This action might not be possible to undo. Are you sure you want to continue?

**Copyright c 2010 by Nob Hill Publishing, LLC
**

1

General Mole Balance

R

j

V

Q

1

c

j1

Q

0

c

j0

Conservation of mass

¸

¸

¸

¸

¸

¸

rate of

accumulation

of component j

¸

¸

¸

¸

¸

¸

¸

=

rate of inﬂow

of component j

−

rate of outﬂow

of component j

+

¸

¸

¸

¸

¸

¸

rate of generation

of component j by

chemical reactions

¸

¸

¸

¸

¸

¸

¸

(1)

2

General Mole Balance

R

j

V

Q

1

c

j1

Q

0

c

j0

Conservation of mass

d

dt

V

c

j

dV = Q

0

c

j0

−Q

1

c

j1

+

V

R

j

dV (2)

3

General Mole Balance

d

dt

V

c

j

dV = Q

0

c

j0

−Q

1

c

j1

+

V

R

j

dV

• Equation 2 applies to every chemical component in the system, j =

1, 2, . . . , n

s

, including inerts, which do not take place in any reactions.

• Assuming component j enters and leaves the volume element only by con-

vection with the inﬂow and outﬂow streams, i.e. neglecting diffusional ﬂux

through the boundary of the volume element due to a concentration gradient.

• The diffusional ﬂux will be considered during the development of the material

balance for the packed-bed reactor.

4

Rate expressions

• To solve the reactor material balance, we require an expression for the pro-

duction rates, R

j

R

j

=

¸

i

ν

ij

r

i

• Therefore we require r

i

as a function of c

j

• This is the subject of chemical kinetics, Chapter 5

• Here we use common reaction-rate expressions without derivation

5

The Batch Reactor

R

j

• The batch reactor is assumed well stirred

• Let the entire reactor contents be the reactor volume element

6

Batch Reactor

d

dt

V

c

j

dV = Q

0

c

j0

−Q

1

c

j1

+

V

R

j

dV

• Because the reactor is well stirred, the integrals in Equation 2 are simple to

evaluate,

V

R

c

j

dV = c

j

V

R

V

R

R

j

dV = R

j

V

R

• The inﬂow and outﬂow stream ﬂowrates are zero, Q

0

= Q

1

= 0.

d

c

j

V

R

dt

= R

j

V

R

(3)

7

Reactor Volume

• Equation 3 applies whether the reactor volume is constant or changes during

the reaction.

• If the reactor volume is constant (liquid-phase reactions)

dc

j

dt

= R

j

(4)

• Use Equation 3 rather than Equation 4 if the reactor volume changes signiﬁ-

cantly during the course of the reaction.

8

Analytical Solutions for Simple Rate Laws

• In general the material balance must be solved numerically.

• If the reactor is isothermal, we have few components, the rate expressions

are simple, then analytical solutions of the material balance are possible.

• We next examine derive analytical solutions for some classic cases.

9

First-order, irreversible

• Consider the ﬁrst-order, irreversible reaction

A

k

→ B, r = kc

A

(5)

• The material balance for a constant-volume reactor gives

dc

A

dt

= −kc

A

(6)

• Watch the sign!

10

First-order, irreversible

• We denote the initial concentration of A as c

A0

,

c

A

(t) = c

A0

, t = 0

• The solution to the differential equation with this boundary condition is

c

A

= c

A0

e

−kt

(7)

11

First-order, irreversible

12

0

0.2

0.4

0.6

0.8

1

0 1 2 3 4 5

t

k = 0.5

k = 1

k = 2

k = 5

c

A

c

A0

13

First-order, irreversible

• The A concentration decreases exponentially fromits initial value to zero with

increasing time.

• The rate constant determines the shape of this exponential decrease. Rear-

ranging Equation 7 gives

ln(c

A

/c

A0

) = −kt

14

First-order, irreversible

0.001

0.01

0.1

1

0 1 2 3 4 5

t

k = 0.5

k = 1

k = 2

k = 5

c

A

c

A0

• One can get an approximate value of the rate constant from the slope of the

straight line.

15

• This procedure is a poor way to determine a rate constant and should be

viewed only as a rough approximation (Chapter 9).

16

First-order, irreversible

The B concentration is determined from the A concentration.

1. Solve the material balance for component B,

dc

B

dt

= R

B

= kc

A

(8)

with the initial condition for B, c

B

(0) = c

B0

2. Note that the sum of c

A

and c

B

is constant.

d(c

A

+c

B

)

dt

= R

A

+R

B

= 0

Therefore, c

A

+c

B

is a constant.

17

First-order, reversible

• The value is known at t = 0,

c

A

+c

B

= c

A0

+c

B0

• So we have an expression for c

B

c

B

= c

A0

+c

B0

−c

A

(9)

c

B

= c

B0

+c

A0

(1 −e

−kt

) (10)

18

First-order, reversible

• Consider now the same ﬁrst-order reaction, but assume it is reversible

A

k

1

k

−1

B (11)

• The reaction rate is r = k

1

c

A

−k

−1

c

B

.

• The material balances for A and B are now

dc

A

dt

= −r = −k

1

c

A

+k

−1

c

B

c

A

(0) = cA0

dc

B

dt

= r = k

1

c

A

−k

−1

c

B

c

B

(0) = c

B0

• Notice that c

A

+c

B

= c

A0

+c

B0

remains constante

19

First-order, reversible

• Eliminate c

B

in the material balance for A gives

dc

A

dt

= −k

1

c

A

+k

−1

(c

A0

+c

B0

−c

A

) (12)

• How do we want to solve this one?

• Particular solution and homogeneous solution (see text)

• Laplace transforms (control course)

• Separation!

20

First-order, reversible

dc

A

dt

= ac

A

+b

c

A

c

A0

dc

A

ac

A

+b

=

t

0

dt

1

a

ln(ac

A

+b)

¸

¸

¸

¸

c

A

c

A0

= t

c

A

= c

A0

e

at

−

b

a

(1 −e

at

)

Substitute a = −(k

1

+k

−1

), b = k

−1

(c

A0

+c

B0

)

21

First-order, reversible

c

A

= c

A0

e

−(k

1

+k

−1

)t

+

k

−1

k

1

+k

−1

(c

A0

+c

B0

)

¸

1 −e

−(k

1

+k

−1

)t

¸

(13)

• The B concentration can be determined by switching the roles of A and B and

k

1

and k

−1

in Reaction 11, yielding

c

B

= c

B0

e

−(k

1

+k

−1

)t

+

k

1

k

1

+k

−1

(c

A0

+c

B0

)

¸

1 −e

−(k

1

+k

−1

)t

¸

(14)

22

First-order, reversible

0

0.2

0.4

0.6

0.8

1

0 1 2 3 4 5

t

c

B

(t)

c

A

(t)

c

As

c

Bs

c

Figure 1: First-order, reversible kinetics in a batch reactor, k

1

= 1, k

−1

= 0.5,

c

A0

= 1, c

B0

= 0.

23

Nonzero steady state

• For the reversible reaction, the concentration of A does not go to zero.

• Taking the limit t → ∞ in Equation 13 gives

c

As

=

k

−1

k

1

+k

−1

(c

A0

+c

B0

)

in which c

As

is the steady-state concentration of A.

24

Nonzero steady state

• Deﬁning K

1

= k

1

/k

−1

allows us to rewrite this as

c

As

=

1

1 +K

1

(c

A0

+c

B0

)

• Performing the same calculation for c

B

gives

c

Bs

=

K

1

1 +K

1

(c

A0

+c

B0

)

25

Second-order, irreversible

• Consider the irreversible reaction

A

k

→ B (15)

in which the rate expression is second order, r = kc

2

A

.

• The material balance and initial condition are

dc

A

dt

= −kc

2

A

, c

A

(0) = c

A0

(16)

• Our ﬁrst nonlinear differential equation.

26

Second-order, irreversible

• Separation works here

dc

A

c

2

A

= −kdt

c

A

c

A0

dc

A

c

2

A

= −k

t

0

dt

1

c

A0

−

1

c

A

= −kt

• Solving for c

A

gives

c

A

=

1

c

A0

+kt

¸

−1

(17)

• Check that this solution satisﬁes the differential equation and initial condition

27

Second-order, irreversible

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 1 2 3 4 5

ﬁrst order

second order

c

A

c

A0

• The second-order reaction decays more slowly to zero than the ﬁrst-order

reaction.

28

Another second-order, irreversible

A+B

k

→ C r = kc

A

c

B

(18)

• The material balance for components A and B are

dc

A

dt

= −r = −kc

A

c

B

(19)

dc

B

dt

= −r = −kc

A

c

B

(20)

• Subtract B’s material balance from A’s to obtain

d(c

A

−c

B

)

dt

= 0

29

Another second-order, irreversible

• Therefore, c

A

−c

B

is constant, and

c

B

= c

A

−c

A0

+c

B0

(21)

• Substituting this expression into the material balance for A yields

dc

A

dt

= −kc

A

(c

A

−c

A0

+c

B0

)

• This equation also is separable and can be integrated to give (you should

work through these steps),

c

A

= (c

A0

−c

B0

)

¸

1 −

c

B0

c

A0

e

(c

B0

−c

A0

)kt

−1

, c

A0

≠ c

B0

(22)

30

Another second-order, irreversible

• Component B can be computed from Equation 21, or by switching the roles

of A and B in Reaction 18, giving

c

B

= (c

B0

−c

A0

)

¸

1 −

c

A0

c

B0

e

(c

A0

−c

B0

)kt

−1

(23)

• What about component C? C’s material balance is

dc

C

dt

= kc

A

c

B

and therefore, d(c

A

+c

C

)/dt = 0. The concentration of C is given by

c

C

= c

A0

−c

A

+c

C0

31

Another second-order, irreversible

• Notice that if c

A0

> c

B0

(Excess A), the steady state

c

As

= c

A0

−c

B0

c

Bs

= 0

c

Cs

= c

B0

+c

C0

• For c

B0

> c

A0

(Excess B), the steady state is

c

As

= 0

c

Bs

= c

B0

−c

A0

c

Cs

= c

A0

+c

C0

32

nth-order, irreversible

The nth-order rate expression r = kc

n

A

33

0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

5

0 0.5 1 1.5 2 2.5 3

c

A

r

n = 3

2

1

1/2

0

34

nth-order, irreversible

A

k

→ B r = kc

n

A

dc

A

dt

= −r = −kc

n

A

• This equation also is separable and can be rearranged to

dc

A

c

n

A

= −kdt

• Performing the integration and solving for c

A

gives

c

A

=

¸

c

−n+1

A0

+(n−1)kt

¸

1

−n+1

, n ≠ 1

35

nth-order, irreversible

• We can divide both sides by c

A0

to obtain

c

A

c

A0

= [1 +(n−1)k

0

t]

1

−n+1

, n ≠ 1 (24)

in which

k

0

= kc

n−1

A0

has units of inverse time.

36

nth-order, irreversible

0

0.2

0.4

0.6

0.8

1

0 1 2 3 4 5

t

1

2

3

4

n = 5

c

A

c

A0

• The larger the value of n, the more slowly the A concentration approaches

zero at large time.

37

nth-order, irreversible

• Exercise care for n < 1, c

A

reaches zero in ﬁnite time.

0

0.2

0.4

0.6

0.8

1

0 0.5 1 1.5 2

t

c

A

c

A0 n = 2

1

1/2

0

−1/2 −1 −2

38

Negative order, inhibition

• For n < 0, the rate decreases with increasing reactant concentration; the

reactant inhibits the reaction.

0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

5

0 0.5 1 1.5 2 2.5 3

c

A

r

0

−1/2

−1

n=−2

39

Negative order, inhibition

• Inhibition reactions are not uncommon, but watch out for small concentra-

tions. Notice the rate becomes unbounded as c

A

approaches zero, which is

not physically realistic.

• When using an ODE solver we may modify the right-hand sides of the material

balance

dc

A

dt

=

−kc

n

A

, c

A

> 0

0, c

A

= 0

• Examine the solution carefully if the concentration reaches zero.

40

Two reactions in series

• Consider the following two irreversible reactions,

A

k

1

→ B (25)

B

k

2

→ C (26)

• Reactant A decomposes to form an intermediate B that can further react to

form a ﬁnal product C.

• Let the reaction rates be given by simple ﬁrst-order rate expressions in the

corresponding reactants,

r

1

= k

1

c

A

r

2

= k

2

c

B

41

Two reactions in series

• The material balances for the three components are

dc

A

dt

= R

A

= −r

1

= −k

1

c

A

dc

B

dt

= R

B

= r

1

−r

2

= k

1

c

A

−k

2

c

B

dc

C

dt

= R

C

= r

2

= k

2

c

B

• The material balance for component A can be solved immediately to give

c

A

= c

A0

e

−k

1

t

as before.

42

Two reactions in series – B

• The material balance for B becomes

dc

B

dt

+k

2

c

B

= k

1

c

A0

e

−k

1

t

Oops, not separable, now what?

• Either Laplace transform or particular solution, homogeneous equation ap-

proach produces

c

B

= c

B0

e

−k

2

t

+c

A0

k

1

k

2

−k

1

¸

e

−k

1

t

−e

−k

2

t

¸

, k

1

≠ k

2

(27)

43

Two reactions in series – C

• To determine the C concentration, notice from the material balances that

d(c

A

+c

B

+c

C

)/dt = 0. Therefore, c

C

is

c

C

= c

A0

+c

B0

+c

C0

−c

A

−c

B

44

Two reactions in series

0

0.2

0.4

0.6

0.8

1

0 1 2 3 4 5

t

c

A

(t)

c

B

(t)

c

C

(t)

c

Figure 2: Two ﬁrst-order reactions in series in a batch reactor, c

A0

= 1,

c

B0

= c

C0

= 0, k

1

= 2, k

2

= 1.

45

Two reactions in parallel

• Consider next two parallel reactions of A to two different products, B and C,

A

k

1

→ B (28)

A

k

2

→ C (29)

• Assume the rates of the two irreversible reactions are given by r

1

= k

1

c

A

and

r

2

= k

2

c

A

.

