• Author / Uploaded
• Nhat Cao

Citation preview

Il.IT,a~~l .'

' l,R

(:!.:!)

where:

j

reaction of rank j

vij stolchlometrlc

~~effjcient of

component Ai in reaction j

R 'total number of reactions

As a descriptive example of such a system, the set of stoichiometric relations below can describe the methane steam reforming reaction, without taking any possible side reactions into account: -CH 4 - 2 H20 + 4 H2 + CO2 = 0

(2.3a)

-CH 4 - H20 + 3 H2 + CO = 0

(2.30)

-CO - H20 + CO2 + Hz = 0

(2.3c)

The fact that expressions (2.3a), (2.3b) and (2.3c) are not independent is not particularly significant. since it is the rate of the reaction act itself that counts as will be seen later on. Anyway, the reaction as such is probably not described, or is only accidentally described, by one of the relations listed above. It is important to emphasize one point: since a stoichiometric relation is only a material balance, it can not say anything about the actual feasibili~y of the reaction as proposed. It can say even less about the rate at which the assumed change would occur. In the same way as molar balances are written: IVijA i =0

(2.2)

elemental balances can also be written: i= 1,5

j

=l,R

k= 1,£

(2.4)

in this expression: £ik

number of atoms of element k in component Ai

E total number of elements involved

Accordingly, there will be Ex R elemental relations that will have to be complied with simultaneously.

Chapcer 2.

BASIC PRINCiPlES GOVEFWING CHEMICAL

CHANGES

13

The methane steam reforming example shows that: • Equation 2.3a can be broken down as:

-c + C =0

-4 H -4 H + 8 H =0 -20+20=0

• Equation 2.3b as:

-c

+ C =0 -4 H - 2 H + 6 H =0

-0 + 0=0 • and Equation 2.3c as:

-c + C=O

-2 H + 2 H =0 -0-0+20=0 which makes precisely 3 x 3 = 9 elemental relations complied with simultaneously. In the case of complex reaction feeds consisting of petroleum cuts or ;~ctroleum residues, moles will no longer be identifiable, except on a few rare occasions. Only elemental balances can be written and they will have to be complied with as per the accuracy allowed by sample representativity and analyzer performance. This will be the case, for example, of the following elements: C. H. 0, S, N, Ni, V. An intermediate case is when the components of a cut can be grouped together in families: aromatics, paraffins, naphthenes, etc. The lumping technique. which will be discussed later on (see Section 2.3.3.6) can be used to follow a reaction by observing the changes in the families. For example, in a catalytic reforming operation, hydrogen production is' Closely related to the increase in the "aromatics" family. Quasi-stoichiometry can thus be estab.ished by combining elemental balances and balances by families.

: .1.1 The Concept of Advancement (or Molar Extent) Advancement (or molar extent). which should not be confused with conversion. defined later on in Section 2.1.2, can be used to quantify the variation in the number of moles of components, reactants or products that take part in the reaction. It is of no interest as regards elements. since their quantity does not vary. For a closed system and a single reaction, advancement is defined by the -::;fJ!ar extent" ~: (2.5a) n, amount of species Ai (mol)

n.. initial amount of species A. (mol)

For an open system, as are most of the reactors in the oil industry, advancenent or "molar extent" ~' is defined by: (2.5b) vith:

F; F;o

molar flow rate of species Ai (mol/s) inlet molar flow rate of species Ai (mol/s)

With the preceding definlttons.j; is expressed in moles and ~' in moles per econd. ~ and ~' can be negative if the stoichiometric relation has been writ.en in the reverse direction from the one in which the reaction actually proJresses. Sapproaches - ni/vi and;' approaches - Fia/V i when the reaction is com"Jlet,,:. . . the case of a set of R reactions, relations 2.5a and 2.5b are written:

n,=nio + I vi/c,j t.-r; +I vi/t.j

(2.6a) (2.Gb)

Although the concept of advancement (molar extent) is quite useful in molecular stoichiometry, it is of little interest for complex systems where moles are elusive entities. Here, the concept of "conversion- will be used I instead, at least as far as reactants are concerned.

