to the behavior of petroleum and its distillates. Kinetic data derived from model compounds cannot be expected to include contributions from the various steric effects that are a consequence of complex molecules containing three-dimensional structures. Indeed, such steric effects can lead to the requirement of additional catalyst and process parameters for the various heteroatoms removed.

For the HDT of real feeds, nth-order kinetics with respect to total concentration of the heteroatom is usually employed, in which the n value depends on several factors, such as type and concentration of the heteroatom, catalyst properties, type of feed, operating conditions, and experimental system, among others.

Hydrodesulfurization The structural differences between the various sulfur-containing molecules make it impractical to have a single rate expression applicable to all reactions in HDS. Each sulfur-containing molecule has its own hydrogenolysis kinetics, which is usually complex.

The complex nature of oil fractions with sulfur compounds exhibiting very different reactivities as well as the presence of others, such as nitrogen (basic and nonbasic), aromatics, and so on, reacting at the same time and competing for the same active sites and also inhibiting the effects of by-products of the same reactions (e.g., hydrogen sulfide) have limited HDS experimental studies to model compounds ranging from easy to desulfurize (e.g., thiophene) to difficult to desulfurize (e.g., 4,6-dimethyldibenzothiophene). These are the main reasons why few works have been reported dealing with experiments with real petroleum feedstocks under industrial conditions, since most of the time it is not simple to extract individual effects and one does not know which one to blame. However, when a catalyst formulation is almost ready for commercial application, experiments with real feeds are mandatory. Testing with real feeds not only for exploration of the commercial application of new catalyst formulations but also for process design and optimization studies is a very important step for new technology development. For the latter issue, kinetic data obtained from experiments with real feeds are of great interest, since they are employed for reactor modeling, simulation, and optimization.

Figure 3.14. First-order kinetic constant values for the HDS of different sulfur compounds in a diesel fraction.

When real feeds and their hydrotreated products are characterized in detail, as in the case of sulfur compounds during HDS of straight- run gas oil and other light petroleum distillates, experimental results obtained by gas chroma-tography with a sulfur chemiluminescence detector have indicated that simple first-order kinetics with respect to the heteroatom is the predominant mechanism by which it is removed from the organic material. However, the various molecules have very different reactivity, as illustrated in Figure 3.14 for several sulfur compounds included in diesel fractions. The differences in HDS reactivity of the sulfur compounds are clearly distinguished. As is well known, DBTs with 4-, 6-, or 4, 6-alkyl positions are the most refractory compounds.

When the analysis is performed as the total content of the heteroatom (e.g., total sulfur content, total nitrogen content), the kinetics is typically represented by nth-order kinetics with respect to the total concentration of that heteroatom. The value of n for most hydrotreating reactions is in general larger than 1. For example, Table 3.4 reports a compilation of reaction orders and activation energies obtained for the hydrodesulfurization of different real feedstocks. The table has two parts; one shows reaction orders as calculated with experimental data, and the second presents assumed values of reaction order. For calculated reaction orders, in general an increase in their values in the range 1.5 to 2.5 is observed as the sulfur content is increased. Some data do not follow this trend, which may be due to differences in conditions used during experiments. This tendency with respect to sulfur content in the feed is not observed for activation energies. For different sulfur contents (3.72 and 5.68 wt%), two feeds were reported to have the same reaction order and activation energy (nS = 2 and EA = 29kcal/mol), or for almost the same sulfur content (3.45 and 3.72 wt%), two other feeds presented very different activation energies (68.6 and 29 kcal/mol) for the same reaction order (nS = 2). Therefore, it is clear that reaction order and activation energy depend on the type and distribution of heteroatom compounds in the oil fraction as well as on the catalyst and reaction conditions employed. Figure 3.15 corroborates the fact that reaction orders can be different for the HDS of middle distillates having the same amount of sulfur but coming from different sources (i.e., crude oils). The development of general kinetic data for the hydrodesulfuriza-tion of different feedstocks is complicated by the presence of a large number of sulfur compounds, each of which may react at a different rate because of structural differences as well as differences in molecular weight.