46

Two reactions in parallel

• The material balances for the components are

dc

A

dt

= R

A

= −r

1

−r

2

= −k

1

c

A

−k

2

c

A

dc

B

dt

= R

B

= r

1

= k

1

c

A

dc

C

dt

= R

C

= r

2

= k

2

c

A

47

Two reactions in parallel

• The material balance for A can be solved directly to give

c

A

= c

A0

e

−(k

1

+k

2

)t

(30)

• Substituting c

A

(t) into B’s material balance gives

dc

B

dt

= k

1

c

A0

e

−(k

1

+k

2

)t

• This equation is now separable and can be integrated directly to give

c

B

= c

B0

+c

A0

k

1

k

1

+k

2

1 −e

−(k

1

+k

2

)t

(31)

48

Two reactions in parallel

• Finally, component C can be determined from the condition that c

A

+c

B

+c

C

is constant or by switching the roles of B and C, and k

1

and k

2

in Equation 31,

c

C

= c

C0

+c

A0

k

2

k

1

+k

2

1 −e

−(k

1

+k

2

)t

(32)

49

Two reactions in parallel

0

0.2

0.4

0.6

0.8

1

0 0.5 1 1.5 2 2.5 3

t

c

A

(t)

c

B

(t)

c

C

(t)

c

Figure 3: Two ﬁrst-order reactions in parallel in a batch reactor, c

A0

= 1,

c

B0

= c

C0

= 0, k

1

= 1, k

2

= 2.

50

Two reactions in parallel

• Notice that the two parallel reactions compete for the same reactant, A

• The rate constants determine which product is favored

• Large values of k

1

/k

2

favor the formation of component B compared to C and

vice versa

51

Conversion, Yield, Selectivity

There are several ways to deﬁne selectivity, yield and conversion, so be clear

about the deﬁnition you choose.

Point selectivity: The point (or instantaneous) selectivity is the ratio of the production rate of

one component to the production rate of another component.

Overall selectivity: The overall selectivity is the ratio of the amount of one component pro-

duced to the amount of another component produced.

Yield: The yield of component j is the fraction of a reactant that is converted into component

j.

Conversion: Conversion is normally deﬁned to be the fraction of a component that has been

converted to products by the reaction network. Conversion has several deﬁnitions and

conventions. It is best to state the deﬁnition in the context of the problem being solved.

52

The Continuous-Stirred-Tank Reactor (CSTR)

Q

c

j

Q

f

c

jf

R

j

Writing the material balance for this reactor gives

d

c

j

V

R

dt

= Q

f

c

jf

−Qc

j

+R

j

V

R

, j = 1, . . . , n

s

(33)

53

CSTR – Constant Density

• If the reactor volume is constant and the volumetric ﬂowrates of the inﬂow

and outﬂow streams are the same, Equation 33 reduces to

dc

j

dt

=

1

τ

(c

jf

−c

j

) +R

j

(34)

• The parameter

τ = V

R

/Q

f

is called the mean residence time of the CSTR.

• We refer to this balance as the constant-density case. It is often a good ap-

proximation for liquid-phase reactions.

54

CSTR – Steady State

• The steady state of the CSTR is described by setting the time derivative in

Equation 33 to zero,

0 = Q

f

c

jf

−Qc

j

+R

j

V

R

(35)

• Conversion of reactant j is deﬁned for a steady-state CSTR as follows

x

j

=

Q

f

c

jf

−Qc

j

Q

f

c

jf

(steady state) (36)

• One can divide Equation 35 through by V

R

to obtain for the constant-density

case

c

j

= c

jf

+R

j

τ (steady state, constant density) (37)

55

Transient behavior of the CSTR

• Consider a ﬁrst-order, liquid-phase reaction in an isothermal CSTR

A

k

→ 2B r = kc

A

the feed concentration of A is c

Af

= 2 mol/L, the residence time of the reactor

is τ = 100 min, and the rate constant is k = 0.1 min

−1

.

1. Find the steady-state concentration of A in the efﬂuent for the given feed.

2. Plot the concentration of A versus time for constant feed concentration c

Af

= 2 mol/L if

the reactor is initially ﬁlled with an inert so c

A0

= 0 mol/L.

3. Plot the concentration of A versus time for constant feed concentration c

Af

= 2 mol/L if

the reactor is initially ﬁlled with feed so c

A0

= 2 mol/L.

56

Transient CSTR Solution. Part 1

• Liquid phase: assume the ﬂuid density is constant.

c

A

= c

Af

+R

A

τ

• Substituting the production rate R

A

= −kc

A

and solving for c

A

gives the

steady-state concentration

c

As

=

c

Af

1 +kτ

• Substituting in the numerical values gives

c

As

=

2 mol/L

1 +(0.1 min

−1

)(100 min)

= 0.182 mol/L

57

Parts 2 and 3

dc

A

dt

=

1

τ

c

Af

−c

A

−kc

A

(38)

c

A

(0) = c

A0

• This equation is also separable. The analytical solution is

c

A

(t) = c

A0

e

−(1/τ+k)t

+

c

Af

1 +kτ

¸

1 −e

−(1/τ+k)t

¸

(39)

58

Parts 2 and 3

0

0.5

1

1.5

2

0 10 20 30 40 50 60 70 80

t (min)

c

A

(

t

)

(

m

o

l

/

L

)

c

A0

= 0

c

A0

= 2

c

As

= 0.182

• Both solutions converge to the same steady-state even though the starting

conditions are quite different.

59

Phenol production in a CSTR

• Consider the reaction of cumene hydroperoxide (CHP) to phenol and acetone

(C

6

H

5

)C(CH

3

)

2

OOH → (C

6

H

5

)OH+(CH

3

)

2

CO (40)

CHP → phenol + acetone (41)

• The reaction is pseudo-ﬁrst-order

r = kc

CHP

• Find the reactor volume to achieve 85% conversion of CHP at steady state.

The ﬂowrate into the reactor is Q

f

= 26.9 m

3

/hr and k = 4.12 hr

−1

.

60

Phenol production

• Liquids at 85

◦

C, so assume constant density and Q = Q

f

.

c

A

= c

Af

+R

A

τ

R

A

= −kc

A

, and solving for the CHP concentration gives

c

A

=

c

Af

1 +kτ

(42)

61

Phenol production

• The conversion of CHP (for Q = Q

f

) is

x

A

=

c

Af

−c

A

c

Af

= 1 −

c

A

c

Af

x

A

=

kτ

1 +kτ

• Solving this equation for τ gives

τ =

1

k

x

A

1 −x

A

62

Phenol production

• Substituting the relation τ = V

R

/Q

f

and solving for V

R

gives

V

R

=

Q

f

x

A

k(1 −x

A

)

• Substituting in the known values gives the required CSTR volume

V

R

=

(26.9 m

3

/hr)(0.85)

(4.12 hr

−1

)(0.15)

= 37 m

3

63

The Semi-Batch Reactor

• The semi-batch reactor is a cross between the batch reactor and CSTR.

• The semi-batch reactor is initially charged with reactant, like the batch reac-

tor, but allows a feed addition policy while the reaction takes place, like the

CSTR.

• Normally there is no outﬂow stream.

64

The Semi-Batch Reactor

• We set Q = 0 in the CSTR material balance to obtain

d

c

j

V

R

dt

= Q

f

c

jf

+R

j

V

R

, j = 1, . . . , n

s

(43)

• One may choose to operate a semi-batch reactor to control the reaction rate

or heat release during reaction by slowly adding one of the reactants in the

feed stream.

• Compared to the batch reactor, the semi-batch reactor provides more com-

plete use of the reactor volume in reactions such as polymerizations that

convert from lower density to higher density during the course of the reac-

tion.

65

Volume Change Upon Reaction

d

c

j

V

R

dt

= Q

f

c

jf

−Qc

j

+R

j

V

R

(44)

• Equation 44 covers both the batch, CSTR and semi-batch reactors, depending

on how we specify Q

f

and Q.

• If we multiply Equation 44 by the molecular weight of species j and sum over

all species we obtain,

d(

¸

j

c

j

M

j

V

R

)

dt

= Q

f

¸

j

c

jf

M

j

−Q

¸

j

c

j

M

j

+

¸

j

R

j

M

j

V

R

(45)

• The term

¸

j

c

j

M

j

is simply the mass density of the reactor contents, which

66

we denote ρ

ρ =

n

s

¸

j=1

c

j

M

j

(46)

67

Volume Change Upon Reaction

• The term

¸

j

c

jf

M

j

is the mass density of the feedstream, ρ

f

.

• We knowthat conservation of mass in chemical reactions implies

¸

j

R

j

M

j

= 0

(see Chapter 2). Substitution into Equation 45 leads to

d(ρV

R

)

dt

= Q

f

ρ

f

−Qρ (47)

• Equation 47 is clearly a total mass balance, in which the total mass in the

reactor changes in time due to the inﬂow and outﬂow of mass.

• Notice that chemical reactions play no role in the total mass balance.

68

Equation of state for the mixture

• If we have a single-phase system at equilibrium, the intensive variables c

j

, T,

P, completely specify all intensive variables of the system.

• In this chapter we consider T and P to be known, ﬁxed quantities. Therefore,

the density of the reaction mixture, which is an intensive variable, is known

if the c

j

are known.

• This relationship is one form of the equation of state for the mixture

ρ =

`

f(T, P, c

1

, c

2

, . . . , c

n

s

)

• Substituting the deﬁnition of density, we can express the equation of state

as

f(c

1

, c

2

, . . . , c

n

s

) = 0

69

Equation of state for the mixture

• For example, we could express the equation of state in terms of the partial

molar volumes as

¸

j

c

j

V

j

= 1

in which V

j

is the partial molar volume of component j in the mixture.

• The partial molar volumes are functions of T, P and c

j

.

70

Equation of state for the mixture — Ideal mixture

• If we assume an ideal mixture, this reduces to

¸

j

c

j

V

◦

j

= 1, ideal mixture

in which V

◦

j

is the speciﬁc volume of pure component j, which is a function

of only T and P.

• We assume that a thermodynamic equation of state is valid even when the

reactor is not at equilibrium.

71

Constant density

• Because the reaction mixture density, ρ, is independent of composition, it

does not vary with time either and we can set it to the feed value,

ρ = ρ

f

(48)

• The total mass balance then reduces to

dV

R

dt

= Q

f

−Q (49)

which is sometimes referred to as a “volume balance.”

• This terminology should be avoided.

72

Constant density

dV

R

dt

= Q

f

−Q

• Batch reactor. For the batch reactor, Q = Q

f

= 0. We can therefore conclude

from Equation 49 that a batch reactor with constant density has constant

volume.

• CSTR (dynamic and steady state). If the outﬂow of the CSTR is regulated so

that the CSTR has constant volume, then we can conclude from Equation 49

that Q = Q

f

.

• Semi-batch reactor. In the semi-batch reactor, the reactor is ﬁlled during

operation so Q

f

is speciﬁed and positive for some time and Q = 0. The

solution to Equation 49 then determines the change in volume of the reactor

during the ﬁlling operation.

73

Nonconstant density

Unknowns.

• In the general case, consider the following variables to fully determine the

state of the reactor: T, P, n

j

, V

R

.

• We also require the value of Qto specify the right-hand sides of the material

balances.

• The set of unknowns is n

j

, V

R

, Q.

• We therefore have n

s

+2 unknowns.

Equations.

• We have the n

s

equations from the component mole balances.

• The equation of state provides one additional equation.

• The ﬁnal equation is provided by a statement of reactor operation.

74

Nonconstant density – reactor operation

1. Constant-volume reactor. The constant-volume reactor can be achieved by

allowing overﬂow of the reactor to determine ﬂowrate out of the reactor. In

this situation, V

R

is speciﬁed as the additional equation.

2. Constant-mass reactor. The constant-mass reactor can be achieved if a dif-

ferential pressure measurement is used to control the ﬂowrate out of the

reactor and the reactor has constant cross-sectional area. In this situation

ρV

R

is speciﬁed as the additional equation.

3. Flowrate out of the reactor is speciﬁed. This type of operation may be

achieved if the ﬂowrate out of the reactor is controlled by a ﬂow controller.

In this case Q(t) is speciﬁed. A semi-batch reactor is operated in this way

with Q = 0 until the reactor is ﬁlled with the reactants.

75

Nonconstant density – reactor operation

• See the text for the derivation.

dV

R

dt

= Q

f

¸

j

f

j

c

jf

¸

j

f

j

c

j

−Q+

¸

i

∆f

i

r

i

V

R

¸

j

f

j

c

j

(50)

• in which f

j

is

f

j

=

∂f

∂c

j

• and ∆f

i

is

∆f

i

=

¸

j

ν

ij

f

j

=

¸

j

ν

ij

∂f

∂c

j

which is a change in a derivative property upon reaction.

76

Nonconstant density – idea mixture

• For the ideal mixture we have f(c

j

) =

¸

j

c

j

V

◦

j

−1 = 0.

f

j

= V

◦

j

the pure component speciﬁc volumes

• The ∆f

i

are given by

∆f

i

= ∆V

◦

i

the change in speciﬁc volume upon reaction i.

• So the reactor volume can be calculated from

dV

R

dt

= Q

f

−Q+

¸

i

∆V

◦

i

r

i

V

R

77

Nonconstant density

Unknowns (n

s

+2): V

R

, Q, n

j

, j = 1, . . . , n

s

Component balances:

dn

j

dt

= Q

f

c

jf

−Qc

j

+R

j

V

R

, j = 1, . . . , n

s

Deﬁned quantities: n

j

= c

j

V

R

ρ =

¸

j

c

j

M

j

∆V

◦

i

=

¸

j

ν

ij

V

◦

j

(i) constant density: ρ = ρ

0

(ii) ideal mixture:

¸

j

c

j

V

◦

j

= 1

1. vol V

R

= V

R0

Q = Q

f

V

R

= V

R0

Q = Q

f

+

¸

i

∆V

◦

i

r

i

V

R

2. mass V

R

= V

R0

Q = Q

f

dV

R

dt

= Q

f

(1 −ρ

f

/ρ) +

¸

i

∆V

◦

i

r

i

V

R

Q = Q

f

ρ

f

/ρ

3. Q

dV

R

dt

= Q

f

−Q Q speciﬁed

dV

R

dt

= Q

f

−Q+

¸

i

∆V

◦

i

r

i

V

R

Q speciﬁed

Table 1: Reactor balances for constant-density and ideal-mixture assumptions.