2.1.2 Conversion, Selectivity, and Yield In a closed system, the conversion of a reactant Ai is designated by: (n · -n.) X=_'o__'_ nio and in an open system: (F -F) X=_Jo__J_

r;

(2.7a)

(2.7b)

':onversion is dimensionless and is frequently expressed in %. for petroleum cuts, characterized by distillation range for example, a conversion can generally be defined unambiguously. Take for example the case of cracking a 550°C + residue. The cracking (or conversion) ratio at the cracking reactor exit is equal to:

x = mass flow rateof 5SO·C· feed - mass flowrateof 550·C+ residue in tile reactor effiuent mass flowrateof the feed

Product selectivity expresses the (molar or mass) amount of the product obtained over the theoretical amount that could be expected if the reaction were totally oriented toward getting this product alone.

Chapter2. BASIC PRINCIPLES GOVeRNING CH~MICAL CHA/I;GES

15

With the example of cracking the 550°C+ residue, gasoline mass selectivity is equal to:

s=

mass flow rate of CS_ 80 gasoline in the emuen~__.-:-~~_ mass flow rate of feed - mass flow rate of 550'C + residue in the effluent

Accordingly, if the feed were processed to produce only gasoline exclusively, selectivity would be 100%. Product yield expresses the amount of the product obtained over the amount of feed processed. The gasoline yield in the preceding example is:

Y =mass flow rate of gasoline in the effluent ----m-as-s-f1-o..;.w-r-a-te-o-ff-e-ed---This gives the basic expression: (2.8)

Y=SxX

The yield is equal to the product of selectivity and conversion. However, great care is required in expressing the terms selectivity and yield, for authors differ as to their meaning. For example, selectivity is sometimes defined as the ratio between the mass flow rate of one product and that of another one, or as the ratio between the quotient of mass flow retes of two products and that of theoretical flow rates. Additionally, no confusion should be made between overall conversion and conversion per pass, or between oeerall yield and yield per pass. In Figure2.1 there are two material balances lnvelving a chemical change that can be schematically represented by: .. H(heavy)

~

MCmedium)

~ ~light)

The material balances are fictitious, shown only to illustrate the definition of conversion, selectivity and yield.

In case A M and L are separated before recycling the non-converted part of H and purging amount E: • conversion per pass of H =(129 - 39)/129 = iO% • overall conversion of H =(100- 10)/100 =90% • selectivity for M in relation to H = 45/90 =50% • selectivity for L in relation to H =45/90 =50% • yield per pass of M in relation to H =45/129 =35?~ • yield per pass of L in /elation to H =45il29 =35% • overall yield of M in relation to H = 45/100 =45?6 • overall yield of L in relation to H =45/100 =4576

Chapter 2, BASIC PRINCiPlES GOVERNING OEAIICAL CHANGES

17

It is easy to imagine that overall conversion can exceed conversion per pass by a great deal, and even reach 100% if amount purged E approaches zero. The overall yield can then come close to, if not attain, the selectivity value.

2.2 Thermodynamics of Chemical Reactions The two thermodynamic parameters to be considered when analyzing the thermodynamics of chemical reactions are enthalpy H and Gibbs energy G. Generally speaking, the variations in enthalpy and in Gibbs energy associated with a chemical change will have to be assessed so as to draw the relevant conclusions for the heat balance and for the equilibrium advancement. It should be noted that there is absolute continuity between the thermodynamics of physical equilibria and that of the chemical reaction. This is why the reader is requested to read through Chapter 4, Volume 1 (Methods for the Calculation of Hydrocarbon Physical Properties) and Chapter 2, Volume 2 (Thermodynamics. Phase equilibria) before approaching the subject presented here. In this way, he or she will be familiar with the different concepts used in thermodynamics. Other references [40,31] could also be consulted for ~ more thorough understanding of the topic. First and foremost it is important to define two concepts: the standard state and the origin of enthalpies. a. The Standard State The standard state of a component corresponds to the ideal gas state under an absolute pressure of 1 bar and is identified by the exponent ". It does not imply a reference temperature. Table 2.1 lists the standard states that are most commonly adoptedfor elements in tables. .