TABLE 3.4. Reaction Orders and Activation Energies for Hydrodesulfurization of Different Feedstocks

Feed2	Density at 15°C	Sulfur (wt%)	Distillation Range ( ° C)	n	Ea (kcal/mol)
	Reaction Orders Calculated from Experimental Data
SRGO	0.861	1.31	213-368	1.57	20.3
SRGO	0.843	1.32	188-345	1.53
HSRGO	0.862	1.33	142-390	1.65
SRGO-LCO	0.879	1.78	209-369	1.63	16.5
VGO	0.907	2.14	243-514	2.09	33.1
SRGO-LCO	0.909	2.44	199-370	1.78	16.37
Residue oil	0.910	3.45	—	2.0	68.6
Residue oil	0.950	3.72	281-538	2.0	29.0
Residue oil	1.007	5.30	—	2.5	36.1
AR	0.995	5.86	—	2.0	29.0
	Assumed Values of Reaction Orders
Used oil	0.900	0.70	—	1.0	19.6
Residue oil	0.969	1.45	—	1.0	24.0
SRGO	—	1.47	—	1.65	25.0
Residue oil	0.964	2.90	—	1.0	18.3
CGO	0.984	4.27	196-515	1.5	33.0

"SRGO, straight-run gas oil; HSRGO, heavy straight-run gas oil; LCO, light cycle oil; VGO, vacuum gas oil; AR, atmospheric residue; CGO, coker gas oil.

Figure 3.15. Reaction order for the HDS of middles distillates from different crude oils (CoMo/y-Al2O3, 340 to 360°C, 1.5 to 2.0 h-1 LHSV, 54 kg/cm2, H2/oil ratio of 2000ft3/bbl).

One way to represent the HDS reaction is the practical and widely accepted generalized stoichiometric equation, which lumps the HDS reaction of all the sulfur compounds into a single expression:

where vS, Uh2, Uhc, and Uh2s are the stoichiometric coefficients of the organic sulfur compounds, hydrogen, sulfur-free hydrocarbon, and hydrogen sulfide, respectively.

The simplest model that can be used to represent HDS kinetics is the power-law type, which does not take into account the inhibiting effect of H2S (Cotta et al., 2000):

Another simplified kinetic model considers the existence of only two reactive sulfur components by means of the following first-order rate equation (Gates et al., 1979):

where y, k1, and k2 are adjustable parameters. The parameter y represents the fraction of the easy-to-react sulfur-containing compounds, and 1 – y is the fraction of the more refractory sulfur-containing compounds. These simplified kinetic models [Eqs. (3.11) and (3.12)] are valid only when partial pressures of hydrogen and hydrogen sulfide are held constant, and these values are incorporated in kHDs, k1, and k2.

The most frequently used kinetic expression of HDS is the following Langmuir-Hinshelwood model:

The exponent 2 in the denominator of Eq. (3.13) represents the number of sites in adsorption for hydrogen sulfide.

When HDS is studied together with HDA, the following reaction has been proposed, which is assumed to be irreversible under normal hydrotreating conditions (Chowdhury et al., 2002):where A represents aromatic compounds. For this HDS reaction, the following form of the Langmuir-Hinshelwood rate equation has been considered:

Equations -3.13) and -3.15) include an adsorption- equilibrium constant of hydrogen sulfide (Kads), which is a function of temperature and can be estimated using the van’t Hoff equation:

The reaction orders for sulfur and hydrogen, nS and m, respectively, have been reported to range between 1.5 and 2.5 for sulfur, depending on the type of feed as well as the amount and type of sulfur compounds, and between 0.5 and 1.0 for hydrogen. According to the chemical dissociation of the hydrogen molecule on the catalyst surface, the theoretical value of m should be 0.5. However, it approaches 1.0 if the mass transfer rate of hydrogen becomes the limiting step (Cheng et al., 2004).

Apart from H2S, hydrodesulfurization can also be inhibited by other compounds, so that the term indicating competitive adsorption can be expressed by a sum of terms as follows:

A combination of the use of two reactive sulfur components with inhibition by H2S and aromatics yields the following kinetic model (Avraam and Vasalos, 2003 ):

The reaction orders nS and m in Eqs. (3.18) and (3.19) have been reported to be equal to 1.