78

Nonconstant density

Unknowns (n

s

+2): V

R

, Q, n

j

, j = 1, . . . , n

s

Component balances:

dn

j

dt

= Q

f

c

jf

−Qc

j

+R

j

V

R

, j = 1, . . . , n

s

Deﬁned quantities: n

j

= c

j

V

R

ρ =

¸

j

c

j

M

j

∆f

i

=

¸

j

ν

ij

∂f

∂c

j

Equation of state: f(c

1

, c

2

, . . . , c

n

s

) = 0

DAEs ODEs

1. vol V

R

= V

R0

f(c

j

) = 0 V

R

= V

R0

Q = Q

f

¸

j

f

j

c

jf

¸

j

f

j

c

j

+

¸

i

∆f

i

r

i

V

R

¸

j

f

j

c

j

2. mass ρV

R

= ρ

0

V

R0

f(c

j

) = 0

dV

R

dt

= Q

f

¸

j

f

j

c

jf

¸

j

f

j

c

j

−Q+

¸

i

∆f

i

r

i

V

R

¸

j

f

j

c

j

Q = Q

f

ρ

f

/ρ

3. Q Q speciﬁed f(c

j

) = 0

dV

R

dt

= Q

f

¸

j

f

j

c

jf

¸

j

f

j

c

j

−Q+

¸

i

∆f

i

r

i

V

R

¸

j

f

j

c

j

Q speciﬁed

Table 2: Reactor balances for general equation of state.

79

Semi-batch polymerization

• Consider a solution polymerization reaction, which can be modeled as a ﬁrst-

order, irreversible reaction

M

k

→ P r = kc

M

• A 20 m

3

semi-batch reactor is initially charged with solvent and initiator to

half its total volume.

• A pure monomer feed is slowly added at ﬂowrate Q

f0

= 1 m

3

/min to ﬁll the

reactor in semi-batch operation to control the heat release.

80

Semi-batch polymerization

• Consider two cases for the subsequent reactor operation.

1. The monomer feed is shut off and the reaction goes to completion.

2. The monomer feed is adjusted to keep the reactor ﬁlled while the reaction

goes to completion.

• Calculate the total polymer mass production, and the percentage increase in

polymer production achieved in the second operation.

81

The physical properties

• You may assume an ideal mixture

• The densities of monomer and polymer are

ρ

M

= 800 kg/m

3

ρ

P

= 1100 kg/m

3

• The monomer molecular weight is M

M

= 100 kg/kmol

• The rate constant is k = 0.1 min

−1

.

82

Semi-batch polymerization

• While the reactor is ﬁlling, the monomer mole balance is

d(c

M

V

R

)

dt

= Q

f0

c

Mf

−kc

M

V

R

in which c

Mf

= ρ

M

/M

M

is given, and Q

f

= Q

f0

is constant during the ﬁlling

operation.

• We denote the total number of moles of monomer by M = c

M

V

R

, and can

write the monomer balance as

dM

dt

= Q

f0

c

Mf

−kM (51)

M(0) = 0

83

Semi-batch polymerization

• For an ideal mixture, the volume is given by

dV

R

dt

= Q

f0

+∆VkM (52)

V

R

(0) = 10 m

3

in which ∆V = (1/ρ

P

−1/ρ

M

) M

M

84

The polymer mass production

• To compute the polymer mass, we note from the stoichiometry that the mass

production rate of polymer

`

R

P

is

`

R

P

= −R

M

M

M

• The mass balance for total polymer

`

P is given by

d

`

P

dt

=

`

R

p

V

R

= kc

M

M

M

V

R

= (kM

M

)M (53)

85

Semi-batch polymerization

• The text solves this problem analytically. Instead, let’s solve it numerically.

• Let t

1

be the time that the reactor ﬁlls.

• We need an ODE solver that is smart enough to stop when the reactor ﬁlls,

because we do not know this time t

1

. The ODE solver needs to ﬁnd it for us.

• dasrt is an ODE solver with the added capability to ﬁnd the time at which

some event of interest occurs.

86

Finding the time for ﬁlling the reactor

• The ODE solver ﬁnds the time at which V

R

= 20 m

3

t

1

= 11.2 min

• Note the reactor would have ﬁlled in 10 min if the density were constant.

• The extra time reﬂects the available volume created by converting some of

the monomer to polymer during ﬁlling.

• After t

1

we consider the two operations.

87

Operation 1.

• In the ﬁrst operation, Q

f

= 0 after t

1

.

• Notice the reactor volume decreases after t

1

because ∆V is negative.

88

Semi-batch polymerization

89

0

5

10

15

20

0 10 20 30 40 50

time (min)

V

R

(

m

3

)

2

1

Figure 4: Semi-batch reactor volume for primary monomer addition (operation

1) and primary plus secondary monomer additions (operation 2).

90

-0.2

0

0.2

0.4

0.6

0.8

1

0 10 20 30 40 50

time (min)

Q

f

(

m

3

/

m

i

n

)

2

1

Figure 5: Semi-batch reactor feed ﬂowrate for primary monomer addition (op-

eration 1) and primary plus secondary monomer additions (operation 2).

91

0

1000

2000

3000

4000

5000

6000

0 10 20 30 40 50

time (min)

m

o

n

o

m

e

r

(

k

g

)

2

1

Figure 6: Semi-batch reactor monomer content for primary monomer addition

(operation 1) and primary plus secondary monomer additions (operation 2).

92

0

2000

4000

6000

8000

10000

12000

0 10 20 30 40 50

time (min)

p

o

l

y

m

e

r

(

k

g

)

2

1

Figure 7: Semi-batch reactor polymer content for primary monomer addition

(operation 1) and primary plus secondary monomer additions (operation 2).

93

Operation 2.

• Because the reactor volume is constant, we can solve Equation 52 for the feed

ﬂowrate during the secondary monomer addition

Q

f

= −∆VkM

• Operation 2 is also shown in the ﬁgures.

• Notice the ﬁnal polymer production is larger in operation 2 because of the

extra monomer addition.

94

Polymer production rate

• We can perform an independent, simple calculation of the total polymer in

operation 2. Useful for debugging the computation.

• In operation 2, 10 m

3

of polymer are produced because in an ideal mixture,

the volumes are additive. Therefore

`

P

2

= (V

R

−V

R0

)ρ

P

= 10 m

3

×1100 kg/m

3

= 11000 kg

in good agreement with the long-time solution for operation 2.

• The increase in production rate is

`

P

2

−

`

P

1

`

P

1

×100% = 22.5%

95

• By using the volume of the reactor more efﬁciently, the total polymer produc-

tion increases 22.5%.

96

The Plug-Flow Reactor (PFR)

• Plug ﬂow in a tube is an ideal-ﬂow assumption in which the ﬂuid is well mixed

in the radial and angular directions.

• The ﬂuid velocity is assumed to be a function of only the axial position in the

tube.

• Plug ﬂow is often used to approximate ﬂuid ﬂow in tubes at high Reynolds

number. The turbulent ﬂow mixes the ﬂuid in the radial and angular direc-

tions.

• Also in turbulent ﬂow, the velocity proﬁle is expected to be reasonably ﬂat in

the radial direction except near the tube wall.

97

Thin Disk Volume Element

Given the plug-ﬂowassumption, it is natural to take a thin disk for the reactor

volume element

Q

c

j

Q(z +∆z)

c

j

(z +∆z)

Q(z)

c

j

(z)

Q

f

c

jf

z

R

j

∆V

z +∆z

. .. .

98

Thin Disk Volume Element

• Expressing the material balance for the volume element

∂

c

j

∆V

∂t

= c

j

Q

¸

¸

¸

z

− c

j

Q

¸

¸

¸

z+∆z

+R

j

∆V (54)

• Dividing the above equation by ∆V and taking the limit as ∆V goes to zero

yields,

∂c

j

∂t

. .. .

accumulation

= −

∂

c

j

Q

∂V

. .. .

convection

+ R

j

....

reaction

(55)

99

Length or volume as independent variable

• If the tube has constant cross section, A

c

, then velocity, v, is related to vol-

umetric ﬂowrate by v = Q/A

c

, and axial length is related to tube volume by

z = V/A

c

,

• Equation 55 can be rearranged to the familiar form [1, p.584]

∂c

j

∂t

= −

∂

c

j

v

∂z

+R

j

(56)

100

Steady-State Operation

• Setting the time derivative in Equation 55 to zero gives,

d(c

j

Q)

dV

= R

j

(57)

• The product c

j

Q = N

j

is the total molar ﬂow of component j. One also can

express the PFR mole balance in terms of the molar ﬂow,

dN

j

dV

= R

j

(58)

101

Volumetric Flowrate for Gas-Phase Reactions

• To use Equation 58 for designing a gas-phase reactor, one has to be able to

relate the volumetric ﬂowrate, Q, to the molar ﬂows, N

j

, j = 1, 2, . . . , n

s

.

• The important piece of information tying these quantities together is, again,

the equation of state for the reaction mixture, f(T, P, c

j

) = 0.

• Because the molar ﬂow and concentration are simply related,

N

j

= c

j

Q (59)

the equation of state is also a relation between temperature, pressure, molar

ﬂows, and volumetric ﬂowrate.

102

Ideal Gas Equation of State

• The ideal-gas equation of state, c = P/RT, can be stated in terms of molar

concentrations, c

j

, as

¸

j

c

j

=

P

RT

• In terms of molar ﬂows, the equation of state is

¸

j

N

j

Q

=

P

RT

• One can solve the previous equation for the volumetric ﬂowrate,

Q =

RT

P

¸

j

N

j

(60)

103

Ideal Gas Equation of State

• To evaluate the concentrations for use with the reaction rate expressions, one

simply rearranges Equation 59 to obtain

c

j

=

N

j

Q

=

P

RT

N

j

¸

j

N

j

(61)

104

Volumetric Flowrate for Liquid-Phase Reactions

• Consider the equation of state for a liquid-phase system to be arranged in

the form

ρ = f(T, P, c

j

)

• The mass density is related to the volumetric ﬂowrate and total mass ﬂow,

M =

¸

j

N

j

M

j

, via

M = ρQ (62)

• Multiplying Equation 58 by M

j

and summing on j produces

dM

dV

= 0, M(0) = M

f

in which M

f

is the feed mass ﬂowrate.

• The total mass ﬂow in a PFR is constant.

105

Volumetric Flowrate for Liquid-Phase Reactions

• We can solve for the volumetric ﬂowrate by rearranging Equation 62

Q =

M

f

ρ

(63)

• If the liquid density is considered constant, ρ = ρ

f

, then

Q = Q

f

, constant density (64)

and the volumetric ﬂowrate is constant and equal to the feed value.

• Equation 64 is used often for liquid-phase reactions.

106

Volumetric Flowrate for Liquid-Phase Reactions

• If we denote the time spent in the tube by τ = V/Q, if Q is constant, we can

rewrite Equation 57 as

dc

j

dτ

= R

j

, constant ﬂowrate (65)

which is identical to the constant-volume batch reactor.

• For the constant-ﬂowrate case, the steady-state proﬁle in a PFR starting from

a given feed condition is also the transient proﬁle in a batch reactor starting

from the equivalent initial condition.

107

Single Reaction Systems – Changing ﬂowrate in a PFR

• A pure vapor stream of A is decomposed in a PFR to form B and C

A

k

→ B +C (66)

• Determine the length of 2.5 cm inner-diameter tube required to achieve

35% conversion of A. The reactor temperature is 518

◦

C and the pressure is

2.0 atm. Assume the pressure drop is negligible.

• The reaction rate is ﬁrst order in A, k = 0.05 sec

−1

at the reactor temperature,

and the feed ﬂowrate is 35 L/min.

108

Changing ﬂowrate in a PFR

• The mole balance for component A gives

dN

A

dV

= R

A

• The production rate of A is R

A

= −r = −kc

A

.

• Substituting the production rate into the above equation gives,

dN

A

dV

= −kN

A

/Q (67)

• The volumetric ﬂowrate is not constant, so we use Equation 60, which as-

sumes an ideal-gas equation of state,

Q =

RT

P

(N

A

+N

B

+N

C

) (68)

109

Changing ﬂowrate in a PFR

• The ideal-gas assumption is reasonable at this reactor temperature and pres-

sure.

• One can relate the molar ﬂows of B and C to A using the reaction stoichiom-

etry. The mole balances for B and C are

dN

B

dV

= R

B

= r

dN

C

dV

= R

C

= r

• Adding the mole balance for A to those of B and C gives

d(N

A

+N

B

)

dV

= 0

d(N

A

+N

C

)

dV

= 0

• The stoichiometry does not allow the molar ﬂow N

A

+N

B

or N

A

+N

C

to change

with position in the tube.

110

Changing ﬂowrate in a PFR

• Because N

A

+N

B

and N

B

+N

C

are known at the tube entrance, one can relate

N

B

and N

C

to N

A

,

N

A

+N

B

= N

Af

+N

Bf

N

A

+N

C

= N

Af

+N

Cf

• Rearranging the previous equations gives,

N

B

= N

Af

+N

Bf

−N

A

N

C

= N

Af

+N

Cf

−N

A

• Substituting the relations for N

B

and N

C

into Equation 68 gives

Q =

RT

P

2N

Af

+N

Bf

+N

Cf

−N

A

111

Changing ﬂowrate in a PFR

• Because the feed stream is pure A, N

Bf

= N

Cf

= 0, yielding

Q =

RT

P

2N

Af

−N

A

**• Substituting this expression in Equation 67 gives the ﬁnal mole balance,
**

dN

A

dV

= −k

P

RT

N

A

2N

Af

−N

A

• The above differential equation can be separated and integrated,

N

A

N

Af

2N

Af

−N

A

N

A

dN

A

=

V

0

−

kP

RT

dV

112

Changing ﬂowrate in a PFR

• Performing the integration gives,

2N

Af

ln

N

A

/N

Af

+

N

Af

−N

A

= −

kP

RT

V

• The conversion of component j for a plug-ﬂow reactor operating at steady

state is deﬁned as

x

j

=

N

jf

−N

j

N

jf

• Because we are interested in the V corresponding to 35% conversion of A, we

substitute N

A

= (1 −x

A

)N

Af

into the previous equation and solve for V,

V = −

RT

kP

N

Af

[2ln(1 −x

A

) +x

A

]

113

Changing ﬂowrate in a PFR

• Because Q

f

= N

Af

RT/P is given in the problem statement and the tube

length is desired, it is convenient to rearrange the previous equation to obtain

z = −

Q

f

kA

c

[2ln(1 −x

A

) +x

A

]

• Substituting in the known values gives

z = −

35 ×10

3

cm

3

/min

0.05 sec

−1

60 sec/min

¸

4

π(2.5 cm)

2

¸

[2ln(1 −.35) +.35]

z = 1216 cm= 12.2 m

114

Multiple-Reaction Systems

• The modeler has some freedomin setting up the material balances for a plug-

ﬂow reactor with several reactions.