Element

Standard state

Carbon Hydrogen Oxygen Nitrogen Sulfur

Crystalline. graphite form Biatomic molecule Biatomic molecule Biatomic molecule Crystalline. orthorhombic form

"""'1' Table . 2.1 Standard state of a number of elements.

I

b. The Origin of Enthalpies

:tis known that the origin of enthalpies can be chosen arbitrarily. Consultation 0i available data suggests choosing elements in their standard state as the

A. Separation of Land M before recycle

[TI] = 39 M=45 L =45 I

E

B. Separation of L before recycle

II rJgUre

2.1

Conversion per pass, overall conversion. selectivity, yield per pass and overall yield.

In case B. M is not separated, It is recycled with the non-converted part 011-1: , conversion per pass of H = (136 - 40)/136 = 70% >

overall conversion of H =(100 - 4)/100

=96%'

) selectivity for L in relation to H = 92/96 = 96%• >

yield per pass of L in relation to H = 92/136 =68%

) over~1l yield of L in relation to H = 92/100 =92%

origin of components' enthalpies. This procedure proves to be very convenient when heats of reaction are to be evaluated.

2.2.1 Enthalpy Variation Associated with a Chemical Reaction The enthalpy variation associated with a chemical reaction measures the heat released or absorbed during a chemical change. It is obtained from the evaluation of partial molar enthalpies of each of the components. 2.2.1.1 Partial Molar Enthalpy

By definition, the partial molar enthalpy of a component Aj at temperature T and pressure P inCl. mixture with other components is written as follows: (2.9)

It can be evaluated by the following expression, provided the convention on standard state and the choice of a reference temperature To are taken into account: hr(A j) = (MlDro(A;) + [hTCA;) - hhCAj)] + hf(A j) + h¥(Aj)

(2.10)

(~()ro (A;) is the variation in standard enthalpy of formation from the elements at reference temperature To. [hr(A;) - hh (Aj) ] is the standard enthalpy variation between temperatures To and T. h;(A;) is an enthalpy corrective term to be considered if the state is different from the standard state. If the conditions laid down correspond to the (liquid or solid) condensed state, the term contains the heat of condensation. h ¥(A;) is a term that takes into account the contribution of component Aj to the heat of mixing. It is generally a relatively insignificant term that is difficult to estimate, and this is why it is frequently omitted.

2.2.1.2 Heat of Reaction

For a system characterized by the stoichiometric equation:

L vjA j = 0

(2.1)

the quantity MlR.Tis called the heat of reaction at temperature T, such that: (2.11)

where hr(A;) is the partial molar enthalpy of component Aj • The tables available generally give access to standard heats of formation at a reference temperature To and to molar heat capacities Cp(A;) [20, 28, 36, 37, 45,46J.

Chapter2. BASIC PRINCIPLES GC1IISNING CHEMICAL CHANGES

1S

This allows calculation of: T [hi-(A;)- hhCAj)} =I [C;CAj dT To and consequently estimation of: T hi-CA j) = CMI,)ToCAj) +I [C;CA;)} dT To In particular, at reference temperature To:

»)

(2.12)

hToCA;) =CtlliDToCA;)

and the standard enthalpy of reaction is defined by:

=r V;(Mli)ToCA;)

(2.13)

CMfFJT= L v. (Mfi)T(A j)

(2.14)

