Chemical Reactor Design-CHEM-E7135

The field that studies the rates and mechanisms of chemical

reactions and the design of the reactors in which they take place

Yongdan Li
Professor of Industrial Chemistry
Department of Chemical and
Metallurgical Engineering
School of Chemical Technology
Aalto University
Email: [Link]@[Link]
Kemistintie 1, E404

1
 3.1. Basics of Non-Ideal Flow

2
Lecture 3.1 Basics of Non-Ideal Flow

plug flow
Two ideal flow pattern
mixed flow

Real equipment always deviates from these ideals.

How to account for this?

This is what this and the following chapters are about.

3
Lecture 3.1 Basics of Non-Ideal Flow

Non-ideal flow

Stagnant regions
The area in the reactor where fluid
flows very slowly

Channeling and short-circuiting

The area in the reactor where
fluid flows very quickly

Short-circuiting :
bad design of reactor
Channeling :
bad packing of the catalyst

4 Figure 3.1
Lecture 3.1 Basics of Non-Ideal Flow

The Residence Time Distribution, RTD

It is evident that elements of fluid taking different routes through the reactor may
take different lengths of time to pass through the vessel.

Shout-circuiting
Take short time to pass through
the vessel

Stagnant regions
Take long time to pass through
the vessel

Figure 3.2

In many cases if we know the distribution of residence times of the flowing

fluid. We can predict the behavior of a vessel as a reactor.

5
Lecture 3.1 Basics of Non-Ideal Flow

E function
E represent the residence time distribution RTD of fluid. E has the units of time-l .
E dt represents the fraction of exit stream of age between t and t + dt.
The fraction younger than age t1, is

Figure 3.3
6
Lecture 3.1 Basics of Non-Ideal Flow

Experimental methods (nonchemical) for finding E

We use the pulse and the step experiments for finding the E curve.

The pulse experiment

A vessel of volume V m3
through which flows υ m3/s of
fluid. For this instantaneously
introduce M units of tracer (kg
or moles) into the fluid entering
the vessel, and record the
concentration-time of tracer
leaving the vessel. This is the
Cpulse curve.

Figure 3.4

7
Lecture 3.1 Basics of Non-Ideal Flow

Figure 3.5

From the material balance for the vessel we find

( 3.1 )

( 3.2 )

8
Lecture 3.1 Basics of Non-Ideal Flow

To find the E curve from the Cpulse curve simply divide the concentration readings
by M/v.

( 3.3 )

We have another RTD function Eθ . Here time is measured in terms of mean

residence time . Thus

( 3.4 )

9
Lecture 3.1 Basics of Non-Ideal Flow

The step experiment

Consider υ m3/s of fluid flowing through a vessel of volume V. Now at time t =0 switch
from ordinary fluid to fluid with tracer of concentration Cmax , and measure the outlet tracer
concentration Cstep versus t, as shown in Fig. 3.6.

Figure 3.6

10
Lecture 3.1 Basics of Non-Ideal Flow

The dimensionless form of the Cstep curve is called the F curve. It is found by having the
tracer concentration rise from zero to unity, as shown in Fig. 3.7.

Figure 3.7.

F(t1) represents the fraction of exit stream younger than age t1

11
Lecture 3.1 Basics of Non-Ideal Flow

Relationship between the F and E Curves

F(t1) represents the fraction of exit stream younger than age t1

represents the fraction of exit stream younger than age t1

( 3.5 )

or ( 3.6 )

12 Figure 3.8
Lecture 3.1 Basics of Non-Ideal Flow

Figure 3.9

13
Lecture 3.1 Basics of Non-Ideal Flow

Figure 3.10

14
 3.2. Compartment Models

15
Lecture 3.2 Compartment Models

In the compartment models we consider the vessel and the flow through it as follows:

Dead or stagnant region

Figure 3.11
16
Lecture 3.2 Compartment Models

bypass flow

recycle flow
Figure 3.12

17
Lecture 3.2 Compartment Models

recycle flow
combine
Compartment
bypass flow Models

Figure 3.13

By comparing the E curve for the real vessel with the theoretical curves for various
combinations of compartments and through flow, we can find which model best fits the
real vessel.

18
Lecture 3.2 Compartment Models

Figure 3.14

19
Lecture 3.2 Compartment Models

Diagnosing reactor ills

These combined models are useful for diagnostic purposes, to pinpoint faulty flow and
suggest causes.
For example, if you expect plug flow and you know , Fig. 3.15 shows what you
could find.