• The most straightforward method is to write the material balance relation for

every component,

dN

j

dV

= R

j

, j = 1, 2, . . . , n

s

R

j

=

n

r

¸

i=1

ν

ij

r

i

, j = 1, 2, . . . , n

s

• The reaction rates are expressed in terms of the species concentrations.

• The c

j

are calculated from the molar ﬂows with Equation 59

• Q is calculated from Equation 60, if an ideal-gas mixture is assumed.

115

Benzene pyrolysis in a PFR

• Hougen and Watson [3] analyzed the rate data for the pyrolysis of benzene

by the following two reactions.

• Diphenyl is produced by the dehydrogenation of benzene,

2C

6

H

6

k

1

k

−1

C

12

H

10

+H

2

2B

D + H

116

Benzene pyrolysis in a PFR

• Triphenyl is formed by the secondary reaction,

C

6

H

6

+C

12

H

10

k

2

k

−2

C

18

H

14

+H

2

(69)

B + D

T + H (70)

• The reactions are assumed to be elementary so that the rate expressions are

r

1

= k

1

c

2

B

−

c

D

c

H

K

1

¸

(71)

r

2

= k

2

c

B

c

D

−

c

T

c

H

K

2

¸

(72)

117

Benzene pyrolysis in a PFR

• Calculate the tube volume required to reach 50% total conversion of the ben-

zene for a 60 kmol/hr feed stream of pure benzene.

• The reactor operates at 1033K and 1.0 atm.

• Plot the mole fractions of the four components versus reactor volume.

118

Benzene pyrolysis in a PFR

• The rate and equilibrium constants at T = 1033K and P = 1.0 atm are given

in Hougen and Watson,

k

1

= 7 ×10

5

L/mol · hr

k

2

= 4 ×10

5

L/mol · hr

K

1

= 0.31

K

2

= 0.48

119

Benzene pyrolysis in a PFR

• The mole balances for the four components follow from the stoichiometry,

dN

B

dV

= −2r

1

−r

2

(73)

dN

D

dV

= r

1

−r

2

(74)

dN

H

dV

= r

1

+r

2

(75)

dN

T

dV

= r

2

(76)

• The initial condition for the ODEs are N

B

(0) = N

Bf

and N

D

(0) = N

H

(0) =

N

T

(0) = 0.

120

Benzene pyrolysis in a PFR

• The total molar ﬂux does not change with reactor volume.

Q =

RT

P

N

Bf

(77)

• The rate expressions are substituted into the four ODEs and they are solved

numerically.

• The total conversion of benzene, x

B

= (N

Bf

− N

B

)/N

Bf

, is plotted versus

reactor volume in Figure 8.

• A reactor volume of 404 L is required to reach 50% conversion. The compo-

sition of the reactor versus reactor volume is plotted in Figure 9.

121

0

0.1

0.2

0.3

0.4

0.5

0.6

0 200 400 600 800 1000 1200 1400 1600

V (L)

x

B

Figure 8: Benzene conversion versus reactor volume.

122

0

0.2

0.4

0.6

0.8

1

0 200 400 600 800 1000 1200 1400 1600

V (L)

y

B

y

H

y

T

y

D

y

j

Figure 9: Component mole fractions versus reactor volume.

123

Ethane pyrolysis in the presence of NO

• See the text for another worked PFR example.

124

Some PFR-CSTR Comparisons

• We have two continuous reactors in this chapter: the CSTR and the PFR.

• Let’s compare their steady-state efﬁciencies in converting reactants to prod-

ucts.

• For simplicity, consider a constant-density, liquid

A

k

→ B r = kc

n

A

(78)

• For this situation, the steady-state PFR material balance is given by Equa-

tion 65

dc

A

dτ

= −r(c

A

)

125

Some PFR-CSTR Comparisons

• We rearrange and solve for the time required to change from the feed condi-

tion c

Af

to some exit concentration c

A

τ =

c

A

c

Af

1

r(c

A

)

dc

A

• The area under the curve 1/r(c

A

) is the total time required to achieve the

desired concentration change.

1

r(c

A

)

c

Af

c

A

c

A

PFR

CSTR

126

Some PFR-CSTR Comparisons

• To achieve this same concentration change in the CSTR, we start with Equa-

tion 37, and solve for τ giving

τ =

c

Af

−c

A

r(c

A

)

• This result also can be interpreted as an area. Notice that this area is the

height, 1/r(c

A

), times the width, c

Af

−c

A

, of the rectangle.

1

r(c

A

)

c

Af

c

A

c

A

PFR

CSTR

127

Some PFR-CSTR Comparisons

• If 1/r(c

A

) is a decreasing function of c

A

, or, equivalently, r(c

A

) is an increas-

ing function of c

A

, to achieve the same conversion, the PFR time (or volume,

V

R

= Q

f

τ) is less than the CSTR time (volume).

• The PFR reaction rate varies with length. The rate is high at the entrance to the

tube where the concentration of A is equal to the feed value, and decreases

with length as the concentration drops. At the exit of the PFR, the rate is the

lowest of any location in the tube.

• Now considering that the entire volume of the CSTR is reacting at this lowest

rate of the PFR, it is intuitively obvious that more volume is required for the

CSTR to achieve the same conversion as the PFR.

128

Some PFR-CSTR Comparisons

• If the reaction order is positive (the usual case), the PFR is more efﬁcient. If

the reaction order is negative, the CSTR is more efﬁcient.

1

r(c

A

)

c

Af

c

A

c

A

CSTR

PFR

129

The PFR versus CSTR with separation

• The PFR achieves higher conversion than an equivalent volume CSTR for the

irreversible reaction with ﬁrst-order kinetics

A → B r = kc

A

• Consider the case in which we add separation.

• Find a single CSTR and separator combination that achieves the same con-

version as the PFR.

130

The PFR versus CSTR with separation

• The issue is to increase the CSTR achievable conversion using separation.

• Education in chemical engineering principles leads one immediately to con-

sider recycle of the unreacted A as a means to increase this conversion.

131

The PFR versus CSTR with separation

N

A0

N

A1

N

A2

N

A

N

A0

N

A

αN

A2

αN

A2

pure B

V

R

V

R

• In the text, we show how to ﬁnd the recycle ﬂowrate so this system achieves

the PFR conversion.

132

CSTR Equivalence Principle.

• This example was motivated by a recent result of Feinberg and Ellison called

the CSTR Equivalence Principle of Reactor-Separator Systems [2].

• This surprising principle states:

For a given reaction network with n

i

linearly independent reactions, any

steady state that is achievable by any reactor-separator design with total

reactor volume V is achievable by a design with not more than n

i

+ 1

CSTRs, also of total reactor volume V. Moreover the concentrations,

temperatures and pressures in the CSTRs are arbitrarily close to those

occurring in the reactors of the original design.

133

Stochastic Simulation of Chemical Reactions

• We wish to introduce next a topic of increasing importance to chemical engi-

neers, stochastic (random) simulation.

• In stochastic models we simulate quite directly the random nature of the

molecules.

• We will see that the deterministic rate laws and material balances presented in

the previous sections can be captured in the stochastic approach by allowing

the numbers of molecules in the simulation to become large.

• The stochastic modeling approach is appropriate if the random nature of the

system is one of the important features to be captured in the model.

• These situations are becoming increasingly important to chemical engineers

as we explore reactions at smaller and smaller length scales.

134

Stochastic Simulation of Chemical Reactions

• For example, if we are modeling the chemical transformation by reaction of

only a few hundreds or thousands of molecules at an interface, we may want

to examine explicitly the random ﬂuctuations taking place.

• In biological problems, we often consider the interactions of only several hun-

dred or several thousand protein molecules and cells.

• In sterilization problems, we may wish to model the transient behavior until

every last organism is eliminated.

135

Stochastic Simulation of Chemical Reactions

• Assume we have only a hundred molecules moving randomly in the gas phase

A

k

1

→ B (79)

B

k

2

→ C (80)

in a constant-volume batch reactor.

• The probability of reaction is assumed proportional to the

r

1

= k

1

x

A

r

2

= k

2

x

B

in which x

j

is the number of component j molecules in the reactor volume.

• Note x

j

is an integer, unlike the deterministic model’s c

j

, which is real.

136

Stochastic Simulation of Chemical Reactions

• The basic idea of the Gillespie algorithm is to: (i) choose randomly the time

at which the next reaction occurs, and (ii) choose randomly which reactions

occurs at that time.

1. Initialize. Set integer counter n to zero. Set the initial species numbers,

x

j

(0), j = 1, . . . n

s

. Determine stoichiometric matrix ν and reaction probabil-

ity laws (rate expressions)

r

i

= k

i

h(x

j

)

for all reactions.

2. Compute reaction probabilities, r

i

= k

i

h(x

j

). Compute total reaction proba-

bility r

tot

=

¸

i

r

i

.

137

Stochastic Simulation of Chemical Reactions

3. Select two random numbers, p

1

, p

2

, from a uniform distribution on the inter-

val (0, 1). Let the time interval until the next reaction be

`

t = −ln(p

1

)/r

tot

(81)

138

Stochastic Simulation of Chemical Reactions

4. Determine reaction m to take place in this time interval. The idea

here is to partition the interval (0,1) by the relative sizes of each

reaction probability and then “throw a dart” at the interval to pick

the reaction that occurs. In this manner, all reactions are pos-

sible, but the reaction is selected in accord with its probability.

p

2

r

1

r

1

+r

2

r

2

r

1

+r

2 0

1

139

Stochastic Simulation of Chemical Reactions

5. Update the simulation time t(n+1) = t(n) +

`

t. Update the species numbers

for the single occurrence of the mth reaction via

x

j

(n+1) = x

j

(n) +ν

mj

, j = 1, . . . n

s

Let n = n+1. Return to Step 2.

140

Stochastic Simulation of Chemical Reactions

• If r

tot

is the total probability for reaction, e

−r

tot

`

t

is the probability that a re-

action has not occurred during time interval

`

t.

• We will derive this fact in Chapter 8 when we develop the residence-time

distribution for a CSTR.

• The next ﬁgure shows the results of this algorithm when starting with x

A

=

100 molecules.

141

Stochastic Simulation of Chemical Reactions

142

0

20

40

60

80

100

0 1 2 3 4 5

t

x

A

x

B

x

C

x

j

143

Stochastic Simulation of Chemical Reactions

• Notice the random aspect of the simulation gives a rough appearance to the

number of molecules versus time, which is quite unlike any of the determin-

istic simulations.

• Because the number of molecules is an integer, the simulation is actually

discontinuous with jumps between simulation times.

• But in spite of the roughness, we already can make out the classic behavior

of the series reaction: loss of starting material A, appearance and then dis-

appearance of the intermediate species B, and slow increase in ﬁnal product

C.

144

Stochastic Simulation of Chemical Reactions

• Next we explore the effect of increasing the initial number of A molecules on

a single simulation. The results for 1000 and 4000 initial A molecules are

shown in the next ﬁgures.

145

Stochastic Simulation of Chemical Reactions

146

0

200

400

600

800

1000

0 1 2 3 4 5

t

x

A

x

B

x

C

x

j

147

Stochastic Simulation of Chemical Reactions

148

0

800

1600

2400

3200

4000

0 1 2 3 4 5

t

x

A

x

B

x

C

x

j

149

Stochastic Simulation of Chemical Reactions

• We see the random ﬂuctuations become less pronounced. Notice that even

with only 4000 starting molecules, the results compare very favorably with

the deterministic simulation shown previously.

• Another striking feature of the stochastic approach is the trivial level of pro-

gramming effort required to make the simulations.

• The biggest numerical challenge is producing the pseudo-random numbers

and many well-developed algorithms are available for that task.

• The computational time required for performing the stochastic simulation

may, however, be large.

150

Hepatitis B virus modeling

151

Hepatitis B virus modeling

nucleotides

cccDNA

−−−−−−→ rcDNA (82)

nucleotides + rcDNA −−−−−−→ cccDNA (83)

amino acids

cccDNA

−−−−−−→ envelope (84)

cccDNA −−−−−−→ degraded (85)

envelope −−−−−−→secreted or degraded (86)

rcDNA + envelope −−−−−−→ secreted virus (87)

152

Hepatitis B virus modeling

• The reaction rates and production rates for Reactions 82–87 are given by

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

r

1

r

2

r

3

r

4

r

5

r

6

=

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

¸

k

1

x

A

k

2

x

B

k

3

x

A

k

4

x

A

k

5

x

C

k

6

x

B

x

C

¸

¸

¸

¸

R

A

R

B

R

C

=

¸

¸

¸

¸

r

2

−r

4

r

1

−r

2

−r

6

r

3

−r

5

−r

6

(88)

in which A is cccDNA, B is rcDNA, and C is envelope.

153

Hepatitis B virus modeling

• Assume the systems starts with a single cccDNA molecule and no rcDNA and

no envelope protein, and the following rate constants

¸

x

A

x

B

x

C

¸

T

=

¸

1 0 0

¸

T

(89)

k

T

=

¸

1 0.025 1000 0.25 2 7.5 ×10

−6

¸

(90)

154

Average stochastic is not deterministic.

0

5

10

15

20

25

0 50 100 150 200

t (days)

c

c

c

D

N

A

deterministic

avg. stochastic

0

50

100

150

200

250

0 50 100 150 200

t (days)

r

c

D

N

A

deterministic

avg. stochastic

155

0

2000

4000

6000

8000

10000

0 50 100 150 200

t (days)

e

n

v

e

l

o

p

e

deterministic

avg. stochastic

156

Hepatitis B virus modeling

157

0

5

10

15

20

25

30

35

0 50 100 150 200

t (days)

c

c

c

D

N

A

deterministic

stoch 1

stoch 2

Figure 10: Species cccDNA versus time for hepatitis B virus model; two repre-

sentative stochastic trajectories.

158

Hepatitis B virus modeling

• The simulation of the deterministic model and an average of 500 stochastic

simulations are not the same.

• Figure 10 shows two representative stochastic simulations for only the cc-

cDNA species.

• Notice the ﬁrst stochastic simulation does ﬂuctuate around the deterministic

simulation as expected.

• The second stochastic simulation, however, shows complete extinction of the

virus. That is another possible steady state for the stochastic model.

159

Average stochastic is not deterministic.

• In fact, it occurs for 125 of the 500 simulations. So the average stochastic

simulation consists of 75% trajectories that ﬂuctuate about the deterministic

trajectory and 25% trajectories that go to zero.