(M-IR)To

It can likewise be shown that: Generally speaking, at moderate pressure (a few bars), the terms corresponding to hP(A;) and hM(A;) can be disregarded. This is true except when the chemical change involves a change in state, as will be seen in the example tJelow. This frequently means that the reference to the standard state is eliminated: (2.15)

2.2.1.3 Example of Calculating Reaction Endlalpy Variation

Given the isopropanol dehydrogenation reaction: --.. CH3-CHOH-CH3 -+ CH3-CO-CH;+ Hz gas at 433 K liquid at 298 K gas at 433K The enthalpy variation corresponding to this reaction is calculated. The thermodynamic data used are the ones published by the Thermodynamic Research Center [37}. The reaction indicated can be broken down into three parts: 1. CH3-CHOH-CH3 (liq., 298 K) -+ CH3-CHOH-CH3 (gas. 298 K) A298 = +45,396 J

(vaporization)

2. CH3-CHOH-CH3 (gas, 298 K) ...,. CH3-CO-CH3 (gas. 298 K) + H2 (gas. 298 K) (..lliR)298 =(llii)298 (CH3-CO-CH~ + (..llif)298

CH:0 -

(~(CH3-CHOH-CH~

=-217150 + 0 + 272295 =+55 145 J

3. CH3-CO-CH3 (gas, 298 K) + H~ (gas, 298 K) -7

CH3-CO-CH3 (gas, 433 K) + H2 (gas, 433 K)

433

f

(.ilig,298~433K) = 298 [Cp(CH3-CO-CH0 + Cp(Hv

1dT= +15 585 J

There is a total of: .ili

=+45 396 + 55145 + 15585 =+116 127 J

In conclusion, for each mole of isopropanol that reacts according to the conditions indicated, the system will require an input of 116 127 J. This is a highly endothermic reaction.

2.2.1.4 Case of Complex Systems Defined by a Set of Several Stoichiometric Relations The enthalpy variation corresponding to reactionj, whose advancement is ~j' is equal to (~h !;j, and the enthalpy variation corresponding to all the reactions taking part in the change is equal to: (2.16)

This of course implies that the reaction can be broken down into its differ-' ent component parts and that the corresponding advancement values can be assigned to them with enough accuracy. As mentioned earlier, this can be quite difficult for reactions involved in refining operations. The difficulties can be circumvented in certain cases. For example in hydrodesulfurization, experience shows that the elimination of a sulfur atom requires 2.7 moles of hydrogen on the average (depending on the nature of the sulfur compounds). Moreover, the disappearance by reaction of a mole of hydrogen releases approximately 60 kJ (depending on the nature of the bonds that need to be saturated) (Thonon in [44]). The desulfurization ratio can therefore be readily. linked to the heat released, or to the temperature increment of the reaction stream going through the adiabatic reactor. The appropriate correlations for different cases (catalytic cracking, hydrocracking, oligomerization, reforming, etc.) can be established without too much difficulty, at least approximately, by designing model systems and applying the principles mentioned above to them.

2.2.2 Gibbs Energy Variation Associated with a Chemical Reaction The Gibbs energy variation associated with a chemical reaction allows the position of the chemical equilibrium state to be situated between reactants and products for specified operating conditions. It also enables the position to be expressed by a value: the equilibrium constant.

Chapter 2.

BAsIC PRINCiPLES GOVERNING CHEMICAL CHANGEs

21

2.2.2.1 . Law of Mass Action

Given the equilibrium reaction: A + 2B::; C or more generally

L ViAi =0

(2.1)

The law of mass action, which defines how far the reaction can go, is written: K

[el

= [AJ(Bj2

and more generally K =n [Ad V ;