It is well accepted that hydrodesulfurization undergoes through two reaction paths: direct HDS (DD) by hydrogenolysis of the reactants, and indirect HDS (ID) by hydrogenation of one aromatic ring followed by C-S bond cleavage of the hydrogenated intermediate products. It is also recognized that deep HDS is inhibited by the nitrogen compounds and aromatics, and they compete with sulfur compounds only on the hydrogenation sites, thus inhibiting the hydrogenation route. Based on these considerations, and assuming that H2 S inhibition is negligible, hydrogen is in excess, and sulfur conversion behaves as a pseudo- first- order reaction, the following kinetic model has been proposed (Liu et al., 2008):

where &hds and &Hds are the rate constants for the direct HDS and indirect HDS routes, respectively.

Hydrodenitrogenation As in the case of HDS, all of the nitrogen compounds present in the feed are sometimes lumped together and the following reaction is assumed for nitrogen removal:

where R-N is the hydrocarbon structure containing nitrogen, and R-H is the nitrogen – free hydrocarbon.

The power-law model and Langmuir-Hinshelwood rate equation have been used for determining nitrogen removal:

For example, when aromatic hydrocarbons and H-S are to be the inhibiting compounds, Eq. (3.17) becomes

where nN is the reaction order of the nitrogen compounds. Equation (3.21) is used most frequently when during experiments it is not possible to identify the ammonia content in the gas-phase exit stream.

Not only can ammonia inhibit HDN reactions but other compounds also, such as H2S and aromatics. To account for the inhibition effect of these other compounds, similar to HDS, the following rate expression can be used:

The kinetics of HDN has been reported to be represented by the power-law model with a value of nN = 1.5 (Bej et al., 2001). For nonbasic nitrogen,

and for basic nitrogen,

Since nitrogen is present in petroleum as basic (NB) and nonbasic (NWB) compounds, the following consecutive reaction scheme in also considered for nitrogen removal, in which nonbasic nitrogen is hydrogenated first to basic nitrogen, which undergoes further reactions to eliminate the nitrogen atom from the molecule:

Hydrodearomatization During hydrotreating, the HDA reaction is controlled kinetically at low temperature but controlled thermodynamically at high temperature. This means that when the reaction temperature increases, the hydrogenation of aromatics increases, passes through the maximum, and then decreases. This behavior can be represented in a general form by the following simple reversible reaction and first – order rate expression:

where CA and Cnaph are the concentrations of aromatics and naphthenes, respectively, and kf and kr are the forward (hydrogenation) and reverse (dehydrogenation) rate constants. In hydrotreating operations, hydrogen is used in excess so that H2 partial pressure can be assumed constant as well as the term kfpm2.

Another approach to modeling the HDA reaction is by the separation of the total aromatic content into three groups of aromatics: monoaromatics, containing a single aromatic ring in their structure; diaromatics, with two aromatic rings; and polyaromatics, with three or more aromatic rings. Based on this consideration, the following stoichiometric equations can be established (Chowdhury et al., 2002):

The naphthenes are the final products of HDA reactions and are composed primarily of alkylmono- to alkylhexacycloparaffins. The HDA reaction rates are expressed as (Cheng et al., 2004)

The first reaction order of the aromatics compounds is in general explained by their strong adsorption on the catalyst surface. There is a preference for hydrogenation of the first ring(s) in polyaromatics, since the hydrogenation rate of the first ring in condensed diaromatics (e.g., naphthalene) has been found to be much faster (typically, 20 to 40 times) than that of monoaromatics (including byphenyl, tetralin, and cyclohexylbenzene) (Cooper and Donnis, 1996). In addition, the reaction DA ^ MA is more severely limited by equilibrium than is the reaction PA ^ DA. Based on this, the previous set of equations can be simplified to (Chowdhury et al., 2002; Bhaskar et al., 2004)

Working at close to constant hydrogen partial pressure allows for lumping rate constants ( kf and kr) with p’mi2 as follows:

Therefore, Eqs. (3.37) to (3.40) reduce to

To determine the forward and reverse reaction rate constants, equilibrium constants are defined as

The equilibrium constants for reversible reactions are determined at different temperatures by using the van’t Hoff correlation:

The values of the equilibrium constants decrease when the number of side chains and the number of carbon atoms in each side chain increases until a maximum temperature, where this behavior shifts. For any equilibrium constant, an increase in reaction temperature results in a lower equilibrium constant and a higher equilibrium concentration of aromatics. At low temperatures (<443°C) the order of the equilibrium constant values is Kma > Kda > Kpa, while at high temperatures (>443°C) the order changes to KPA > KDA > KMA (Chowdhury et al., 2002; Bhaskar et al., 2004).