0

5

10

15

20

25

0 50 100 150 200

t (days)

c

c

c

D

N

A

deterministic

avg. stochastic

160

Summary

• We have introduced four main reactor types in this chapter.

• the batch reactor

• the continuous-stirred-tank reactor (CSTR)

• the semi-batch reactor

• the plug-ﬂow reactor (PFR).

161

Summary

BATCH

d(c

j

V

R

)

dt

= R

j

V

R

(91)

constant volume

dc

j

dt

= R

j

(92)

CSTR

d(c

j

V

R

)

dt

= Q

f

c

jf

−Qc

j

+R

j

V

R

(93)

constant density

dc

j

dt

=

1

τ

(c

jf

−c

j

) +R

j

(94)

steady state c

j

= c

jf

+R

j

τ (95)

SEMI-BATCH

d(c

j

V

R

)

dt

= Q

f

c

jf

+R

j

V

R

(96)

PFR

∂c

j

∂t

= −

∂(c

j

Q)

∂V

+R

j

(97)

steady state

d(c

j

Q)

dV

= R

j

(98)

constant ﬂowrate

dc

j

dτ

= R

j

, τ = V/Q

f

(99)

162

Summary

We also have introduced some of the basic reaction-rate expressions.

• ﬁrst order, irreversible

• ﬁrst order, reversible

• second order, irreversible

• nth order, irreversible

• two ﬁrst-order reactions in series

• two ﬁrst-order reactions in parallel

• two second-order, reversible reactions

163

Summary

• We developed the equations required to compute the volume of the reactor if

there is a signiﬁcant volume change upon reaction. We require an equation

of state for this purpose.

• Several of these simple mass balances with basic rate expressions were solved

analytically.

• In the case of multiple reactions with nonlinear rate expressions (i.e., not

ﬁrst-order reaction rates), the balances must be solved numerically.

• A high-quality ordinary differential equation (ODE) solver is indispensable for

solving these problems.

164

Summary

• We showed that the PFR achieves higher conversion than the CSTR of the

same volume if the reaction rate is an increasing function of a component

composition (n > 0 for an nth-order rate expression).

• Conversely, the CSTR achieves higher conversion than the same-volume PFR

if the rate is a decreasing function of a component composition (n < 0).

• Finally, we introduced stochastic simulation to model chemical reactions oc-

curring with small numbers of molecules.

165

Summary

• The stochastic model uses basic probability to compute reaction rate. A given

reaction’s probability of occurrence is assumed proportional to the number

of possible combinations of reactants for the given stoichiometry.

• Two pseudo-random numbers are chosen to determine: (i) the time of the

next reaction and (ii) the reaction that occurs at that time.

• The smooth behavior of the macroscopic ODE models is recovered by the

random simulations in the limit of large numbers of reacting molecules.

• With small numbers of molecules, however, the average of the stochastic

simulation does not have to be equal to the deterministic simulation. We

demonstrated this fact with the simple, nonlinear hepatitis B virus model.

166

References

[1] R. B. Bird, W. E. Stewart, and E. N. Lightfoot. Transport Phenomena. John

Wiley & Sons, New York, second edition, 2002.

[2] M. Feinberg and P. Ellison. General kinetic bounds on productivity and se-

lectivity in reactor-separator systems of arbitrary design: I. Principles. Ind.

Eng. Chem. Res., 40(14):3181–3194, 2001.

[3] O. A. Hougen and K. M. Watson. Chemical Process Principles. Part Three:

Kinetics and Catalysis. John Wiley & Sons, New York, 1947.

167

General Mole Balance

Rj V Q0 cj0 Q1 cj1

Conservation of mass

rate of accumulation = of component j

rate of inﬂow of component j

−

rate of outﬂow of component j

rate of generation of component j by + chemical reactions

(1)

2

General Mole Balance

Rj V Q0 cj0 Q1 cj1

Conservation of mass d dt

V

cj dV = Q0cj0 − Q1cj1 +

V

Rj dV

(2)

3

. including inerts. 4 . • The diffusional ﬂux will be considered during the development of the material balance for the packed-bed reactor. ns . • Assuming component j enters and leaves the volume element only by convection with the inﬂow and outﬂow streams. i. . 2.General Mole Balance d dt V cj dV = Q0cj0 − Q1cj1 + V Rj dV • Equation 2 applies to every chemical component in the system. j = 1.e. . which do not take place in any reactions. neglecting diffusional ﬂux through the boundary of the volume element due to a concentration gradient. .

Rate expressions • To solve the reactor material balance. Chapter 5 • Here we use common reaction-rate expressions without derivation 5 . Rj Rj = i νij ri • Therefore we require ri as a function of cj • This is the subject of chemical kinetics. we require an expression for the production rates.

The Batch Reactor Rj • The batch reactor is assumed well stirred • Let the entire reactor contents be the reactor volume element 6 .

Batch Reactor d dt V cj dV = Q0cj0 − Q1cj1 + V Rj dV • Because the reactor is well stirred. the integrals in Equation 2 are simple to evaluate. Q0 = Q1 = 0. d cj V R dt = Rj VR (3) 7 . VR cj dV = cj VR VR Rj dV = Rj VR • The inﬂow and outﬂow stream ﬂowrates are zero.

• If the reactor volume is constant (liquid-phase reactions) dcj = Rj dt (4) • Use Equation 3 rather than Equation 4 if the reactor volume changes signiﬁcantly during the course of the reaction.Reactor Volume • Equation 3 applies whether the reactor volume is constant or changes during the reaction. 8 .

9 .Analytical Solutions for Simple Rate Laws • In general the material balance must be solved numerically. • If the reactor is isothermal. the rate expressions are simple. then analytical solutions of the material balance are possible. • We next examine derive analytical solutions for some classic cases. we have few components.

k r = kcA (5) • The material balance for a constant-volume reactor gives dcA = −kcA dt • Watch the sign! (6) 10 . irreversible reaction A → B.First-order. irreversible • Consider the ﬁrst-order.

cA(t) = cA0. t=0 • The solution to the differential equation with this boundary condition is cA = cA0e−kt (7) 11 .First-order. irreversible • We denote the initial concentration of A as cA0.

First-order. irreversible 12 .

2 k=2 k=5 0 1 2 3 4 5 0 t 13 .6 0.5 k=1 0.1 0.8 cA cA0 0.4 k = 0.

First-order, irreversible

• The A concentration decreases exponentially from its initial value to zero with increasing time. • The rate constant determines the shape of this exponential decrease. Rearranging Equation 7 gives ln(cA/cA0) = −kt

14

First-order, irreversible

1

k = 0.5 k=1

0.1

cA cA0

0.01

k=2

k=5

0.001 0 1 2 3 4 5

t

• One can get an approximate value of the rate constant from the slope of the straight line.

15

• This procedure is a poor way to determine a rate constant and should be viewed only as a rough approximation (Chapter 9).

16

Note that the sum of cA and cB is constant.First-order. d(cA + cB ) = RA + RB = 0 dt Therefore. irreversible The B concentration is determined from the A concentration. 17 (8) . cB (0) = cB0 2. dcB = RB = kcA dt with the initial condition for B. Solve the material balance for component B. 1. cA + cB is a constant.

cA + cB = cA0 + cB0 • So we have an expression for cB cB = cA0 + cB0 − cA cB = cB0 + cA0(1 − e−kt ) (9) (10) 18 .First-order. reversible • The value is known at t = 0.

• The material balances for A and B are now dcA = −r = −k1 cA + k−1 cB dt dcB = r = k1 cA − k−1 cB dt • Notice that cA + cB = cA0 + cB0 remains constante cA (0) = cA0 cB (0) = cB0 k1 k−1 B (11) 19 . reversible • Consider now the same ﬁrst-order reaction. but assume it is reversible A • The reaction rate is r = k1 cA − k−1 cB .First-order.

reversible • Eliminate cB in the material balance for A gives dcA = −k1cA + k−1(cA0 + cB0 − cA) dt • How do we want to solve this one? • Particular solution and homogeneous solution (see text) • Laplace transforms (control course) • Separation! (12) 20 .First-order.

b = k−1(cA0 + cB0) 21 . reversible dcA = acA + b dt dcA = cA0 acA + b 1 ln(acA + b) a cA = cA0eat − cA cA t dt 0 =t cA0 b (1 − eat ) a Substitute a = −(k1 + k−1).First-order.

yielding cB = cB0e −(k1 +k−1 )t k1 + (cA0 + cB0) 1 − e−(k1+k−1)t k1 + k−1 (14) 22 . reversible cA = cA0e−(k1+k−1)t + k−1 (cA0 + cB0) 1 − e−(k1+k−1)t k1 + k−1 (13) • The B concentration can be determined by switching the roles of A and B and k1 and k−1 in Reaction 11.First-order.

k1 = 1. reversible kinetics in a batch reactor.First-order. reversible 1 0.5. 23 .2 0 0 1 2 3 4 5 t Figure 1: First-order. cB0 = 0. k−1 = 0.4 cAs cB (t) 0.6 c 0. cA0 = 1.8 cA (t) cBs 0.

24 .Nonzero steady state • For the reversible reaction. the concentration of A does not go to zero. • Taking the limit t → ∞ in Equation 13 gives k−1 (cA0 + cB0) = k1 + k−1 cAs in which cAs is the steady-state concentration of A.

Nonzero steady state • Deﬁning K1 = k1/k−1 allows us to rewrite this as cAs 1 = (cA0 + cB0) 1 + K1 • Performing the same calculation for cB gives cBs K1 (cA0 + cB0) = 1 + K1 25 .

dt • Our ﬁrst nonlinear differential equation.Second-order. cA(0) = cA0 (16) 26 . irreversible • Consider the irreversible reaction k A → B 2 in which the rate expression is second order. r = kcA. (15) • The material balance and initial condition are dcA 2 = −kcA.

irreversible • Separation works here dcA 2 = −kdt cA cA t dcA 2 = −k cA0 cA 1 1 − = −kt cA0 cA • Solving for cA gives 1 + kt cA = cA0 dt 0 −1 (17) • Check that this solution satisﬁes the differential equation and initial condition 27 .Second-order.

2 0.9 0. irreversible 1 0.8 0. 28 .4 0.7 0.Second-order.1 0 0 1 2 3 4 5 ﬁrst order second order • The second-order reaction decays more slowly to zero than the ﬁrst-order reaction.3 0.5 0.6 cA cA0 0.

Another second-order. irreversible A+B → C k r = kcAcB (18) • The material balance for components A and B are dcA = −r = −kcAcB dt dcB = −r = −kcAcB dt • Subtract B’s material balance from A’s to obtain d(cA − cB ) =0 dt (19) (20) 29 .

Another second-order. irreversible • Therefore. and cB = cA − cA0 + cB0 • Substituting this expression into the material balance for A yields dcA = −kcA(cA − cA0 + cB0) dt • This equation also is separable and can be integrated to give (you should work through these steps). cB0 (cB0−cA0)kt cA = (cA0 − cB0) 1 − e cA0 −1 (21) . cA0 ≠ cB0 (22) 30 . cA − cB is constant.

d(cA + cC )/dt = 0. irreversible • Component B can be computed from Equation 21. giving cA0 (cA0−cB0)kt e cB = (cB0 − cA0) 1 − cB0 • What about component C? C’s material balance is dcC = kcAcB dt and therefore. The concentration of C is given by cC = cA0 − cA + cC0 −1 (23) 31 .Another second-order. or by switching the roles of A and B in Reaction 18.

Another second-order. the steady state is cAs = 0 cBs = cB0 − cA0 cCs = cA0 + cC0 32 . irreversible • Notice that if cA0 > cB0 (Excess A). the steady state cAs = cA0 − cB0 cBs = 0 cCs = cB0 + cC0 • For cB0 > cA0 (Excess B).

nth-order. irreversible n The nth-order rate expression r = kcA 33 .

5 3 34 .5 4.5 0 0 0.5 4 3.5 1 0.5 1 1.5 3 n=3 2 r 2.5 cA 2 1 1/2 0 2.5 2 1.

n≠1 35 . irreversible A → B k n r = kcA dcA n = −r = −kcA dt • This equation also is separable and can be rearranged to dcA n = −kdt cA • Performing the integration and solving for cA gives −n+1 cA = cA0 + (n − 1)kt 1 −n+1 .nth-order.

irreversible • We can divide both sides by cA0 to obtain 1 cA −n+1 . 36 .nth-order. = [1 + (n − 1)k0t] cA0 n≠1 (24) in which n−1 k0 = kcA0 has units of inverse time.

8 cA cA0 0. irreversible 1 0.6 n=5 0.2 0 0 1 2 3 4 5 t • The larger the value of n.4 4 3 2 1 0. 37 . the more slowly the A concentration approaches zero at large time.nth-order.

irreversible • Exercise care for n < 1.5 2 38 .8 cA cA0 0. 1 0.6 0.5 −1/2 0 1/2 1 t 1.2 −2 −1 0 0 0.nth-order.4 n=2 1 0. cA reaches zero in ﬁnite time.

the rate decreases with increasing reactant concentration.5 1 0. 5 4.5 n=−2 −1 −1/2 0 1 1.5 2 1.Negative order. the reactant inhibits the reaction.5 3 r 2.5 0 0 0. inhibition • For n < 0.5 4 3.5 cA 2 2.5 3 39 .

• When using an ODE solver we may modify the right-hand sides of the material balance n −kcA . cA = 0 dt • Examine the solution carefully if the concentration reaches zero. 40 . but watch out for small concentrations.Negative order. which is not physically realistic. cA > 0 dcA = 0. inhibition • Inhibition reactions are not uncommon. Notice the rate becomes unbounded as cA approaches zero.

• Let the reaction rates be given by simple ﬁrst-order rate expressions in the corresponding reactants. k1 k2 A → B B → C (25) (26) • Reactant A decomposes to form an intermediate B that can further react to form a ﬁnal product C. r1 = k1cA r2 = k2cB 41 .Two reactions in series • Consider the following two irreversible reactions.

42 .Two reactions in series • The material balances for the three components are dcA = RA = −r1 = −k1cA dt dcB = RB = r1 − r2 = k1cA − k2cB dt dcC = RC = r2 = k2cB dt • The material balance for component A can be solved immediately to give cA = cA0e−k1t as before.

homogeneous equation approach produces cB = cB0e −k2 t k1 + cA0 e−k1t − e−k2t . now what? • Either Laplace transform or particular solution. not separable.Two reactions in series – B • The material balance for B becomes dcB + k2cB = k1cA0e−k1t dt Oops. k2 − k1 k1 ≠ k2 (27) 43 .

notice from the material balances that d(cA + cB + cC )/dt = 0. cC is cC = cA0 + cB0 + cC0 − cA − cB 44 .Two reactions in series – C • To determine the C concentration. Therefore.