(2.17)

where [ ] stands for a concentration, a partial pressure, a mole fraction, a mass fraction, etc. depending on what was chosen as unit of mass density. !\very high Kvalue means that the-reaction can go right to the end. It can be considered complete from a thermodynamic standpoint. This in no way predicts the reaction rate, which may be zero if there is no means of activating it selectively or initiating it (e.g. by combustion). A very low K value means that the reactions advances little. This does not mean that it is not feasible, since difterent solutions can be imagined to overcome this thermodynamic difficulty. Some examples are: wide disproportion among reactants, removal of one of the products as it is formed, separation of products and recycling of the unchanged reactant, etc. ~.~.2.2

Calculating the Equilibrium Constant

The equilibrium constant is related to Gibbs energy variation by the following formula: (2.18) (AGR>r=- RTin K'

KO =n(f;/f;OY'j

where:

(2.19)

The first problem consists in evaluating (AGR)r' It can be shown that, for a reaction represented by the stoichiometric relation L ViA; =0: (2.20)

(ilGnr(A;) is the variation in standard Gibbs energy of formation for component A;. In tables, (~Gf)T(A;) is seklom found directly, however it is easier· to find out or estimate CilGf) To(A;), (AFlOTo (A;) and Cp(A;). The most common To is 298 K. It is then possible to calculate (MiR)r:

(2.21)

which can often be simpl~fied to: (.:lGR)r = (j/ffJro - T(.lSVT.)

as long as I

V; Cp(A;) is

small. which is usually the case.

(2.22)

When (.1GR)r, and therefore K=. have been estimated. the "concentrations" of the different components at equilibrium should be specified. The following was defined: (2.19) with:

f; fugacity of component Ai in the mixture at equilibrium f;0 fugacity of component Ai in the standard state Fugacity f;0 is equal to 1 bar by definition of the standard state. Fugacity f;. also expressed in bars, can be evaluated either in the liquid or in the vapor state depending on the problem data (see Vol.2, Chapter 2). If the gas phase is involved: (2.23) Pyi = p, is the partial pressure of component Ai (to be expressed in bars) ~Y is the fugacity coefficient in the gas phase If the liquid phase is involved:

(2.24) where:

Pf Xi y;L

is the vapor pressure of component Ai (to be expressed in bars) is the mole fraction of component Ai in the liquid phase is the activity coefficient of component Ai in the liquid phase

When component Ai has no definite vapor pressure (supercritical state), the Henry relation is used to express f;L: (2.25) where 'Je; is the Henry constant for component i. The equilibrium between phases is obviously considered to have been achieved for these estimates and' therefore:

It is also possible to express the equilibrium constant in terms of mole fractions, partial pressures or molar concentrations:

KG = n(f;/f;°)Vi= n(Py;,Y/f;°)VI

=(P/1)r. v; n(+nv; Ky =n (¢Y)viKp

(2.26)

Note that Kp defined in this way is dimensionless and that at a relatively moderate pressure:

The following can also be written:

KG

=n(f;/ft)v; =nCI1x;y;Llfj)v; =n(P7/1)v; ncy;L)Y;Kx ..

nV

n.l n

x-

vt»,

vm

(2.27)

and. additionally since C; = ...!. = -'-' = --!.., vm being the average molar volume, the result is:

KG

=n (11/I)V; n (YiL)V; (v~)rV; K~

(2.28a)

for a liquid phase, and:

KG

=(P/I)~V; n (,n

V

;

(v~) !v; K8

(2.28b)

for a gas phase. Note that K~ and K8 are not cimensionless, as long as L v; * O.