Theoretical residence time

Actual residence time

The volume of reactor

The active volume of reactor

Figure 3.15

20
Lecture 3.2 Compartment Models

Figure 3.16

21
Lecture 3.2 Compartment Models

Figure 3.17

22
 3.3. The Dispersion Model

23
Lecture 3.3 The Dispersion Model

Physical model
 Consider plug flow of a fluid, on top of which is superimposed some degree of backmixing.
 The backmixing is independent of position within the vessel.
 there exist no stagnant pockets and no gross bypassing or short-circuiting of fluid in the
vessel.
This is called the dispersed plug flow model, or simply the dispersion model.

Figure 3.18
24
Lecture 3.3 The Dispersion Model

Mathematical model

For molecular diffusion in the x-direction the governing differential equation is

given by Fick's law:

( 3.7 ) Where , the coefficient of molecular diffusion.

x : position in the vessel t : residence time in the vessel C : concentration of tracer

In an analogous manner we may consider all the contributions to intermixing of

fluid flowing in the x-direction to be described by a similar form of expression, or

Where the parameter D, which we call the longitudinal or

( 3.8 )
axial dispersion coeficient.

25
Lecture 3.3 The Dispersion Model

Mathematical model
In dimensionless form where and , the basic differential
equation representing this dispersion model becomes

( 3.9 ) Mathematical model

L: length of the reactor u: velocity of fluid C : concentration of tracer

Where the dimensionless group , called the vessel dispersion number, is the
parameter that measures the extent of axial dispersion. Thus

negligible dispersion, hence plug flow

large dispersion, hence mixed flow

26
Lecture 3.3 The Dispersion Model

In this model，the D or D/uL is the most important. If we know the D or D/uL, we can
evaluate the degree that a real flow deviates from plug flow.

We evaluate D or D/uL by recording the shape of the tracer curve as it passes the exit of
the vessel. In particular, we measure

= mean time of passage, or when the curve passes by the exit

( 3.10 )

= variance, or a measure of the spread of the curve

( 3.11 )

27 Figure 3.19
Lecture 3.3 The Dispersion Model

Fitting the dispersion model for small extents of dispersion, D/uL < 0.01

For small extents of dispersion (if D/uL is small) the solution to Eq. 3.9 is not
difficult.

D/uL is small
( 3.12 )

L: length of the reactor u: velocity of fluid t : residence time in the vessel

x : position in the vessel

28
Lecture 3.3 The Dispersion Model

29 Figure 3.20
Lecture 3.3 The Dispersion Model

Ways to evaluate D/uL from an experimental curve

 By measuring its maximum height

 By finding that width which includes

68% of the area.

Point of
 By width at the point of inflection
inflection

 By calculating its variance

Figure 3.21

30
Lecture 3.3 The Dispersion Model

Large deviation from plug flow, D/uL >0.01

What happens right at the entrance and exit of the vessel strongly affects the shape of
the tracer curve as well as the relationship between the parameters of the curve and D/uL.

Plug flow outside the vessel up to The flow is undisturbed as it passes

the boundaries the entrance and exit boundaries
Figure 3.22

31
Lecture 3.3 The Dispersion Model

Closed vessel, D/uL >0.01

Here an analytic expression for the E

curve is not available. However, we can
construct the curve by numerical methods.

Or evaluate its variance exactly, as was

first done by van der Laan(1958). Thus

(3.14)

32 Figure 3.23
Lecture 3.3 The Dispersion Model

Open Vessel, D/uL >0.01

(3.15)

33 Figure 3.24
Lecture 3.3 The Dispersion Model
The dispersion model with chemical reaction
Consider a steady-flow chemical reactor of length L through which fluid is flowing at a
constant velocity u, and in which material is mixing axially with a dispersion coefficient D.
Let an nth-order reaction be occurring.

(3.16)

34
Lecture 3.3 The Dispersion Model

Figure 3.25.

35
Lecture 3.3 The Dispersion Model

Figure 3.25.

Entering by axial diffusion

Leaving by axial diffusion

36
Lecture 3.3 The Dispersion Model

Entering by axial diffusion

Leaving by axial diffusion

Entering all these terms into Eq. 3.16 and dividing by S Δl gives

(3.17)

37
Lecture 3.3 The Dispersion Model

So taking limits as Δl → 0 we obtain

(3.18 )

where and
First-order reaction.

Equation 3.18 has been solved analytically by Wehner and Wilhelm (1956) for first-
order reactions.

(3.19)

38
 3.4. The Tanks-In-Series Model

39
Lecture 3.4 The Tanks-In-Series Model

This model can be used whenever the dispersion model is used; and for not too large a
deviation from plug flow both models give identical results.

Physical model
Figure 3.26 shows the system we are considering.