2 0 0 1 2 3 4 5 t Figure 2: Two ﬁrst-order reactions in series in a batch reactor.8 cA (t) 0. k1 = 2. 45 .Two reactions in series 1 0. cB0 = cC0 = 0. k2 = 1.6 cC (t) c 0. cA0 = 1.4 cB (t) 0.

A → B A → C k2 k1 (28) (29) • Assume the rates of the two irreversible reactions are given by r1 = k1cA and r 2 = k 2 cA .Two reactions in parallel • Consider next two parallel reactions of A to two different products. 46 . B and C.

Two reactions in parallel • The material balances for the components are dcA = RA = −r1 − r2 = −k1cA − k2cA dt dcB = R B = r1 = k 1 cA dt dcC = R C = r2 = k 2 cA dt 47 .

Two reactions in parallel • The material balance for A can be solved directly to give cA = cA0e−(k1+k2)t • Substituting cA(t) into B’s material balance gives dcB = k1cA0e−(k1+k2)t dt • This equation is now separable and can be integrated directly to give cB = cB0 + cA0 k1 1 − e−(k1+k2)t k1 + k2 (31) (30) 48 .

component C can be determined from the condition that cA + cB + cC is constant or by switching the roles of B and C. cC = cC0 + cA0 k2 1 − e−(k1+k2)t k1 + k2 (32) 49 . and k1 and k2 in Equation 31.Two reactions in parallel • Finally.

k1 = 1.Two reactions in parallel 1 0. k2 = 2.5 3 Figure 3: Two ﬁrst-order reactions in parallel in a batch reactor.5 1 1.5 t 2 2. cB0 = cC0 = 0.2 0 0 0.8 cA (t) 0.6 cC (t) c 0. cA0 = 1. 50 .4 cB (t) 0.

A • The rate constants determine which product is favored • Large values of k1/k2 favor the formation of component B compared to C and vice versa 51 .Two reactions in parallel • Notice that the two parallel reactions compete for the same reactant.

so be clear about the deﬁnition you choose. Yield: The yield of component j is the fraction of a reactant that is converted into component j. 52 . It is best to state the deﬁnition in the context of the problem being solved. Yield. Overall selectivity: The overall selectivity is the ratio of the amount of one component produced to the amount of another component produced. Point selectivity: The point (or instantaneous) selectivity is the ratio of the production rate of one component to the production rate of another component. Conversion: Conversion is normally deﬁned to be the fraction of a component that has been converted to products by the reaction network. Selectivity There are several ways to deﬁne selectivity.Conversion. Conversion has several deﬁnitions and conventions. yield and conversion.

.The Continuous-Stirred-Tank Reactor (CSTR) Qf cjf Rj Q cj Writing the material balance for this reactor gives d cj V R dt = Qf cjf − Qcj + Rj VR . . ns (33) 53 . . . j = 1.

Equation 33 reduces to dcj 1 = (cjf − cj ) + Rj dt τ • The parameter τ = VR /Qf is called the mean residence time of the CSTR. • We refer to this balance as the constant-density case.CSTR – Constant Density • If the reactor volume is constant and the volumetric ﬂowrates of the inﬂow and outﬂow streams are the same. It is often a good approximation for liquid-phase reactions. (34) 54 .

0 = Qf cjf − Qcj + Rj VR (35) • Conversion of reactant j is deﬁned for a steady-state CSTR as follows xj = Qf cjf − Qcj Qf cjf (steady state) (36) • One can divide Equation 35 through by VR to obtain for the constant-density case cj = cjf + Rj τ (steady state. constant density) (37) 55 .CSTR – Steady State • The steady state of the CSTR is described by setting the time derivative in Equation 33 to zero.

1. 3. the residence time of the reactor is τ = 100 min.Transient behavior of the CSTR • Consider a ﬁrst-order. Plot the concentration of A versus time for constant feed concentration cAf = 2 mol/L if the reactor is initially ﬁlled with an inert so cA0 = 0 mol/L. liquid-phase reaction in an isothermal CSTR A → 2B k r = kcA the feed concentration of A is cAf = 2 mol/L. Find the steady-state concentration of A in the efﬂuent for the given feed. 2. 56 . Plot the concentration of A versus time for constant feed concentration cAf = 2 mol/L if the reactor is initially ﬁlled with feed so cA0 = 2 mol/L. and the rate constant is k = 0.1 min−1.

cA = cAf + RAτ • Substituting the production rate RA = −kcA and solving for cA gives the steady-state concentration cAf cAs = 1 + kτ • Substituting in the numerical values gives cAs 2 mol/L = = 0.Transient CSTR Solution.1 min−1)(100 min) 57 . Part 1 • Liquid phase: assume the ﬂuid density is constant.182 mol/L 1 + (0.

The analytical solution is cA(t) = cA0e−(1/τ+k)t + cAf 1 − e−(1/τ+k)t 1 + kτ (38) (39) 58 .Parts 2 and 3 1 dcA = cAf − cA − kcA dt τ cA(0) = cA0 • This equation is also separable.

5 cA (t) (mol/L) 1 cA0 = 2 0.Parts 2 and 3 2 1.5 cAs = 0.182 0 0 cA0 = 0 10 20 30 40 50 t (min) 60 70 80 • Both solutions converge to the same steady-state even though the starting conditions are quite different. 59 .

Phenol production in a CSTR • Consider the reaction of cumene hydroperoxide (CHP) to phenol and acetone (C6H5)C(CH3)2OOH → (C6H5)OH + (CH3)2CO CHP → phenol + acetone (40) (41) • The reaction is pseudo-ﬁrst-order r = kcCHP • Find the reactor volume to achieve 85% conversion of CHP at steady state.12 hr−1.9 m3/hr and k = 4. 60 . The ﬂowrate into the reactor is Qf = 26.

and solving for the CHP concentration gives cA = cAf 1 + kτ (42) 61 . so assume constant density and Q = Qf . cA = cAf + RAτ RA = −kcA.Phenol production • Liquids at 85◦C.

Phenol production • The conversion of CHP (for Q = Qf ) is cAf − cA cA xA = =1− cAf cAf xA = kτ 1 + kτ • Solving this equation for τ gives 1 xA τ= k 1 − xA 62 .

85) (4.Phenol production • Substituting the relation τ = VR /Qf and solving for VR gives VR = Qf xA k(1 − xA) • Substituting in the known values gives the required CSTR volume VR = (26.9 m3/hr)(0.15) = 37 m3 63 .12 hr−1)(0.

but allows a feed addition policy while the reaction takes place. 64 . • Normally there is no outﬂow stream. • The semi-batch reactor is initially charged with reactant.The Semi-Batch Reactor • The semi-batch reactor is a cross between the batch reactor and CSTR. like the batch reactor. like the CSTR.

• Compared to the batch reactor. ns (43) • One may choose to operate a semi-batch reactor to control the reaction rate or heat release during reaction by slowly adding one of the reactants in the feed stream. . 65 .The Semi-Batch Reactor • We set Q = 0 in the CSTR material balance to obtain d cj VR dt = Qf cjf + Rj VR . . the semi-batch reactor provides more complete use of the reactor volume in reactions such as polymerizations that convert from lower density to higher density during the course of the reaction. . j = 1. .

CSTR and semi-batch reactors. d( j cj Mj VR ) dt = Qf j cjf Mj − Q j cj Mj + j Rj Mj VR (45) • The term j cj Mj is simply the mass density of the reactor contents. which 66 . depending on how we specify Qf and Q. • If we multiply Equation 44 by the molecular weight of species j and sum over all species we obtain.Volume Change Upon Reaction d cj VR dt = Qf cjf − Qcj + Rj VR (44) • Equation 44 covers both the batch.

we denote ρ ρ= ns cj M j j=1 (46) 67 .

Volume Change Upon Reaction • The term j cjf Mj is the mass density of the feedstream. ρf . 68 . j • We know that conservation of mass in chemical reactions implies (see Chapter 2). in which the total mass in the reactor changes in time due to the inﬂow and outﬂow of mass. • Notice that chemical reactions play no role in the total mass balance. Substitution into Equation 45 leads to d(ρVR ) = Qf ρf − Qρ dt Rj Mj = 0 (47) • Equation 47 is clearly a total mass balance.

**Equation of state for the mixture
**

• If we have a single-phase system at equilibrium, the intensive variables cj , T , P , completely specify all intensive variables of the system. • In this chapter we consider T and P to be known, ﬁxed quantities. Therefore, the density of the reaction mixture, which is an intensive variable, is known if the cj are known. • This relationship is one form of the equation of state for the mixture ρ = f (T , P , c1, c2, . . . , cns ) • Substituting the deﬁnition of density, we can express the equation of state as f (c1, c2, . . . , cns ) = 0

69

Equation of state for the mixture

**• For example, we could express the equation of state in terms of the partial molar volumes as cj V j = 1
**

j

in which V j is the partial molar volume of component j in the mixture. • The partial molar volumes are functions of T , P and cj .

70

Equation of state for the mixture — Ideal mixture

**• If we assume an ideal mixture, this reduces to
**

◦ cj Vj = 1, j ◦ in which Vj is the speciﬁc volume of pure component j, which is a function of only T and P .

ideal mixture

• We assume that a thermodynamic equation of state is valid even when the reactor is not at equilibrium.

71

it does not vary with time either and we can set it to the feed value.” • This terminology should be avoided.Constant density • Because the reaction mixture density. is independent of composition. ρ = ρf (48) • The total mass balance then reduces to dVR = Qf − Q dt which is sometimes referred to as a “volume balance. (49) 72 . ρ.

• CSTR (dynamic and steady state). 73 . the reactor is ﬁlled during operation so Qf is speciﬁed and positive for some time and Q = 0. • Semi-batch reactor. Q = Qf = 0. In the semi-batch reactor. The solution to Equation 49 then determines the change in volume of the reactor during the ﬁlling operation. If the outﬂow of the CSTR is regulated so that the CSTR has constant volume. For the batch reactor. then we can conclude from Equation 49 that Q = Qf . We can therefore conclude from Equation 49 that a batch reactor with constant density has constant volume.Constant density dVR = Qf − Q dt • Batch reactor.

• In the general case. • The ﬁnal equation is provided by a statement of reactor operation. • We have the ns equations from the component mole balances. • The equation of state provides one additional equation. • The set of unknowns is nj . VR . • We therefore have ns + 2 unknowns. consider the following variables to fully determine the state of the reactor: T . nj . P . • We also require the value of Q to specify the right-hand sides of the material balances. 74 . Q.Nonconstant density Unknowns. VR . Equations.

A semi-batch reactor is operated in this way with Q = 0 until the reactor is ﬁlled with the reactants. Constant-mass reactor.Nonconstant density – reactor operation 1. Flowrate out of the reactor is speciﬁed. In this case Q(t) is speciﬁed. The constant-mass reactor can be achieved if a differential pressure measurement is used to control the ﬂowrate out of the reactor and the reactor has constant cross-sectional area. In this situation ρVR is speciﬁed as the additional equation. 2. VR is speciﬁed as the additional equation. 75 . 3. Constant-volume reactor. In this situation. This type of operation may be achieved if the ﬂowrate out of the reactor is controlled by a ﬂow controller. The constant-volume reactor can be achieved by allowing overﬂow of the reactor to determine ﬂowrate out of the reactor.

dVR = Qf dt • in which fj is fj = • and ∆fi is ∆fi = j j fj cjf j fj cj −Q+ i ∆fi ri VR j f j cj (50) ∂f ∂cj ∂f ∂cj νij fj = j νij which is a change in a derivative property upon reaction. 76 .Nonconstant density – reactor operation • See the text for the derivation.

• So the reactor volume can be calculated from dVR = Qf − Q + dt ∆Vi◦riVR i 77 . ◦ fj = Vj the pure component speciﬁc volumes • The ∆fi are given by ∆fi = ∆Vi◦ the change in speciﬁc volume upon reaction i.Nonconstant density – idea mixture • For the ideal mixture we have f (cj ) = j ◦ cj Vj − 1 = 0.

vol VR = VR0 VR = VR0 dVR = Qf − Q dt Q = Qf Q = Qf Q speciﬁed VR = VR0 dVR = Qf (1 − ρf /ρ) + dt dVR = Qf − Q + dt ◦ i ∆Vi ri VR Q = Qf + ◦ i ∆Vi ri VR 2. mass Q = Qf ρf /ρ Q speciﬁed 3. Q. ns j = 1. . Q ◦ i ∆Vi ri VR Table 1: Reactor balances for constant-density and ideal-mixture assumptions. dt nj = cj VR ρ= j cj Mj Deﬁned quantities: ∆V ◦ = i ◦ j cj Vj = 1 (i) constant density: ρ = ρ0 (ii) ideal mixture: 1. . . 78 . . . .Nonconstant density Unknowns (ns + 2): Component balances: VR . . ns ◦ j νij Vj dnj = Qf cjf − Qcj + Rj VR . nj . . j = 1.

ns ∂f j νij ∂c j dnj = Qf cjf − Qcj + Rj VR . . . . . ns j = 1. 79 . nj . Q Table 2: Reactor balances for general equation of state. c2 . . . . . . dt nj = cj VR ρ= j cj Mj Deﬁned quantities: ∆fi = Equation of state: f (c1 . vol VR = VR0 ρVR = ρ0 VR0 Q speciﬁed f (cj ) = 0 f (cj ) = 0 f (cj ) = 0 dVR = Qf dt dVR = Qf dt VR = VR0 j fj cjf −Q+ j fj cj j fj cjf −Q+ j fj cj i ∆fi ri VR j f j cj i ∆fi ri VR j f j cj Q = Qf j fj cjf + j fj cj i ∆fi ri VR j fj cj 2. . . cns ) = 0 DAEs ODEs 1. j = 1. mass Q = Qf ρf /ρ Q speciﬁed 3. Q. .Nonconstant density Unknowns (ns + 2): Component balances: VR .