~.~.2.3

Application Example

Use the example of isopropanol dehydrogenation:

a. Calculating the Variation in Standard Gibbs Energy Given that the reaction is carried out in the gas phase at 433 K at an absolute pressure of 1 bar. The TRC tables [37} give:

=

(tllif)298 (CH3-CO-oI:i) -217 150 j (tllii)298 (CH3-CHOH-CH:i) -272 295-J

=

(tllij)298

(Hv =0 (elements' heat of formation is zero at any temperature)

(.lGf)298 (CH3-CO-oI:i) =-152716 J

(ilG f)298 (CH3-CHOH-CH:i) =-173 385 J (ilGj)298

(Hv =0 (true at any temperature)

The following is calculated first of all:

(tlliR)298 = -21; ISO + 272 295 = +55 145 J then: (ilG R) 298 =-152 i16 + 173 385 =+20669 J

and:

~, = (.llipJ~ - (~GiD298 = +115.69 J/K (~ RJ298 298

C; (Ai)

Using Kobe et als

data [20), the following is then calculated:

(~GR)433 = 55 145 433

+

f

298

4.184(i.i83-1.887·1O-2T+ 1.3i5·1O-5~-0.3-l6·1O-8T3)dT-433x 115.69 433

- 433

f

[4.184 (7.783 - 1.88i·1Q-2T + 1.3i5·1O- 5T2 - 0.346.10-8 T3)] dT,fT

298

.

The result is: (j,G R) 433 =55 145 + 1 452 - 50095 - 2770 =+3732 J Using the simplified relation: (L\G R) 433 =(MfR) 298 - 433 (~R)298

would give: (L\G R) 433 =55 145 - 50095

=+5 050 J

b. Calculating the Equilibrium Constant KO (,1GR)-l33

=- R x 433 In KO

KO

The result is:

K _

and therefore:

p-

=0.35 0.35

nccpy)\';

For example the law of corresponding states and the general graph for determining fugacity coefficients like the one found in most thermodynamics reference works [17] are used to calculate C¢}")\'( An equation of state, such as the Soave Redlich Kwong or the Peng Robinson equations (see Vol. 2. Chapter 2) can also be used. In the case under consideratlon, since the pressure is low arid the temperature is moderate, the following is found:

and therefore: Kp

iO!!

KO

=0.35

c. Maximum Molar Extent If the initial system were made up of no moles of isopropanol and kno moles of inerts, the composition at.equlllbrlurn can be estimated which corresponds to the maximum molar extent Se' At equilibrium, by definition of Se' there are: ~

l;e nokno

moles moles ~e moles moles

of hydrogen of acetone of isopropanol of inerts.

C/I6pref' 2.

8AsJcPRINC6'l.ES GOVERNING CHEMICAL CHANGEs

25

Partial pressures at equilibrium will be: . 110-1: P. (isopropanol) = .• • p (1 + k)no + ;

CH,

CH,

CH,

CH CH. CH, ICH,

CH,

RH+ /

/

r< / '

0Q)

"':~: 11 bar ~'7bar

rK;7

"'3 bar

0.6

'i

c .2 ti

~ Q)

(5

0.4

:E

3 bar

0.2

7 bar 11 bar 15bar

480

520

560

600

T(°C)

Figutt: 4.4

Dehydrogenationequilibrium of n-neptane t?n-heptenes.

• Dehydrogenation of cyclohexanic naphthenes is highly influenced by the hydrogen partial pressure. For instance, as indicated in Figure 4.5, at 40 bar over 10% of the cyclohexane remains present at equilibrium. while at 5 bar total conversion is achieved from 500'C on. Replacement of the aromatic ring by alkyl groups at Iso-condltlons moves the equilibrium toward more thorough dehydrogenation: - low hydrogen pressures and high temperatures promote aromatics production. Light feeds require processing at higher temperatures in order to achieve the same conversions. • For n-paraffin dehydrocyclization, the equilibrium is more complex. Along with olefins (not shown in Figure -4.6), it involves isoparafflr:s. cyclohe..xanlc and cyclopentanic naphthenes. aromatics and hydrogen of course. To convert over 90% of the heptanes at 500'C, pressures lower than 10 bar are requtred,

116

ChaotlH 4.

C,.;TALY7'IC REF01IMlNG

, « + E z

~ •

-- O+=@+3H

Z

0.1

0.05

0.01

450

500

550

T(OC)

Figure

4.5

Influence oftemperature and pressure on the mole fractions of naphttienes present at equilibrium for hydrocarbons with 6 and 7 carbon atoms.