Sulfur- and nitrogen-containing compounds can reversibly and slowly inhibit aromatic hydrogenation, and the poisoning tendency of various nitrogen compounds seems to be related not only to its basicity, but probably also to its structure.

Hydrogenation of Olefins Olefins react with hydrogen to form saturated hydrocarbons. The following stoichiometric equation is assumed to represent olefin hydrogenation:

The hydrogenation of olefins (HGO) is represented by a pseudo-first – order reaction with respect to the total concentration of olefins:

the power-law kinetics is represented by

Hydrodeasphaltenization The way in which asphaltenes react is very important during hydrotreating. It is to be remembered that asphaltenes reduce the reaction rate of other reactions during hydrotreating, since asphaltenes are coke precursors that deactivate catalysts by plugging the catalytic sites. Different approaches have been developed to represent the manner in which asphaltenes react, and kinetic parameters (e.g., reaction order and kinetic constant) can be extracted from such models.

The power-law model is by far the most simplified model to use to represent the kinetics of HDAsp. With this approach, different types of reactivity of asphaltenes are not taken into account. Instead, all asphaltene molecules are considered to react at an average rate expressed in the kinetic constant. For the hydrodeasphaltenization reaction.

To simplify the kinetic modeling, HDA is also represented by the following irreversible reactions in series and rate equations:

where Casp is the concentration of asphaltene and nasp the reaction order for asphaltenes.

Another approach based on the power-law model is to separate the global HDasp reaction into two reactions in parallel:

where y is the fraction of hard-to-react asphaltenes (asp- ) and (1 – y) is the fraction of easy-to-react asphaltenes (asp2). According to the assumptions of Kwak et al. (1992)- the reaction order for both types of asphaltenes is 1. The kinetic model is

Hydrodemetallization Studies with individual metalloporphyrin model compounds have indicated that the hydrodemetallization mechanisms involve a reversible hydrogenation step of a Ni (or V)-EP to form a Ni (or V)-EPH2 followed by an irreversible hydrogenolysis, which results in the fragmentation of the porphyrin ring and deposition of the metal on the catalyst, according to the following reaction network:

where M is Ni or V.

The removal of metals contained mostly in heavy petroleum fractions, such as residues, cannot be represented by the same reaction network, since apart from metalloporphyrin, nonporphyrin metal compounds are also present. The kinetics of HDM is thus commonly expressed by a power-law model:

Hydrocracking The extent of hydrocracking is determined primarily by reaction conditions. During hydrotreating, in general, light and middle distillates exhibit low hydrocracking, whereas for heavy feed it is very high. Kinetic models for HYC, particularly of heavy feeds, have been discussed in detail in Section 2.3, so that in this section, only kinetics of hydrocracking occurring during HDT of light feeds is considered.

The typical support of hydrotreating catalysts is 7-alumina (/-Al2O3),whose acid sites promote mild hydrocracking reactions and produce light hydrocarbons in gas and lighter liquid products. The common hydrocracked products for different feeds are:

Heavy gas oil: diesel and light gases Light gas oil: naphtha and light gases Naphtha: C5-C6 and light gases

The kinetics of the hydrocracking reaction taking place during hydrotreating can be represented by a three-lump model, as shown in Figure 3.16. From this figure the following pseudo-first-order kinetic equations can be derived. For example, when the feed is light gas oil (LGO), the hydrocracked products are naphtha (NT) and light gases (gas):

The concentrations of heteroatoms (sulfur, nitrogen, oxygen, metals) and asphaltenes in a petroleum fraction are usually reported in weight percent. The transformation of mass concentration to molar concentration, which is needed unit for kinetic and reactor modeling purposes, can be done easily using the equation

For olefins, the common analysis to determine the content of unsaturates is the bromide number (Br No.). Therefore, the molar concentration of double bonds can be expressed as follows:

Figure 3.16. Three-lump kinetic model for hydrocracking reaction during catalytic hydrotreating.

where poil is the density of the hydrocarbon, which varies as feed passes through the catalyst bed, and PMi , PMBr, and PMoil are the molecular weights of each heteroatom, bromide, and the hydrocarbon, respectively.