• A pure monomer feed is slowly added at ﬂowrate Qf 0 = 1 m3/min to ﬁll the reactor in semi-batch operation to control the heat release.Semi-batch polymerization • Consider a solution polymerization reaction. which can be modeled as a ﬁrstorder. 80 . irreversible reaction M → P k r = kcM • A 20 m3 semi-batch reactor is initially charged with solvent and initiator to half its total volume.

and the percentage increase in polymer production achieved in the second operation. 2.Semi-batch polymerization • Consider two cases for the subsequent reactor operation. 81 . The monomer feed is shut off and the reaction goes to completion. 1. • Calculate the total polymer mass production. The monomer feed is adjusted to keep the reactor ﬁlled while the reaction goes to completion.

82 .1 min−1.The physical properties • You may assume an ideal mixture • The densities of monomer and polymer are ρM = 800 kg/m3 ρP = 1100 kg/m3 • The monomer molecular weight is MM = 100 kg/kmol • The rate constant is k = 0.

• We denote the total number of moles of monomer by M = cM VR . and can write the monomer balance as dM = Qf 0cMf − kM dt M(0) = 0 (51) 83 . the monomer mole balance is d(cM VR ) = Qf 0cMf − kcM VR dt in which cMf = ρM /MM is given.Semi-batch polymerization • While the reactor is ﬁlling. and Qf = Qf 0 is constant during the ﬁlling operation.

Semi-batch polymerization • For an ideal mixture. the volume is given by dVR = Qf 0 + ∆V kM dt VR (0) = 10 m3 in which ∆V = (1/ρP − 1/ρM ) MM (52) 84 .

we note from the stoichiometry that the mass production rate of polymer R P is R P = −RM MM • The mass balance for total polymer P is given by dP = R p VR = kcM MM VR = (kMM )M dt (53) 85 .The polymer mass production • To compute the polymer mass.

• dasrt is an ODE solver with the added capability to ﬁnd the time at which some event of interest occurs. 86 . • We need an ODE solver that is smart enough to stop when the reactor ﬁlls. The ODE solver needs to ﬁnd it for us. because we do not know this time t1.Semi-batch polymerization • The text solves this problem analytically. let’s solve it numerically. • Let t1 be the time that the reactor ﬁlls. Instead.

2 min • Note the reactor would have ﬁlled in 10 min if the density were constant. • The extra time reﬂects the available volume created by converting some of the monomer to polymer during ﬁlling.Finding the time for ﬁlling the reactor • The ODE solver ﬁnds the time at which VR = 20 m3 t1 = 11. • After t1 we consider the two operations. 87 .

Operation 1. • In the ﬁrst operation. Qf = 0 after t1. 88 . • Notice the reactor volume decreases after t1 because ∆V is negative.

Semi-batch polymerization 89 .

20 15 2 1 VR (m3 ) 10 5 0 0 10 20 30 time (min) 40 50 Figure 4: Semi-batch reactor volume for primary monomer addition (operation 1) and primary plus secondary monomer additions (operation 2). 90 .

1 0.2 0 10 20 30 time (min) 40 50 2 1 Figure 5: Semi-batch reactor feed ﬂowrate for primary monomer addition (operation 1) and primary plus secondary monomer additions (operation 2).2 0 -0.8 Qf (m3 /min) 0.6 0.4 0. 91 .

92 .6000 5000 monomer (kg) 4000 3000 2000 1000 0 0 10 20 30 time (min) 40 50 2 1 Figure 6: Semi-batch reactor monomer content for primary monomer addition (operation 1) and primary plus secondary monomer additions (operation 2).

12000 10000 polymer (kg) 8000 6000 4000 2000 0 0 10 20 30 time (min) 40 50 2 1 Figure 7: Semi-batch reactor polymer content for primary monomer addition (operation 1) and primary plus secondary monomer additions (operation 2). 93 .

Operation 2. • Because the reactor volume is constant. 94 . • Notice the ﬁnal polymer production is larger in operation 2 because of the extra monomer addition. we can solve Equation 52 for the feed ﬂowrate during the secondary monomer addition Qf = −∆V kM • Operation 2 is also shown in the ﬁgures.

Useful for debugging the computation. • In operation 2.Polymer production rate • We can perform an independent. simple calculation of the total polymer in operation 2. the volumes are additive.5% 95 . • The increase in production rate is P2 − P1 P1 × 100% = 22. 10 m3 of polymer are produced because in an ideal mixture. Therefore P 2 = (VR − VR0)ρP = 10 m3 × 1100 kg/m3 = 11000 kg in good agreement with the long-time solution for operation 2.

the total polymer production increases 22. 96 .5%.• By using the volume of the reactor more efﬁciently.

the velocity proﬁle is expected to be reasonably ﬂat in the radial direction except near the tube wall. • Plug ﬂow is often used to approximate ﬂuid ﬂow in tubes at high Reynolds number. The turbulent ﬂow mixes the ﬂuid in the radial and angular directions.The Plug-Flow Reactor (PFR) • Plug ﬂow in a tube is an ideal-ﬂow assumption in which the ﬂuid is well mixed in the radial and angular directions. • Also in turbulent ﬂow. • The ﬂuid velocity is assumed to be a function of only the axial position in the tube. 97 .

Thin Disk Volume Element Given the plug-ﬂow assumption. it is natural to take a thin disk for the reactor volume element Qf cjf Q cj Q(z) cj (z) z ∆V Rj Q(z + ∆z) cj (z + ∆z) z + ∆z 98 .

∂ cj Q ∂cj + Rj (55) =− ∂t ∂V accumulation convection reaction 99 .Thin Disk Volume Element • Expressing the material balance for the volume element ∂ cj ∆V ∂t = cj Q z − cj Q z+∆z + Rj ∆V (54) • Dividing the above equation by ∆V and taking the limit as ∆V goes to zero yields.

and axial length is related to tube volume by z = V /Ac . is related to volumetric ﬂowrate by v = Q/Ac .584] ∂ cj v ∂cj =− + Rj ∂t ∂z (56) 100 .Length or volume as independent variable • If the tube has constant cross section. p. Ac . v. then velocity. • Equation 55 can be rearranged to the familiar form [1.

Steady-State Operation • Setting the time derivative in Equation 55 to zero gives. d(cj Q) = Rj dV (57) • The product cj Q = Nj is the total molar ﬂow of component j. dNj = Rj dV (58) 101 . One also can express the PFR mole balance in terms of the molar ﬂow.

pressure. j = 1. the equation of state for the reaction mixture. • Because the molar ﬂow and concentration are simply related. cj ) = 0. P . 102 . molar ﬂows. ns . . . • The important piece of information tying these quantities together is. again. N j = cj Q (59) the equation of state is also a relation between temperature. . . f (T . Q. one has to be able to relate the volumetric ﬂowrate. to the molar ﬂows.Volumetric Flowrate for Gas-Phase Reactions • To use Equation 58 for designing a gas-phase reactor. and volumetric ﬂowrate. Nj . 2.

the equation of state is j Nj Q = P RT • One can solve the previous equation for the volumetric ﬂowrate. Q= RT P Nj j (60) 103 . as P cj = RT j • In terms of molar ﬂows.Ideal Gas Equation of State • The ideal-gas equation of state. c = P /RT . can be stated in terms of molar concentrations. cj .

one simply rearranges Equation 59 to obtain cj = Nj P = Q RT Nj j Nj (61) 104 .Ideal Gas Equation of State • To evaluate the concentrations for use with the reaction rate expressions.

M = j Nj Mj . cj ) • The mass density is related to the volumetric ﬂowrate and total mass ﬂow. dV in which Mf is the feed mass ﬂowrate. • The total mass ﬂow in a PFR is constant. P .Volumetric Flowrate for Liquid-Phase Reactions • Consider the equation of state for a liquid-phase system to be arranged in the form ρ = f (T . via M = ρQ (62) • Multiplying Equation 58 by Mj and summing on j produces dM = 0. 105 M(0) = Mf .

106 . then Q = Qf . • Equation 64 is used often for liquid-phase reactions.Volumetric Flowrate for Liquid-Phase Reactions • We can solve for the volumetric ﬂowrate by rearranging Equation 62 Mf Q= ρ • If the liquid density is considered constant. constant density (64) (63) and the volumetric ﬂowrate is constant and equal to the feed value. ρ = ρf .

the steady-state proﬁle in a PFR starting from a given feed condition is also the transient proﬁle in a batch reactor starting from the equivalent initial condition. dτ constant ﬂowrate (65) which is identical to the constant-volume batch reactor. we can rewrite Equation 57 as dcj = Rj . • For the constant-ﬂowrate case. if Q is constant. 107 .Volumetric Flowrate for Liquid-Phase Reactions • If we denote the time spent in the tube by τ = V /Q.

0 atm.05 sec−1 at the reactor temperature. k = 0. and the feed ﬂowrate is 35 L/min. 108 . The reactor temperature is 518◦C and the pressure is 2.5 cm inner-diameter tube required to achieve 35% conversion of A. • The reaction rate is ﬁrst order in A. Assume the pressure drop is negligible.Single Reaction Systems – Changing ﬂowrate in a PFR • A pure vapor stream of A is decomposed in a PFR to form B and C A → B+C k (66) • Determine the length of 2.

so we use Equation 60. • Substituting the production rate into the above equation gives. which assumes an ideal-gas equation of state. dNA = −kNA/Q dV (67) • The volumetric ﬂowrate is not constant. RT Q= (NA + NB + NC) P (68) 109 .Changing ﬂowrate in a PFR • The mole balance for component A gives dNA = RA dV • The production rate of A is RA = −r = −kcA.

Changing ﬂowrate in a PFR • The ideal-gas assumption is reasonable at this reactor temperature and pressure. The mole balances for B and C are dNB = RB = r dV dNC = RC = r dV • Adding the mole balance for A to those of B and C gives d (NA + NB) =0 dV d (NA + NC) =0 dV • The stoichiometry does not allow the molar ﬂow NA +NB or NA +NC to change with position in the tube. 110 . • One can relate the molar ﬂows of B and C to A using the reaction stoichiometry.

one can relate NB and NC to NA. NA + NB = NAf + NBf NA + NC = NAf + NCf • Rearranging the previous equations gives. NB = NAf + NBf − NA NC = NAf + NCf − NA • Substituting the relations for NB and NC into Equation 68 gives RT Q= 2NAf + NBf + NCf − NA P 111 .Changing ﬂowrate in a PFR • Because NA + NB and NB + NC are known at the tube entrance.

Changing ﬂowrate in a PFR • Because the feed stream is pure A. dNA P NA = −k dV RT 2NAf − NA • The above differential equation can be separated and integrated. yielding Q= RT 2NAf − NA P • Substituting this expression in Equation 67 gives the ﬁnal mole balance. NA NAf 2NAf − NA dNA = NA V 0 kP − dV RT 112 . NBf = NCf = 0.

2NAf ln NA/NAf + NAf − NA kP =− V RT • The conversion of component j for a plug-ﬂow reactor operating at steady state is deﬁned as Njf − Nj xj = Njf • Because we are interested in the V corresponding to 35% conversion of A. we substitute NA = (1 − xA)NAf into the previous equation and solve for V. RT V =− NAf [2 ln(1 − xA) + xA] kP 113 .Changing ﬂowrate in a PFR • Performing the integration gives.

35) + . it is convenient to rearrange the previous equation to obtain Qf z=− [2 ln(1 − xA) + xA] kAc • Substituting in the known values gives 35 × 103 cm3/min z=− 0.05 sec−1 60 sec/min z = 1216 cm = 12.Changing ﬂowrate in a PFR • Because Qf = NAf RT /P is given in the problem statement and the tube length is desired.2 m 4 [2 ln(1 − .35] π (2.5 cm)2 114 .

if an ideal-gas mixture is assumed. . • The cj are calculated from the molar ﬂows with Equation 59 • Q is calculated from Equation 60. ns • The reaction rates are expressed in terms of the species concentrations. . dNj = Rj .Multiple-Reaction Systems • The modeler has some freedom in setting up the material balances for a plugﬂow reactor with several reactions. . j = 1. . 2. . . . . 115 . j = 1. ns dV nr Rj = i=1 νij ri. • The most straightforward method is to write the material balance relation for every component. 2.

• Diphenyl is produced by the dehydrogenation of benzene. k1 k−1 2C6H6 2B C12H10 + H2 D + H 116 .Benzene pyrolysis in a PFR • Hougen and Watson [3] analyzed the rate data for the pyrolysis of benzene by the following two reactions.

k2 k−2 C6H6 + C12H10 B + D C18H14 + H2 T + H (69) (70) • The reactions are assumed to be elementary so that the rate expressions are 2 r 1 = k 1 cB − r2 = k2 cD cH K1 cT cH cB cD − K2 (71) (72) 117 .Benzene pyrolysis in a PFR • Triphenyl is formed by the secondary reaction.

• The reactor operates at 1033K and 1.0 atm.Benzene pyrolysis in a PFR • Calculate the tube volume required to reach 50% total conversion of the benzene for a 60 kmol/hr feed stream of pure benzene. • Plot the mole fractions of the four components versus reactor volume. 118 .

Benzene pyrolysis in a PFR • The rate and equilibrium constants at T = 1033K and P = 1. k1 = 7 × 105 L/mol · hr k2 = 4 × 105 L/mol · hr K1 = 0.48 119 .31 K2 = 0.0 atm are given in Hougen and Watson.

Benzene pyrolysis in a PFR • The mole balances for the four components follow from the stoichiometry. dNB dV dND dV dNH dV dNT dV = −2r1 − r2 = r1 − r2 = r1 + r2 = r2 (73) (74) (75) (76) • The initial condition for the ODEs are NB (0) = NBf and ND (0) = NH (0) = NT (0) = 0. 120 .

Benzene pyrolysis in a PFR

• The total molar ﬂux does not change with reactor volume. Q= RT NBf P (77)

• The rate expressions are substituted into the four ODEs and they are solved numerically. • The total conversion of benzene, xB = (NBf − NB )/NBf , is plotted versus reactor volume in Figure 8. • A reactor volume of 404 L is required to reach 50% conversion. The composition of the reactor versus reactor volume is plotted in Figure 9.

121

0.6 0.5 0.4

xB

0.3 0.2 0.1 0 0 200 400 600 800 1000 1200 1400 1600 V (L)

**Figure 8: Benzene conversion versus reactor volume.
**

122

1

0.8

0.6

yB

yj

0.4

yH yD yT

0.2

0 0 200 400 600 800 1000 1200 1400 1600 V (L)

**Figure 9: Component mole fractions versus reactor volume.
**

123

124 .Ethane pyrolysis in the presence of NO • See the text for another worked PFR example.

the steady-state PFR material balance is given by Equation 65 dcA = −r (cA) dτ 125 . • Let’s compare their steady-state efﬁciencies in converting reactants to products.Some PFR-CSTR Comparisons • We have two continuous reactors in this chapter: the CSTR and the PFR. consider a constant-density. • For simplicity. liquid A → B k n r = kcA (78) • For this situation.

Some PFR-CSTR Comparisons • We rearrange and solve for the time required to change from the feed condition cAf to some exit concentration cA cA τ= cAf 1 dcA r (cA) • The area under the curve 1/r (cA) is the total time required to achieve the desired concentration change. 1 r (cA ) PFR cA CSTR cA cAf 126 .

times the width. cAf − cA. of the rectangle. and solve for τ giving cAf − cA τ= r (cA) • This result also can be interpreted as an area. we start with Equation 37.Some PFR-CSTR Comparisons • To achieve this same concentration change in the CSTR. 1 r (cA ) PFR cA CSTR cA cAf 127 . 1/r (cA). Notice that this area is the height.

The rate is high at the entrance to the tube where the concentration of A is equal to the feed value. • Now considering that the entire volume of the CSTR is reacting at this lowest rate of the PFR. the PFR time (or volume. At the exit of the PFR. the rate is the lowest of any location in the tube. equivalently. r (cA) is an increasing function of cA. and decreases with length as the concentration drops. • The PFR reaction rate varies with length. to achieve the same conversion.Some PFR-CSTR Comparisons • If 1/r (cA) is a decreasing function of cA. it is intuitively obvious that more volume is required for the CSTR to achieve the same conversion as the PFR. 128 . or. VR = Qf τ) is less than the CSTR time (volume).

the CSTR is more efﬁcient. 1 r (cA ) CSTR cA PFR cA cAf 129 .Some PFR-CSTR Comparisons • If the reaction order is positive (the usual case). If the reaction order is negative. the PFR is more efﬁcient.

The PFR versus CSTR with separation • The PFR achieves higher conversion than an equivalent volume CSTR for the irreversible reaction with ﬁrst-order kinetics A → B r = kcA • Consider the case in which we add separation. • Find a single CSTR and separator combination that achieves the same conversion as the PFR. 130 .

131 .The PFR versus CSTR with separation • The issue is to increase the CSTR achievable conversion using separation. • Education in chemical engineering principles leads one immediately to consider recycle of the unreacted A as a means to increase this conversion.

we show how to ﬁnd the recycle ﬂowrate so this system achieves the PFR conversion. 132 .The PFR versus CSTR with separation NA0 NA VR NA0 NA1 NA2 NA VR αNA2 αNA2 pure B • In the text.

also of total reactor volume V . • This surprising principle states: For a given reaction network with ni linearly independent reactions. 133 . any steady state that is achievable by any reactor-separator design with total reactor volume V is achievable by a design with not more than ni + 1 CSTRs. • This example was motivated by a recent result of Feinberg and Ellison called the CSTR Equivalence Principle of Reactor-Separator Systems [2]. temperatures and pressures in the CSTRs are arbitrarily close to those occurring in the reactors of the original design.CSTR Equivalence Principle. Moreover the concentrations.

• In stochastic models we simulate quite directly the random nature of the molecules.Stochastic Simulation of Chemical Reactions • We wish to introduce next a topic of increasing importance to chemical engineers. • These situations are becoming increasingly important to chemical engineers as we explore reactions at smaller and smaller length scales. stochastic (random) simulation. • The stochastic modeling approach is appropriate if the random nature of the system is one of the important features to be captured in the model. • We will see that the deterministic rate laws and material balances presented in the previous sections can be captured in the stochastic approach by allowing the numbers of molecules in the simulation to become large. 134 .

Stochastic Simulation of Chemical Reactions

• For example, if we are modeling the chemical transformation by reaction of only a few hundreds or thousands of molecules at an interface, we may want to examine explicitly the random ﬂuctuations taking place. • In biological problems, we often consider the interactions of only several hundred or several thousand protein molecules and cells. • In sterilization problems, we may wish to model the transient behavior until every last organism is eliminated.

135

**Stochastic Simulation of Chemical Reactions
**

• Assume we have only a hundred molecules moving randomly in the gas phase A → B B → C in a constant-volume batch reactor. • The probability of reaction is assumed proportional to the r1 = k1xA r2 = k2 xB

k2 k1

(79) (80)

in which xj is the number of component j molecules in the reactor volume. • Note xj is an integer, unlike the deterministic model’s cj , which is real.

136

Stochastic Simulation of Chemical Reactions

• The basic idea of the Gillespie algorithm is to: (i) choose randomly the time at which the next reaction occurs, and (ii) choose randomly which reactions occurs at that time. 1. Initialize. Set integer counter n to zero. Set the initial species numbers, xj (0), j = 1, . . . ns . Determine stoichiometric matrix ν and reaction probability laws (rate expressions) ri = kih(xj ) for all reactions. 2. Compute reaction probabilities, ri = kih(xj ). Compute total reaction probability rtot = i ri.

137

Let the time interval until the next reaction be t = − ln(p1)/rtot (81) 138 . p1. 1). p2.Stochastic Simulation of Chemical Reactions 3. from a uniform distribution on the interval (0. Select two random numbers.

1) by the relative sizes of each reaction probability and then “throw a dart” at the interval to pick the reaction that occurs. In this manner. The idea here is to partition the interval (0.Stochastic Simulation of Chemical Reactions 4. all reactions are possible. but the reaction is selected in accord with its probability. Determine reaction m to take place in this time interval. p2 0 r1 r1 +r2 r2 r1 +r2 1 139 .

Let n = n + 1. Return to Step 2. j = 1. Update the species numbers for the single occurrence of the mth reaction via xj (n + 1) = xj (n) + νmj . ns 140 .Stochastic Simulation of Chemical Reactions 5. . . Update the simulation time t(n + 1) = t(n) + t. .

141 . e−rtot t is the probability that a reaction has not occurred during time interval t. • The next ﬁgure shows the results of this algorithm when starting with xA = 100 molecules.Stochastic Simulation of Chemical Reactions • If rtot is the total probability for reaction. • We will derive this fact in Chapter 8 when we develop the residence-time distribution for a CSTR.

Stochastic Simulation of Chemical Reactions 142 .

100 80 xA 60 xC xB xj 40 20 0 0 1 2 3 4 5 t 143 .

which is quite unlike any of the deterministic simulations. 144 . appearance and then disappearance of the intermediate species B. the simulation is actually discontinuous with jumps between simulation times.Stochastic Simulation of Chemical Reactions • Notice the random aspect of the simulation gives a rough appearance to the number of molecules versus time. and slow increase in ﬁnal product C. • Because the number of molecules is an integer. we already can make out the classic behavior of the series reaction: loss of starting material A. • But in spite of the roughness.

The results for 1000 and 4000 initial A molecules are shown in the next ﬁgures.Stochastic Simulation of Chemical Reactions • Next we explore the effect of increasing the initial number of A molecules on a single simulation. 145 .

Stochastic Simulation of Chemical Reactions 146 .

1000 800 xA 600 xC xB xj 400 200 0 0 1 2 3 4 5 t 147 .

Stochastic Simulation of Chemical Reactions 148 .

4000 3200 xA 2400 xC xB xj 1600 800 0 0 1 2 3 4 5 t 149 .

150 . • The biggest numerical challenge is producing the pseudo-random numbers and many well-developed algorithms are available for that task. • The computational time required for performing the stochastic simulation may. Notice that even with only 4000 starting molecules.Stochastic Simulation of Chemical Reactions • We see the random ﬂuctuations become less pronounced. the results compare very favorably with the deterministic simulation shown previously. be large. • Another striking feature of the stochastic approach is the trivial level of programming effort required to make the simulations. however.

Hepatitis B virus modeling 151 .

Hepatitis B virus modeling nucleotides −− → −− − − nucleotides + cccDNA rcDNA cccDNA envelope degraded (82) (83) (84) (85) (86) (87) rcDNA −− → −− − − amino acids −− → −− − − cccDNA −− → −− − − cccDNA envelope −− → secreted or degraded −− − − rcDNA + envelope −− → −− − − secreted virus 152 .

153 . and C is envelope.Hepatitis B virus modeling • The reaction rates and production rates for Reactions 82–87 are given by r1 r2 r3 r4 r5 r6 k1 xA k2 xB k3 xA k4 xA k5xC k6xB xC R r − r4 A 2 RB = r1 − r2 − r6 RC r3 − r5 − r6 (88) = in which A is cccDNA. B is rcDNA.

25 7. and the following rate constants xA kT = 1 xB xC T = 1 0 2 0 T (89) (90) 0.5 × 10−6 154 .025 1000 0.Hepatitis B virus modeling • Assume the systems starts with a single cccDNA molecule and no rcDNA and no envelope protein.

25 20 cccDNA 15 10 5 0 0 250 200 rcDNA 150 100 50 0 0 50 100 t (days) 150 200 deterministic avg. stochastic 155 .Average stochastic is not deterministic. stochastic 50 100 t (days) 150 200 deterministic avg.

stochastic 156 .10000 8000 envelope 6000 4000 2000 0 0 50 100 t (days) 150 200 deterministic avg.

Hepatitis B virus modeling 157 .

two representative stochastic trajectories. 158 .35 30 25 cccDNA 20 15 10 5 0 0 50 100 t (days) stoch 2 deterministic stoch 1 150 200 Figure 10: Species cccDNA versus time for hepatitis B virus model.

shows complete extinction of the virus. • Figure 10 shows two representative stochastic simulations for only the cccDNA species. 159 . • Notice the ﬁrst stochastic simulation does ﬂuctuate around the deterministic simulation as expected. • The second stochastic simulation. That is another possible steady state for the stochastic model. however.Hepatitis B virus modeling • The simulation of the deterministic model and an average of 500 stochastic simulations are not the same.

Average stochastic is not deterministic. stochastic 160 . 25 20 cccDNA 15 10 5 0 0 50 100 t (days) 150 200 deterministic avg. So the average stochastic simulation consists of 75% trajectories that ﬂuctuate about the deterministic trajectory and 25% trajectories that go to zero. it occurs for 125 of the 500 simulations. • In fact.

Summary • We have introduced four main reactor types in this chapter. • the batch reactor • the continuous-stirred-tank reactor (CSTR) • the semi-batch reactor • the plug-ﬂow reactor (PFR). 161 .

Summary BATCH d(cj VR ) = Rj VR dt dcj = Rj dt d(cj VR ) = Qf cjf − Qcj + Rj VR dt dcj 1 = (c − cj ) + Rj dt τ jf cj = cjf + Rj τ d(cj VR ) = Qf cjf + Rj VR dt ∂cj ∂(cj Q) = − + Rj ∂t ∂V d(cj Q) dV = Rj (91) constant volume (92) CSTR (93) constant density (94) steady state (95) SEMI-BATCH (96) PFR (97) steady state (98) constant ﬂowrate dcj = Rj . dτ τ = V /Qf (99) 162 .

reversible • second order. irreversible • two ﬁrst-order reactions in series • two ﬁrst-order reactions in parallel • two second-order. • ﬁrst order. reversible reactions 163 . irreversible • nth order.Summary We also have introduced some of the basic reaction-rate expressions. irreversible • ﬁrst order.

the balances must be solved numerically..Summary • We developed the equations required to compute the volume of the reactor if there is a signiﬁcant volume change upon reaction. 164 . not ﬁrst-order reaction rates). • Several of these simple mass balances with basic rate expressions were solved analytically.e. • A high-quality ordinary differential equation (ODE) solver is indispensable for solving these problems. • In the case of multiple reactions with nonlinear rate expressions (i. We require an equation of state for this purpose.

the CSTR achieves higher conversion than the same-volume PFR if the rate is a decreasing function of a component composition (n < 0).Summary • We showed that the PFR achieves higher conversion than the CSTR of the same volume if the reaction rate is an increasing function of a component composition (n > 0 for an nth-order rate expression). 165 . we introduced stochastic simulation to model chemical reactions occurring with small numbers of molecules. • Conversely. • Finally.

• With small numbers of molecules. 166 .Summary • The stochastic model uses basic probability to compute reaction rate. nonlinear hepatitis B virus model. • Two pseudo-random numbers are chosen to determine: (i) the time of the next reaction and (ii) the reaction that occurs at that time. We demonstrated this fact with the simple. however. the average of the stochastic simulation does not have to be equal to the deterministic simulation. • The smooth behavior of the macroscopic ODE models is recovered by the random simulations in the limit of large numbers of reacting molecules. A given reaction’s probability of occurrence is assumed proportional to the number of possible combinations of reactants for the given stoichiometry.

[2] M. and E. Stewart. W. Feinberg and P. B. second edition. 2001. N. Part Three: Kinetics and Catalysis. Watson. A. Chemical Process Principles. Chem. Lightfoot. General kinetic bounds on productivity and selectivity in reactor-separator systems of arbitrary design: I. John Wiley & Sons.. New York. John Wiley & Sons. 167 . Eng. Ind. 1947. Bird. 2002.References [1] R. Ellison. 40(14):3181–3194. New York. [3] O. Transport Phenomena. M. Principles. Hougen and K. Res. E.

Sign up to vote on this title

UsefulNot useful- Shashank ion of Cumene Production
- cumene
- Slides Matbal
- Isothermal Continuous Stirred Tank
- cstr 40 L
- Kinetics
- CPE624 (2)
- Levenspiel O- Chemical Reaction Engineering, 1999, 3ed
- Chemical Reaction Engineering-octave Levenspiel (Www.ebookbyte.com)
- Chemical Reaction Engineering-octave Levenspiel
- VI sem PCK
- LEVENSPIEL Chemical reaction engineering-ch1-ch2
- Book.
- Levenspiel Chemical Reaction Engineering
- Chemical Reaction Engineering
- Chemical Reaction Engineering, 3rd Edition
- Chemical Reaction Engineering-octave Levenspiel-third Edition
- DBPS Project
- Homework 2
- 5 - Reactor Design
- Table of Contents
- HYSYS_styrene_multiple_reactions2005
- DOF Analysis of Ethyl Benzene Process by Shivang Sharma
- Chemical Rxn Engg
- Microreactors, A Newe Effcient Tool for Optimum Reactor Design
- Minor Problems
- Articulo Airlift
- CH4801 Syllabus
- ion and CFD Modelling of Ammonia and Urea Production
- BITSAT2011Broucher
- Kinetic