100

Q+H z

75 tf.

'0

~

50

C1H,6

==:

f1

T=500"C 25

o !t

e

+4H 2

o Figure 4.6

Variation in the mole fraction of C; hydrocarbons versus PHt

ChaP/!' 4.

CAW-YTlCREFoRMING

117

Figure 4.7 shows that at soooC with a total pressure of lO'bar C1 and C9 paraffins are transformed into aromatics but the conversion is limited for C6 and C7's: - dehydrocyclisation requires high temperatures which increase with the lowering of the number of carbon atoms of the feed.

100

III

75

0

~ E

2

III

'0 t!-

50

o

::E 25

400

450

500

550

' 600 T(OC)

'. F r;u.;iguree

I Influence of the number of carbon atoms on the mole fraction of aromatics at

~

equilibrium versus T'C loTPH~ = 10 bar. 6. C6: .. C7 ; 0 Ca; • C9' -

. fJ.2.3 Conclusions The thermodynamics of the desired reactions determines operating conditions: high temperature. around 5OO°C and hydrogen pressure as low as possible. Since the reaction produces hydrogen. the minimum pressure is determined by the desired aromatics conversion.

4.3.3 Catalysts [2. 3] of.3.3.l Type All current catalysts are deri ved from platinum on chlorinated alumina as introduced by VOP in l~49.

a. Platinum on Chlorinated Alumina A few hundred tons of this type of catalyst are still being used in the world. Table 4.9 gives its main characteristics.

Average values

Characteristics Support Impurities: alkalis + alkaline earths + iron (ppm) Specific surface area (m 2{g) Total pore volume (cm 3/g) Pore diameter (nm) Chlorine content (% wt) Platinum content (% wt) Extrudate or bead diameter (mm)

I' I

L

iThbi:l ~

Yo: alumina

~

-,

'.

.............~~~

o Figure 4.31

Various reaction rates versus hydrogen pressure.

Apparent activation energies (Fig. 4.32), or rather the thermal increments measured for these reactions are widely different. For hydrogenation it is around 42 kllmol, while for dehydrogenation it is approximately double. Since these reactions are rapid, they are often limited by diffusion. Isomerization activation energy is of about 100 kJ/mol, whereas the value for dehydrocycIization is around 150 kllmol and for cracking about 190 kl.rnol. This is why cracking prevails over the other reactions, thereby lowering the C5+ yield, when the monometallic catalyst ages and is consequently operated at a higher temperature. Coking, a complex reaction, has a high activation energy (150 kl/rnol), As a result, the higher the operating temperature, the faster deactivation occurs. All reaction rates involved in reforming vary according to the number of carbon atoms. For instance in dehydrocyclization, conversion of n-e 6 is very slow, approximately one-tenth of that or n-e 7, which in tum is around three times lower than n-e\(} (Fig.4.33). Likewise, cracking of n.(s is about four times slower than that of n.(lO'

b. Bimetallic Catalysts Bimetallic catalysts have a more complex behavior. In their range of operation, the reaction rate for dehydrocylization can reach a maximum depending on

CNQtw 4. C~TAJ.YTIC

Hydrogenation 42

450

510

~!

-

T(=C)

~ j/nRuence 01temperauueee reaction rate.

1.0

o C,-C,+C i _ 1 - CZ+C._ 2 """,-C 3+C'_3

o ~ 4.33

6

8 9 10 Number of carbon atoms

Influence of the number 01carbon atoms on reaction rate.

fiEFoRMING

139.

hydrogen pressure. Figure ·1.34 shows that for the dehydrocyclization reaction. platlnum-lridtum is the most active in the 10-30 bar range. platinum and platinum-rhenium are fairly close. but that platinum-tin is less active. At low hydrogen pressures. platinum-tin activity is the highest. except for platinum-iridium which is not used at low pressures for other reasons.