Figure 3.26

40
Lecture 3.4 The Tanks-In-Series Model

Mathematical model
For the first tank. Consider a steady flow υ m3/s of fluid into and out of the first of these
ideal mixed flow units of volume V1. At time t = 0 inject a pulse of tracer into the vessel
which when evenly distributed in the vessel (and it is) has a concentration C0 .

Figure 3.27

At any time t after the tracer is introduced make a material balance, thus

where C1 is the concentration of tracer in tank "1."

41
Lecture 3.4 The Tanks-In-Series Model

Separating

Integrating

(3.20)

: mean residence time of first tank

42
Lecture 3.4 The Tanks-In-Series Model

For the second tank where C1 enters, C2 leaves, a material balance gives

Figure 3.28

Separating gives a first-order differential equation, which when integrated gives

(3.21)

43
Lecture 3.4 The Tanks-In-Series Model

For the Nth tank. Integration for the 3rd, 4th, . . . , Nth tank becomes more complicated
so it is simpler to do all of this by Laplace transforms.

(3.22)

(3.23)

44
Lecture 3.4 The Tanks-In-Series Model
In this model，the number of mixed flow reactors ( N ) is the most important parameter.
It represents how many mixed flow reactor should be used to represent the real flow, and can
evaluate the degree that the real flow deviates from plug flow.
Ways to evaluate N from an experimental curve

 By measuring its maximum

height and its horizontal axis
Eq. 3.23

 By measuring its vertical

axis of inflection point

 By width at the point of inflection

Eq.
2.23
By calculating its variance

Figure 3.29
45
Lecture 3.4 The Tanks-In-Series Model

The tanks-in-series model with chemical reaction

First-order reaction
For first-order reactions in one tank

1-XA =

for N tanks in series

XA XA XA

CA1=CA0(1-XA) CA2=CA1(1-XA) CAN=CAN-1(1-XA)

Figure 3.30

CAN= CA0(1-XA)N (3.24)

46
 3.5. Macrofluid

47
Lecture 3.5 Macrofluid

Physical model

Figure 3.31

In the microfluid, individual molecules are free to move about and intermix.
In the macrofluid, molecules are kept grouped together in aggregates or packets,
and each element of fluid follows its own streamline with no intermixing with
neighboring elements.

The discussion and models before are all based on the hypothesis that fluid is
microfluid. For very viscous liquid, we usually assume it is the macrofluid

48
Lecture 3.5 Macrofluid

Mathematical model
In the macrofluid, each element can be regard as a batch reactor. Their
conversion degree depends on their residence time distribution in the reactor.

Conversion degree

(3.25)

The time they stay in the

Batch reactor reactor

mean concentration of A mean conversion of A

49
Lecture 3.5 Macrofluid

Mixed flow reactor-macrofluid.

(3.26)
Figure 3.32

This is the general equation for determining conversion of macrofluids in mixed flow
reactors, and it may be solved once the kinetics of the reaction is given.

50
Lecture 3.5 Macrofluid

For a zero-order reaction in a batch reactor.

(3.26)

(3.27)

For a first-order reaction in a batch, reactor.

(3.26)

(3.28)

This equation is identical to that obtained for a microfluid. Thus we conclude that the
degree of segregation has no effect on conversion for first-order reactions.

51
Lecture 3.5 Macrofluid
Difference in performance: macro- or microfluids

Figure 3.33 illustrates the

difference in performance of
macrofluids and microfluids in
mixed flow reactors, and they show
clearly that a rise in segregation
improves reactor performance for
reaction orders greater than unity
but lowers performance for reaction
orders smaller than unity.

52 Figure 3.33
 Stagnant regions
Non-Ideal The Residence Time Distribution
Flow Channeling RTD
short-circuiting
E or F function

The pulse experiment E function

or
The step experiment F function

Compartment E real Which compartment model

Compare
Models describes real reactor
E theoretical
 D/uL < 0.01

Closed vessel
Dispersion
D/uL > 0.01
Model
Open Vessel

XA is related to D/uL

Tanks-In-Series
XA is related to N
Model

N and D/uL can evaluate the degree that the real flow deviates from plug flow.
Macrofluid
In the macrofluid, each element can be regard as a batch reactor.

Conversion degree

The time they stay in the

Batch reactor reactor
 Chemical Reactor Design
The field that studies the rates and mechanisms of chemical
reactions and the design of the reactors in which they take place

Yongdan Li
Professor of Industrial Chemistry
Department of Chemical and
Metallurgical Engineering
School of Chemical Technology
Aalto University
Email: [Link]@[Link]
Kemistintie 1, E404

56