Pt-M catalyst

M: Pt= 1:3

T= 470°C PH(; = 2.2 bar M=lr

2:»

Ah Au

2

~

--+•• .0. ••

.--6--

x·

Re

~ >.

Pt alone .._.Sn

iii

iii 0 I

~

, s: a

•.•0(>.-

Ge In

-

Pb

-0-

.-4-

..§.

o

10

r---l;..-

20

--'

l Figure i

I 1

4.34

I:

L-J

Variation in dehydrocyclization rate versus catalyst formula and hydrogen pressure.

The low-pressure advantage of platinum-tin is also found in dehydrogenation of naphthenes (Fig. 4.35). Here Pt-Ir and Pt-Re tum out to be better in all cases than platinum alone. Finally. for the conversion of n-heptane into toluene and light products, the important point is that (Fig. 4.36) for platinum alone the ratio of dehydrocyclization (ED and cracking (ED activation energy is less than 1 whatever the pressure range. Any increase in temperature causes a loss of selectivity, In contrast. the ratio is slightly greater than 1 for platinum-iridium. The same is true but more clear cut for Pt-Re and Pt-Sn,and for Pt-Sn even at low

C~4. CATALYTIC

REFORW«:;

141

1

O+=!:@+3H2 2

5% coke

o Pt·Re • PI

• Pt·lr o Pt·Sn

30

~

X I

Ol "j

s: "0

20

S ~

"> U

-c

10

.... ... . '

.•....'

..' ..-

'

o Figure

4.35

Variation in dehydrogenation rote versus catalyst formula and hydrogen.

pressure.

temperature. This explains why bimetallic catalysts maintain their C5+ selectivity, or lose very little during runs, when operating temperatures are graduaJIy raised. Pseudo-kinetic models based on some fifty compounds and around a hundred reactions have been developed. They can be used to rationalize most observations. 4.3.3.7 Condusions Catalytic reforming catalysts activate a large number of reactions whose exo!endothermicity, thermodynamics and rate are Widely different. They are sensitive to impurities, so proper operation will require prior purification offeeds. They produce coke at a slow rate. which is nevertheless too fast to allow stable operation. The coking rate can be stgniflcantly reduced by the presence of hydrogen. As a result, it is advantageous to implement catalysts under the highest hydrogen pressure compatible with reaction thermodynamics and

Pi 8 for monometallic catalysts) by compressing and recycling part of the hydrogen produced during the reaction. It may contain some lmpurtties: water, hydrochloric acid and sometimes HzS, and needs to be dried, dehvdrochlorineted and desulfurized on specific adsorbers. Such installations are mainly used during start up but not only. , Finally, the catalyst produces light C1 to C.. hydrocarbons in addition to hydrogen, and they are separated from the reformate in a stabilization section. Depending on the operating pressure, separation systems of varying degrees of complexity will be implemented to get relatively pure hydrogen and high gasoline recovery: a simple separating drum, recontacting purification systems, ultra-cooling, etc.

A

B

~

I

I

\J

I I

I I I I I I

I I I

I I I

I I

I I I

I I I

I I I

I I I

R,~-R:!--r----R3-' I

I

c Vol %

R1 - : - R : ! - : - - - A3 - - - - - : I

Po=60 r-----,-~

I I I

No=30

I I

.

Aromatics

I

I I I I r J

I

Ao= 10

I

Paraffins I

Naphthenes

A. Furnace (FJ • F:. FJ ) and reactor (R/. R;:. R1} layout.. B. Variation in temperature in reactors RI • R:. RJ • C. t'oriation in effluent composition (paraffins-naphthenes-aromaticsJ in the reactors.

152

Cnaallff4. CATALYTIC REFOR.~· '.3

c: .2 ~

~

i5

~

~ ~

:i: >:

,..'"

c: