Kinetics of homogeneous

reactions
(part 1)
In homogeneous reactions all the reactants belong from a single
phase, which can be liquid, gaseous or solid.

rA = 1/V (dNA/dt)
= generated moles (A)/volume x time

In this equation: if A is a product of the reaction, the kinetic

rate is positive; if A is a reagent the rate is negative.

rA = f(Temperature, pressure and composition)

If the volume is a constant:

rA = -dCA/dt
Considering irreversible reactions, the reaction rate can be
expressed with respect to the concentration:

r = k CA

r = k CA CB

r = k CA2

r = k

If the reaction is reversible, the kinetic equation is:

r = kd CACB – ki CRCS
 I order reactions

-dCA/dt = k CA

Integral:

CA = CA° e-kt

This relation is useful for introducing the fractionary conversion

(with respect to the limiting reactant A):

yA = (nA° - nA)/nA°

yA = 1 – nA/nA°
Constant volume:
yA = 1 – CA/CA° i.e., CA = CA°(1 – yA)
Substituting in the differential equation :
-dCA = C0A (-dy)

− dCA 0 dy A
= CA
dt dt
dy A
C 0
A = kCA0 (1 − y A )
dt

Integral:

yA t
dy A
0 1− y = k  dt
0
 − ln (1− y ) = kt
yA
0

− ln (1− y A ) = kt
(1− y A ) = e −kt

y = 1− e −kt

I order
1
CA = C e
0 −kt
A ln = kt
1− y A
 2nd order: A + B → products

M − yA  C BC 0A 
ln
1− y B
= ln
M (1− y A )
( )
= ln  0  = C A0 (M −1 )kt = C B0 − C A0 kt
1− y A  C BC A 
C B0
M= 0
M≠1 CA

If C0B >> C0A the reaction is pseudo-first order

If 2A → products

1 1 1 yA
− 0 = = kt
0
C A C A C A 1− y A
 Opposite reactions

A B
CBeq kd
K eq = =
C Aeq ki

C 0A − C Aeq  C 0A + C B0 
ln = (kd + ki )t = k d  0 t
C A − C Aeq  C + C0 − C 
 A B Aeq 

 C0 + C0 
ln (C A − C Aeq ) Angularcoefficient = kd  0 t
A B
 C + C0 − C 
 A B Aeq 

t
Introducing the fractionary conversion:

C A = C (1− y A )
0
A

C Aeq = C A0 (1 − y Aeq )

yeq  C 0A + C B0 
= kd  0 t
kd
ln
− y A + y Aeq C +C y 
yeq
0
 B A eq 

ln(yA − y Aeq ) kd
If C0B= 0
yeq
− kd
ln (y A − y Aeq )= ln yeq t
yeq
t
 Parallel and consecutive reactions
2B
A
C → D

The yield is expressed as:

ni
i
Ri = 0
nA

The yield is reffered to the moles number of A at time 0.

 Parallel and consecutive reactions
2B
A
C → D

The selectivity is expressed as:

ni ni
i i
Si = = 0
n Areacted nA − nA

Is referred to the number of moles of A reacted

Ri n − n A
0
= 0
A
=y
Si nA

Si i =1
 Parallel reaction of equal order
kR R
A+ B

kS S

dC R
rR = = k R C AC B
dt

dC S
rS = = k S C AC B
dt
If A is the limiting reactant:

= (k R + k S )C ACB
dC A
rA = −
dt

The ratio between the first two equations results in:

dCR k R
=
dCS kS

CS
CR kR
CR0 dCR = kS  dCS Often: C0R = C0S = 0
CS0
 kR
CR = C S
kS

C R r1
=
CS r2

The ratio CR/CS is

always the same, at
any time, because it is
equal to a ratio
between constants.
 Parallel reactions at different orders

kR
A+A B 2nd order
kS
A C 1st ordine

dC A
rA = − = k1C A2 + k 2C A Eq.1
dt

dC B k1 2 Eq.2
rB = = CA
dt 2
dCC Eq.3
rC = = k 2C A
dt
The ratio between Eq. 2 and 3 is:

dC B k1C A2 k1C A
= =
dCC k 2 C A 2k2

In that case the ratio value is not constant. When CA

diminishes with respect to the time, also the ratio dCB/dCC
diminishes proportionally.
The higher kinetic order reactions are penalized (the
generating rate (B) diminishes much more than A).
Integral of Eq.1:

t
dCA
CA k1C A2 + k2C A = 0 dt
CA
0
−
 k2
CA =
( )
k1C A0 + k 2 e −k2t − k1C A0

C
C0A

t
In order to determine the products distribution it is
necessary to calculate the ratio between Eq. 3 and 1:

− dCC k2CA k2
= k C2 + k C =
dC A 1 A 2 A k1C A + k 2

Integral:
CA
−CC0 dCC = 
CC k 2dC A
k C + k2
C0 1 A
A

CA = f(t)
Because: CC = f(CA) = f(t)
 A + k2
0
k k C
CC − CC = ln
0 2 1
k1 k1C A + k 2

In order to obtain CB it is possible to perform a mass

balance:
C +
0 C =
+ C 0
= C +C +C
A B C A B C

C 0A − C A − CC Con C0B = C0C = 0

CB =
2
C
CC

CB

t
 Consecutive reactions
k1 k2
A B C

dC A
rA = − = k1C A = k1C A0 e −k1t
dt

dC B
rB = = k1C A − k 2C B
dt

dCC
rC = = k 2C C
dt
Substituting CA in the 2nd equation:

dCB
+ k 2C B = k1C A = k1C A0 e −k1t
dt

+ P(x)y = Q(x)
dy
This is a differential equation, like:
dx

Production of B from A
By solving it occurs:
A→B
B distribution
CB = C B0 e −k1t +
k1C A0
k 2 − k1
(−k1t
e −e )
−k 2t
B→C

Formation of B from B already present at time 0(C0B≠0)

For determining CC is just necessary to perform a mass
balance
C 0A + C B + CC = C A + C B + CC
= 0

C C
A
CC

CB

t
Special cases:
The
k2/k1=1 inflection
point of CC
is exactly on
CA curve

The maximum of B
varies the position
with lowering the
k2/k1= 10 temperature and
concentration.
B is an unstable
intermediate, the
instability
increases when
k2>>k1.
Heterogeneous reactant systems
Problems
 Fluid-particle reactions

A (fluid) + bB (solid) → fluid products

solid products
fluid products + solid products

During the reaction, the solid particles maintain the

same dimension if they contain a great amount of
impurities or if they produce solid products.
The particles dimension diminishes when ashes are
produced or when B is pure.
 2 ZnS + 3 O2 → ZnO (s) + 2 SO2

C (s) + O2 (g) → CO2 (g)

NaNH2 (l) + C (s) → NaCN (l) + H2 (g)

 Fluid-particle reactions

Progressive conversion model

The reactive gas enters in the particles and reacts in any inside
the particle, the reaction rates depends from the specific
position inside the particle. In that way the reactive gas is
continuously and progressively converted across the particle.

Not-reactive core model

The chemical reaction occurs at first on the particle surface;
then the reaction area moves inside the solid particle and leaves
behind completely converted materials or inert materials (ashes).
Not-reactive core model for spherical particles with
invariable dimensions.
Stadio controllante: diffusione del reagente A gassoso
attraverso il film circostante la particella fino alla superficie
del solido.
Let’s consider the external surface of the particle, Sext,
which has constant dimensions, as the free surface area.
A + bB → products
Then dNB = bdNA

1 dNB 1 dNB b dNA

−  =−  =− 
Sest dt 4R2 dt 4R2 dt

= b kg (CAg – CAs)
kg= mass transfer coefficient between the fluid and the
solid particle.
however CAs = 0
Moreover it is possible to determine NB

NB = ρB (mol B /cm3) · V (cm3)

ρB solid density
Then:
-dNB = -bdNA = - ρB dV = - ρB d(4/3 π r3c) = - 4π ρBr2cdrc
By introducing this expression in the reaction rate
equation it is possible to obtain:

1 dNB B rc2 drc

−  =− 2  = bkgCAg
Sest dt R dt
Integral:

B 2
rc
t
− 2  rc drc = bkgCA  dt
g
R R 0

B R   rc  
3
t=  1−   
3bkgCAg   R  

When particle reaction is completed rc = 0

B R t   rc  
3
= = 1−   
3bkgCAg    R  
Limiting stage: diffusion across the ash film
The reaction rate of A, in any time, derives from the
diffusion rate across the reactive surface:

− = 4r 2QA
dNA
dt

dCA
QA = D e Fick’s law
dr De= Effective diffusion coefficient
across the ash film.
It is possible to obtain:

dNA dCA
− = 4r De
2

dt dr

rc CAC =0
dNA dr
− =  2 = 4r De  dC A
2

dt R
r C Ag =CAs

dN A  1 1 
−  −  = 4DeCAg
dt  rc R 
We also know that:

-dNB = -bdNA = - ρB dV = - ρB d(4/3 π r3 ) = - 4π ρ r2 dr

c B c c

By substituting and making the integral:

1 1
rc
t
− B  − drc = bDeCAg  dt
R  rc R  0

BR 
2
 rc 
2
 rc  
3
t= 1− 3  + 2  
6bDeCAg   R   R  
 B R 2
=
6bDeCAg

t   rc 
2
 rc  
3
= 1− 3  + 2  
   R   R  

It is also possible to obtain these expression in function of y:

t  rc 
3

=y   = 1− y
 R
Limiting stage: chemical reaction
The reaction rate doesn’t depend from the presence of an
ashes film.

1 dNB
−  = bk s C Ag
4rc
2
dt

ks = rate constant of the 1st order for the surface

reaction.
By substituting the expression of NB in function of ρB:

1 2 drc
−   B 4rc = bk s C Ag
4rc
2
dt
simplifying:

drc
−  B = bk s C Ag
dt
rc t
−  B  drc = bks C Ag  dt
R 0

B B R
t= (R − rc ) =
bk s CAg bk sC Ag
t rc
= 1− = 1− (1− y)1/ 3

 R
 Not-reactive core model for spherical particles of
diminishing dimensions.

Limiting stages:

• Diffusion of the reagent A from the gas bulk to the

gaseous film across the solid surface.

• Surface reaction of reagent A and the solid.

• Diffusion of the products from the solid surface through

the gaseous film until it reaches the gas bulk.

An ashes film doesn’t exist.

 The global
reaction occurs
1 2 3 in 3 stages.
Resistance summatory:

ttot = tsolo film + tsolo cenere + t sola reazione

− 1 dN B bC A
=
Sest dt 1 + R(R − rc ) + R 2
2
kg rc De rc k s

Regarding particles without ashes:

− 1 dNB bCA
=
Sest dt 1 1
+
k g ks
 Fluid-fluid reactions

Problems:
• The global rate expression should consider that the two
fluid phases must enter in contact before the chemical
reaction occurs. Thus both the reaction rate and the mass
transfer rate are used.

• Solubility of the reactants: the reaction can occur in one or

in both the two phases depending from the solubility of the
reactive components in each phase.

• Contact conditions.
Case A Case B
Istantaneous reaction with Istantaneous reaction: high CB
respect to the mass transfer.

Case D
Case C Quick reaction: high CB,
Quick reaction: 2° order pseudo-first order kinetic
kinetic

Case H
Case E and F
Slow reaction with respect to
Intermediate rate with
the mass transfer.
respect to the mass transfer
 Kinetic expression for instantaneous reactions: case A
Let’s consider an infinitely fast reaction of any kinetic order:

A (gas) + bB (liquid) → products

If CB is small, the situation is represented in Fig. 1. In stationary conditions the flow

«B» is b times bigger than the corresponding flow «A»; thus the rate of disappearance
of A and B is:

A in gaseous film A in liquid film B in liquid film

The constants here reported are defined as MASS TRANSFER COEFFICIENTS, both
in liquid and gaseous phases. They regard only the mass transfer, without considering
the chemical reactions!
In corrispondence of the interface, the relation between pA e CA is provided by the
distribution coefficient, also known as HENRY COEFFICIENT:
Since the mass transfer across a film is regulated only by a diffusional process, it is possible
to state the following relation between the mass transfer coefficients A and B:

By deleting the unknown factors x, x0, pAi, CAi from the previous equations, occurs the
following equation:

When the resistance in gaseous phase is negligible, occurs:

By comparing the last equation with the expression of the maximum rate, when only
mass transfer occurs, it is possible to obtain the following relation:

The term in brackets represents the rate increment due to the absorption of A,
related to the presence of B in liquid phase.
It is possible to define:

Increment factor

𝑟𝑎𝑡𝑒 𝑖𝑛 𝑝𝑟𝑒𝑠𝑒𝑛𝑐𝑒 𝑜𝑓 𝑎 𝑐ℎ𝑒𝑚𝑖𝑐𝑎𝑙 𝑟𝑒𝑎𝑐𝑡𝑖𝑜𝑛

𝐸=( )
𝑟𝑎𝑡𝑒 𝑖𝑛 𝑝𝑟𝑒𝑠𝑒𝑛𝑐𝑒 𝑜𝑓 𝑡ℎ𝑒 𝑜𝑛𝑙𝑦 𝑚𝑎𝑠𝑠 𝑡𝑟𝑎𝑛𝑓𝑒𝑟
 Kinetic of suspension reactions

A gas containing the reactant A is bubbled across a liquid B, which

contains a suspended solid catalyst. The reactant A must reach the
catalyst surface in order to react with B.
Series reactions:

• The reactant A must arrive to the gas-liquid interface.

• A must cross the liquid film in order to arrive to the liquid

bulk.

• A must cross the liquid film around the catalytic particle in

order to reach the catalytic surface.

• A reacts upon the catalyst surface with the B component,

which is in the liquid phase.
SINGLE-PHASE BATCH REACTOR
It is a container that contains the starting
reagents. Temperature and the other variables are
such that each reaction can continue until the
wanted conversion.

V, T(t), Ni(t)
Reaction volumeV;

heat exchange surface S; reaction

Reagent fluid temperatureT(t); Ni(t) matter amount
(i species)
Heating fluid (or coolant)
 SINGLE-PHASE BATCH REACTOR
Time variation Space variation
C/T C/T

t r

Main features

•simple apparatus •only indipendent variable = time

•Inefficiency but simplicity •useful for reactions with high added
value

Symbols: t reaction time; Tatemperature; Xi i species conversion; ij i

species stoichiometric coefficient in the j reaction; subscript “o” refers to
the starting value.
 MATTER BALANCE

General structure E -U = A -
G
Batch reactor general dN i (t )
structure application 0 - 0 = RiV
dt

species “i” accumulation in t reaction time

species “i” generation in the reaction volume V

 dN(t
i = RiV
Then ) dt

If in the system Nr reactions occur is

Ri is the species ima generation rate (it can be negative as well)

rj is the reaction jma rate (always positive)

The “i” species conversion definition is:

Then
t t
Integrating t Batch reactor performance
equation

t is the required reaction time to have the Xfi final conversion

(V is in the integral funztion because it could be both constant or variable with the
conversion)

If the reaction volume V is constant during the reaction, and because Noi/V is the
species “i” starting concentration that could be symbolized Coi, the performance
equation is:

 
Noi Xfi dX X dX i
=-  i =Coi -
t fi

V X oi Ri X oi Ri
 Performance equation geometric


t = XdX i
- fi

1/(-Ri) Coi Xoi Ri

t
Coi

Area below the curve

1/(-Ri) vs X
0 Xoi Xfi 1
Xi
 Deriving the batch reactor performance
t equation, it is assumed that the volume is
constant during the reaction.

But volume could change with the variation if the conversion.

The causes of this volume variation are two:
1. External causes (ex. Extra reagents such as in the semi-batch);
2.Internal causes (ex. Change of the mole number in the reagent system during
the reaction)

Bacause of the external causes are due to external situations, the internal ones are
the most interesting because they are due to the reaction evolution.

The following Chemical reaction

Example: PCl5 PCl3+ Cl2

takes place with an increase of the moles number; the reagent
system volume increases as well with an increase directly
proportional to the reaction progress.
 Variable reaction volume
Considering the generic reaction : AA+ BB QQ+ SS

Reagent conversion definition

Reaction progress definition:

 From the reaction conversion and the reaction progress:

 σ A ++σ+B σ Q σ S 
Member to member sum = + N oA  A
T oT
 σA 
is:
NT= Ni total number of the moles in the reagent system, instant t;
NoT= Noi total numeber of moles in the reagent system, instant t0;

Divide both for NoT and obtain:

 is ε =y δ Called volumetric
A oA A expansion coefficient

Using the law PV=ZNRT (Z compressibility factor)

PV Z o N oT RTo V Z N RTo Initial time (t = 0)

o o o oT
o
P== RT
PV ZN RT  V ZN Final time (t = tf)
=
T
=
T
P

Member to member division

For isobaric and isothermal systems is T=To e P=Po, and ZZo

 Being XA=1
VX =VX =  (1 + ε A )
A =1
0
V=VX e Vo = V X
A

A ==1 A= 0

or, extracting the volumetric expansion coefficient A

The volumetric expansion coefficient is the volume fractional variation in

the reaction when it proceeds to completion.
 Kinetic experiments
differenzial procedure
1. Plor the experimental data Civs. ;
2. Extract Ri from the interpolating
equation or from the graphic
interpolation ;
3. data Rivs.  can be used to obtain the
kinetic parameters. Better is the
“fit”, better is the chosen kinetic
model.
integral procedure
1. Chose a kinetic model;
2. Integrate the chosen kinetic model;
3.Set the integral equation in a linear
form. If the experimental data are
aligned, the choice is correct. The
linear regression of the data leads to
the kinetic parameters.
 Integral analysis I example

Kinetic scheme A → B Conceivable kinetic model -rA= -kCA

dC integration  C 
A = - kC
Ln oA= kt
dt A
C A 
If Plotting Ln(CoA/CA) vs. t you get a sraight
line, the kinetic model is correct.
Reaction rate
Phenomenological
expression Ln(CoA/CA) CA t
C1 t1
C2 t2
Generation rate defintion 2
expression 1

t
 Integral analysis II example

Kinetic scheme 2A → B Conceivable kinetic model -rA=-kCA2

1 1
dC A = - kC 2 integration - C = - kt
dt A
C oA
A
1/CA t

slope=k
2
The straight line in the 1/CA vs. t
CA t plot confirms the good assumption
1 C1 t1 for the kinetic model.
C2 t2
1/CoA

t
 Not-isothermal batch reactor

The mass balance with the null incoming and outgoing flows dN (t )
coincides with this expression: i = r iV
dt

Energy
balance

“i” is the subscript for the generic reagent, so i=-1, Ri=-r

dN i ) (θ dX (θ )
then rV = - R i V = - = N oi i
dt dt
 t

dt

U=0 for adiabatic operation, so dt

That is the maximum achievable

temperature (for exothermic process) for an
adiabatic system with a complete
conversion.
 Operations Optimization
CONSIDERED SITUATIONS

1 2

OPTIMAL TIME BEST VARIABLE PROFILE

CALCULATION FOR THE CALCULATION (usually temperature)
OPERATION MANAGEMENT FOR THE BEST OPERATION
MANAGEMENT
DATA :
Wi species ima unit price
Xi reagent imo conversion
 reaction jma progress degree
 Purpose of a chemical reaction
management
CH3 COOH

CH3 COOH

Para toluene. Low value terephthalic acid. High value

The purpose of the chemical reactor management is evolving a reaction

in which the products have an higher value than that of the starting
reagents.
 Optimization aims
Higher conversion and lower unit costs
increase
1: Higher reactor volume and higher costs
reaction
time Lower conversion and higher unit costs
decrease
Lower volume reactor and lower costs

Starting phase: an higher temperature increases the reaction rate and

2: optimum decreases the volume to obtain the stated conversion.
variation
Ex. Exothermic
reactions Final phase: a lower temperature increases the maximum obtainable
equation conversion, i.e the equilibrium conversion

t = time = θ
 N
=c W  ()
t) 
From the “j” reaction
progress definition j is N i - N oi = R σ ijξ j
i i
i=1 j=1
species “i” starting moles

species “i” moles Stoichiometric Coefficient species

species “i” value ima in the reaction jma

Cost variation Reaction progress jma

Maximum increase of the imo component value in

the jma reaction (constant value for each reaction)
 Single Reaction
Se nel sistema evolve una singola NoAX
reazione interessante il reagente A
“A” ed il prodotto “B” è
σA
ξ-=
N XoA
quindi ) ΔW ξ ΔW
W(θ σ
- A
A

dW(θ ) N dX A
Derivando rispetto al tempo di
= - ΔW oA
reazione  si ha == dθ σ A dθ
dX
E ricordando che - N oi  = R iiV
dθ
dW(θ )  RA
Si potrà scrivere =ΔW V
dθ σA
 Operating costs
The reaction management in a batch reactor consists in different steps
with a proper operating cost.
Aim: maximum profit investigation
 Profit optimization
per unit of time

Massimo della curva W()-WTvs. T

W()-WT

T
Wo
o ’R
Wo+Wp
p+ o R
Wo+Wp+ Wa o+ p+ a+
 Temperature Optimization

The search of the optimum temperature for the batch

reactor management deals with two different problems:
BEST SINGLE BEST TEMPERATURE
TEMPERATURE PROFILE RESEARCH
RESEARCH

For irreversible reactions Per reazioni irreversibili

(esothermic and endothermic) and for endothermic
reactions (even reversible), the optimum temperature is the
highest possible for the used materials (costruction
materials, catalysts, reagent species).
Temperature optimization
 In these three equations there are 4 variables :
XA “A” reagent conversion;
 Reaction time in the batch reactor (s);

q exchanged thermal power (cal/s o W);

T reaction temperature (°C o K).

Only one can be chosen as the indipendent variable.

the choose conditions the solution.
1° case: set conversion (find the optimal temperature for the
reaction)

dX A
The t = C   -
integrate
t=f(T)
R i (X
) A ,T
f
oA 0

dt = = 0  Tott
Using the optimal
dT temperature:

Solving the mass balance dXA

t =CoA 0 f -
Ri (X
)fA,Tott
equation, tott :
ott

Solving the energy balance

equation, qa-ott :
2° case: stated container (you know the reaction time, you must find
the optimal temperature)

Integrating the performance equation t= f(X i ,T)  X = g(t ,T)

from the mass balance:

dX
= You obtain the temperature to have the
dT maximum conversion

from Tott and from the mass balance it is possible to obtain the conversion
(the maximum possible), from the energy balance it is possible to obtain the
heat amount to exchange to obtain the optimal reaction temperature.
 CONTINUOUS STIRED TANK
REACTOR (CSTR)
IDEAL REACTORS WITH INFINITE STIRRING

MIX=0 MIX= 
IDEAL FLUID DYNAMIC IDEAL FLUID DYNAMIC

REAL FLUID DYNAMIC

PFR CSTR
 TANKS CHARACTERIZED BY AN IDEAL
FLUID DYNAMIC, I.E., PERFECTLY STIRRED,
HAVE THE FOLLOWIGN PROPERTIES:

1. THE MIXTURE FEATURES (such as T and C) ARE THE SAME IN

EVERY POINT OF THE TANK;
2. THE FEED MIX WITH THE CONTENT OF THE TANK
INSTANTANEOUSLY;
3. THE FEED DOENSN'T MODIFY THE COMPOSITION AND THE
OTHER FEATURES OF THE MIXTURE CONTAINED IN THE
TANK;
4. THE OUTLET FLOW HAS THE SAME FEATURES OF THE
MIXTURE CONTAINED INSIDE THE TANK.
 SCHEMATIC REPRESENTATION

Cei is the concentration of the “i” component

inside the inlet reacting mixture [mol/L] ENTRATA
ve is the volumetric flowrate of the inlet
reacting mixture [L/s]
Cui is the concentration of “i” inside the
outlet mixture [mol/L]
vu is the volumetric flowrate of the outlet
USCITA
mixture [L/s], at the end of the chemical
reaction

Ci is the concentration of the component “i” FLUID DYNAMIC

inside the reacting mixture [mol/L] CONDITIONS:
Cui=Ci(Tui=T)
V is the volume of the reacting mixture [L]
 CSTR
Continuous reactor with
stirring in stationary regime
• The residence time is the same for each volumetric component.
• The stirring is perfect.
• The reactor content is uniform in every point and the composition of the outlet
flow is the same of the one of the fluid inside the reactor.

Since the composition is uniform in the whole reactor, the balance can be referred
to the entire tank.

Inlet moles= outlet moles + disappeared moles due to chemical reactions

+ accumulated moles

Inlet flowrate = F0A= C0Av0

Inlet flowrate= F0A

Outlet flowrate= F0A (1 - XA) = FA

Quantity of A which disappears due to chemical reactions = (-rA) V
By substituting in the balance equation:
F0A = F0A(1 - XA) + (-rA) V

F0AXA = (-rA) V

Occurs:

V
=
 XA
=
FA0 C 0A − rA

V C A0 − C A = apparent residence time or refill

= =
v0 − rA time

It is possible that the conversion modifies the volume of the reacting system
(linearly):
V = V0 (1 + AXA)
 = Expansion coefficient
Saturated vapor

Condensation
TEMPERATURE
PROFILE IN A
NOT-STIRRED
Edge distance TANK
Saturated vapor

Condensation
TEMPERATURE
PROFILE IN A
INTENSIVE
STIRRED TANK
Edge distance
 Condensation
Saturated vapor

TEMPERATURE
PROFILE IN A
INFINITE STIRRED
TANK
Edge distance
 Temperature profile with
infinite stirring

Intensity
of the Temperature profile with
growing
stirring intense stirring

Temperature profile without

stirring
Edge distance
 GRAPHICAL MEANING
1
−r PFR
i

C
ui
C
ei
C
i
• CSTR IN STATIONARY CONDITIONS IS ISOTHERMAL.
• THE MASS BALANCES AND THE ENERGY BALANCE
CAN BE SOLVED SEPARATELY.
•THE MASS BALANCES ARE NEEDED FOR THE DESIGN
OF THE REACTOR OR FOR THE WORKING TEST; THE
ENERGY BALANCE IS NECESSARY TO DETERMINE THE
THERMAL CONDITIONS RELATED TO THE
PERFORMANCES CHOSEN IN THE MASS BALANCE.
 ENERGY BALANCE
E – U = A-G + Q

• E inlet flow

• U outlet flow

• A accumulated

• G generated

• Q exchanged (outlet)
 cpi = heat specific capacity
It is possible to assume: Hi c pi T−T=o of the component i
To = reference temperature

We can decide that To =Te

When the system is under perfect stirring Tu =T

THE ENERGY BALANCE CAN BE

EXPRESSED AS:
 THE EQUATION CORRESPONDING TO THE ENERGY
BALANCE IS:

i is the ratio between the outlet molar flowrate of the component

i and the inlet molar flowrate of the reactant A; the formule is:
i =Fui /FeA

Ta room temperature.
Te temperature of the inlet stream
 The system can be analyzed in a graphic form both for the
working tests and the parameter check.

Design ( determination)
Experimental data R/C
Test (Cui determination)
R
1/-R

area=
pendenza 

Cu Ce C Cu Ce C

Starting from Ce, a line with slope  is

Between the abscissa Ce and
outlined, the interception with the
Cu it is possible to outline a
curve points out a coordinate R vs C,
rectangle, whose height is
the corresponding value on the abscissa
1/Ru , the rectangle area
represents Cu
corresponds to 
 It is possible to check the accuracy of the kinetic equation
by a graphical test. If the kinetic equation is correct we will
obtain a straight line.

For example, when the kinetic is supposed to be of the 1st

order, the CSTR equation is the following:

When the experimental data CeA - C

primo ordine
(CeA -CA )/CeA vs.  are plotted and a CA
straight line interpolates the data,
the kinetic order supposed is
correct; moreover the slope of the
line corresponds to kinetic
constant k.

Se invece l’ordine di reazione non è definito ma la velocità di
reazione può essere assunta essere della forma r=kCn, un
diagramma come quello in figura consente la derivazione sia
dell’ordine di reazione che della costante cinetica
ordine "n"
Ln CeA-CA

 pendenza=n

Intercetta = Ln(k)

LnC
A

Se r ha una formulazione più complessa occorre trovare il

modo di linearizzare l’equazione in modo tale da poter
diagrammare i dati sperimentali nella forma opportuna
Esempio. Importanza della Selettività
L’ossido di etilene viene prodotto per ossidazione
parziale dell’etilene.
Se l’ossidazione dell’etilene è troppo spinta può aversi la
combustione completa con produzione di CO2 ed H2 O

CO2 + H2 O
CH2 ===CH2 + O2
CH2 CH2
O

L’interesse del progettista dovrà essere rivolto alla

massimizzazione della quantità di ossido di etilene
prodotto
 PARALLEL REACTIONS

B reaction rate r=
1
k 1
C α
A
1

A
C reaction rate r=
2
k 2CαA 2

The selectivity toward product B with respect to product C is defined as:

OBJECTIVE: MAXIMIZE SB/C

Function of the only concentration

Function of the only temperature

 STUDY OF THE INFLUENCE OF THE CONCENTRATION
ON THE SELECTIVITY OF PARALLEL REACTIONS

The term influenced by concentration is:

𝑎
𝑪𝑨𝟏 − 𝟐

The selectivity doesn’t depend from concentration; in order to modify the

𝛼1 = 𝛼 2 selectivity we have to vary the temperature.

The selectivity increases when concentration decreases. In order to have the

𝛼1 < 𝛼 2 maximum selectivity, the concentrations should be the minimum.

𝛼1 > 𝛼 2 The selectivity increases with concentration. In order to have the maximum
selectivity, it is necessary to work with the highest concentrations
 CHOICE OF THE REACTOR FOR
THE MANAGEMENT OF
PARALLEL REACTIONS

1st situation: the reactor choice doesn't influence the

selectivity;

2nd situation: a CSTR, which always operates at the

minimum concentration of reagents (the final
concentration), is the best choice;

3rd situation: a PFR, which operates at every concentration,

between the maximum one (inlet concentration) and the
minimum one (outlet concentration), is the best choice.
 CHOICE OF THE REACTOR FOR THE
MANAGEMENT OF SERIES
REACTIONS
In the case of series reaction it is necessary to maximize the concentration
of the interesting product (in that case B, because obviously we are not
interested in C…)
During series reaction, two factors influence the concentration of B product:
1. The reaction N° 1, which is responsible of B formation, so increases the
concentration of B
2. The reaction N° 2, which is responsible of B consuming, so diminishes the
concentration of B

In that case is not important the choice of the reactor, but the choice of the
reaction time, which has to be chosen in order to maximize the concentration
of B.
Evolving reaction in a PFR Evolving reaction in a CSTR
Optimal reaction time can be obtained considering: Thus
occurs:

PFR CSTR
1
τ p_ott =
k ml
Once the optimal reaction times are obtained, it is
possible to gain the maximum concentration of
product “B”
 TEMPERATURE EFFECT

The optimal temperature will be the one which is able to

maximize the concentration of B with respect to the
concentration of the product C; i.e., the temperature value
must maximize the following ratio:

Constant with respect to T variation

This term depends from temperature

GRAPHICAL ANALYSIS OF THE
TEMPERATURE EFFECT

Influence of temeprature (T) on 𝐤 𝐄𝟐 − 𝐄𝟏

𝐒 = 𝟎𝟏 ∙ 𝐞𝐱𝐩
selectivity (S) 𝐤𝟎𝟐 𝐑𝐓

Temperature area more suitable when E2>E1

Temperature area more suitable when E2<E1

When E2 >E1 is better to use low temperature, when
E2 <E1 is better to use high temperature; then we can
conclude that high temperatures promote chemical
reactions characterized by the higher activation
energy.
The mathematical model of an ideal reactor provides a system
constituted by the performance equations deriving from the mass
balances (NC ) e by the performance equation deriving from the
energy balance (1).
For CSTR, performance equations deriving from mass balances can be
solved separately from the ones deriving from the energy balances.
By solving the first equations it is possible to determine what we
can obtain from the process, by solving the second equation we
can determine the specific conditions necessary for obtaining the
desired product.

The possibility to solve the mass balance equations separately

from the energy balance equation is an advantage, since the first
ones are linear while the second is not.
 Reaction: A → B diagram

PERFORMANCE EQUATION

Heat generated from the reaction Total heat removed

(not linear with respect to T) [qG] (linear with respect to T) [qR]
 REMOVED HEAT(qR )
r qR
3 1 outgoing heat. Straight line
with slope Fi cPi
1
2 Heat transferred with the
environment. Straight line with
2 slope UA

3 Total heat removed. Straight

line with slope Fi cPi +UA
T
 r qU
When the removal efficiency decreases
also the slope Fi cPi +UA decreases.
The reason can be a diminishing of the
quantity of outgoing heat in the outlet
stream or a diminishing of the overall
heat transfer coefficient.
Slope of the straight
line corresponding to
the total heat removed
Fi cPi +UA
When the removal efficieny increases,
also the slope Fi cPi +UA increases.
The reason can be an increase of the
quantity of outgoing heat in the outlet
stream or an increase of the overall heat
transfer coefficient.
 GENERATED HEAT (qG )
Starting from low temperatures the following phases
occur

I phase: the temperature is low, so

the reaction rate is low and the II phase: the temperature is medium, but
quantity of heat to be removed is a temperature increment is sufficient to
small. speed up the reaction rate a lot, so it is
qG necessary to remove a huge quantity of
heat.
III fase
II fase

III phase: the temperature is very high, so

the reaction rate is high enough to permit
the achieve of the maximum conversion in
I fase
short time, in that case the quantity of heat
to be removed increases only a little.
T
Generated heat depends from c

qG

c increases, at the same

temperature, the generated
heat diminishes.

c diminishes, at the same

temperature, the generated
T heat increases

The optimal temperature is those in which the quantity of

heat removed qR is the same of the quantity of heat
generated qG . In the graph it correspond to the abscissa
value of the intersection point between the line of the
removed heat and the "S" curve of the generated heat.
 EQUATION SOLUTION

q
CURVE OF STRAIGHT LINE
GENERATED OF THE
REMOVED HEAT
HEAT

Solutions 1,2,3

T
T1 T2 T3
Situation 1 and 3

The solution for the equation is provided by

the couple T* q*.
q1R
If, for any reason, the temperature decreases,
until it reaches, for example, the value
q1G corresponding to T1, the removed heat (q1R)
q* and the generated heat (q1G) diminish, but
q1G q1R<q1G, so the temperature tends to grow
and the system will return to T*
q1R

T1 T* T2

If, for any reason, the temperature increases until the value T2 , both the
removed heat (q1R) and the generated heat (q1G) increase, however q1R> q1G , so
the temperature tends to diminish and the system will return to T*

We can conclude that solution like type 1 or 3 are intrinsically stable, since
the system tends always to compensate the variations.
 Situation 2

The solution for the equation is provided

by the couple T* q*.
q1G it reaches, for example,If,the
decreases, until forvalue
anycorresponding
reason, the temperature
to
T1, the removed
q1R heat (q1R) and the generated heat (q1G) diminish, but
q1R<q1G, so the temperature tends to diminish further and the system
recedes from
q* T*.
q1R
q1G

T1 T* T2

If, for any reason, the temperature increases until the value T2 , both the removed
heat (q1R) and the generated heat (q1G) increase, however q1R< q1G , so the
temperature tends to increase further and the system recedes from T*

We can conclude that solution like type 2 are intrinsically unstable, since the
system tends to promote variations.
 Variation in removal efficiency
qR
T T
T1
The removal efficiency increases (for
example, FA or U increase), only one
working condition occurs (the one
corresponding to T1 ).
Since temperature T1 is low, the
reactor allows only a low conversion.
qR
T1
The removal efficiency is too low, a fouling
phenomenon on the heat transfer surfaces
and the consequent U decrease can be
responsible of that.
The final conversion is high, but also the
temperature is very high and it can be
responsible of further damages of the
operation materials.
 Variation of the filling time
qG qG

T T

If c increases, maintaining the same If c decreases, maintaining the same T,

T, the heat generated diminishes;
the possible original solutions (that
were three) become only one: low
temperature, i.e., low conversion.
the heat generated increases;the possible
original solutions (that were three)
become only one: high temperature, i.e.,
high conversion. However higher
temperature are detrimental for operation
materials.
 PLUG FLOW REACTOR
(PFR)
IDEAL REACTOR, MIXING IS NULL

MIX=0 MIX= 
IDEAL FLUID DYNAMICS IDEAL FLUID DYNAMICS

REAL FLUID DYNAMICS

PFR CSTR
 PROPERTIES OF THE CONTAINERS WITH FLUID DYNAMICS IN
WHICH MIXING IS NULL

1. Each fluid thread flows parallel to itself;

2. Each fluid thread flows separate from the others (no dragging);
3. No axial diffusion;
4. Plate velocity profile.

Points 1 and 2 are not compatible with the turbulent movement.

Point 3 refers to a slow movement in which the turbulent movement
is negligible;
Point 4 refers to an highly development turbulent flow.
The PFR movement is unfeasible.
 Adopted scheme I

Heat exchange fluid

Suppose that through

ve Reagent fluid vu the side walls there
are heat exchanges
only, no matter
exchange.

0 the only variation direction

 Adopted scheme II
The features (T e C) of the flow that goes through a PFR can
change in the axial direction only. So, a good scheme has to be
directed to a flow in a tubular duct.

No radial diffusive movement, because of the absence of

gradient.
No axial diffusive
movement, also with
gradient.
Plate concentration profile in
each section, but different from
section to section.
The supposed plate velocity profile can occur in a
tubular system in which there is a turbulent flow.

Hydrodynamic Boundery Layer (HBL)

velocity profile with Core Turbolence (CT)

turbulent flow

Excluding HBL with the thickness  in which the velocity

changes from zero to maximum, in the CT section the velocity
profile is plat and it can be considered constant.

The thickness of the boundary layer decreases

L with the increase of the Reynolds numeber
δ=
Re (Re), which increases with the increase of the
velocity as the movement becomes turbulent.
 Profile
study
The Thickness  of the HBL layer
decreases with the increase of the velocity.

With a complete high speed turbulent movement (without

entrances, shrinkage and curves) (high Re values),  is low and
the velocity can be assumed constant in all the section.

To conclude, the complete turbulent movement is a good

approximation of the PF movement.
 Study of the Diffusion

More the movement is turbulent, more the turbulent diffusivity is high

and the axial diffusion is not negligible.

CeA CuA
PFR
CA
CeA PFR concentration profile

CuA
/z
=0 =p
z=0 z=L
Even if in the PFR there is a concentration gradient in the axial
direction, the diffusion in this direction is negligible.
In the laminar movement the diffusive processes are negligible
because the diffusion in just molecular and slow.
The turbulent movement is characterized by the turbulent
diffusion that increases with the increase of the turbulence. In this
case the the axial transport is not negligible.

The PFR model refers to a laminar movement with a slow

fluid velocity, referring to the diffusion, and it refers to a
turbulent movement with the highest velocity, referring to the
velocity.
UTILITY OF THE PFR MODEL
The PFR system is ideal, but some of his elements show
why it can be useful:

1. The model is an extreme of the fluid dynamic ideality;

2. The analytical structure of the model is reasonably simple;
3.The model can be used (low error rate) both for the reactors with
complete turbulent movement ( negligible axial transport) and fixed-bed
reactors. The approximation is good also for the reactors with two
countercurrent phases.
C Tubular duct in which the
reactions occur
z z+z
Cylindrical duct with
z
constant section S
Concentration profile

Flow direction Reference volume

V=Sz
• for the convective movement a linear velocity profile is assumed.
• No fluid mixing phenomena between one element and the other.
•All the elements characterized by a volume dV have the same velocity in the
movement direction and they have also the same residence time in the tubular
reactor.
The matter balance refers to an infinitesimal element with a volume dV.
Incoming amount = outgoing amount+ disappeared amount during the reaction +
accumulated amount
Considering the volume dV:
Incoming amount of A: FA
Outgoing amount of A: FA + dFA

A disappeared amount during the reaction = (-rA)dV

FA = (FA + dFA) + (-rA)dV
but:
FA = F0 (1 – X )
A A
dFA = dF0A(1 – XA) = -F0 dX
A A

replacing:
F0 dX = (-r ) dV
A A A

The balance for an infinitesimal portion of the reactor with a volume dV is

obtained; for the whole reactor you must integrate:

V XA rA changes for a PFR, is

FA0
 dV =  − rA
dX A constant for a CSTR
0 0


XA
V dX A
0
=
FA C A 0
=  − rA
0

XA
V dX A
 = = C A0
v0 
0
− rA
 Is the time to fill the reactor with the volume
V= V. It is called FILLING TIME and its unit of
τ
v measure in the SI is s.

The inverse of the filling time is the SPACE VELOCITY [s-1]

 General Moles balance for packed-bed reactors.

Input = Output + Gen + Accumul.

FA w FA w+w
XA
dX w

w = F A0 w+dw
0
(−rA' )

Catalyst mass = dw

(rA' )= in mol/t·catalyst mass

 Systems consisting of many reactors
PFR in series a/o shunt
Considering N reactors PFR in series:

Xi
V dX
FA0
=  − rA
X i−1

XiN
Vi V1 + V2 + V3 + ...
N
=  i=1 0 =
V dX
FA0
F F 0
=  ... +  ... +  ... =  − rA
0

So, N PFR reactors in series with a total volume V, have the same conversion of one
reactor with the same volume.
About the shunt reactors, the system can be considered as a single reactor only if
the supply ensures that each fluid current has the same composition.
 Systems consisting of many reactors
CSTR in series with the same dimensions
In the PFR the reagents concentration decreases along the reactor axis.
In the CSTR the concentration goes immediately to the final value.
 CSTR//PFR
Signal in step answer

The answer of the reactors is analyzed referring to two

different schemes:
•Signal in step
•pulse signal
It is possible to obtain an expression about the outgoing
concentration trend over time.
 MATTER BALANCE IN
VARIABLE VOLUME SYSTEMS
In systems in which the volume is variable you have to use a
differential formulation of the residence time, because  changes
with the abscissa.

dV’’ The volume changes with the position z. in

general is dV’dV’’, and d’different
dV’ from d’’

z
Because the next data will refer to a flow
reactor, all the calculations will refer to the
flow rates v [m3/s] and F [moli/s] instead
z’ z’+dz of V [m3] and N [moli]
z’’ z’’+dz
The next picture shows a geometric variation, but the
volume can change with the conversion. For example, if
the gas phase reaction A → 2B occurs with the
complete transformation of the reagent A, the final
volume will be twice the starting volume..

“v” (volumetric flow rate)

increases linearly with the
increase of the conversion.

“X” increases exponentially

with the increase of the filling
time.
 Performance equationof a PFR
with variable volume

-ri=Ri
If the variation of the volumetric flow rate depends on the conversion
only, it is possible to divide the variables writing:
 Performance equationof a PFR
with variable volume
Integrating:

The equation is formally equal to the performance equation of a

PFR with constant volume.

The equality is just formal, because the filling time is calculated

referring to the incoming current and (at variable volume) the
incoming current flow rate is different from the internal one.
 CONCLUSIONS
1. being the PFR ideal structures, in the SS they are described by a relatively
simple model. So the maximum utilization of this model is to give a
reference and determine the scope of applicability of this structure.

2. The high turbulence reactors are not comparable to the PFR, but
they are describable with their model with a reasonable degree of
approximation;
3. The description of fixed-bed reactors using the PFR model often
gives good results;
4.The PFR model is accepted for the description of countcurrent
reagent systems.
 ENERGY BALANCE IN
SYSTEM WITH A CONSTANT VOLUME

If in the PFR there is a not-isothermal reaction, the model must be

part of a performance equation that results from the energy balance.
Starting from the null mixing, these are the constraints:
1. Plat temperature profile in all the section (as the velocity profile is
supposed plat) and no conductive heat transport in the radial
direction for lack of gradient;
2. Even if the temperature is different from a section to another,
there is not conductive heat transport in the axial direction;
3. heat exchange through the outer wall only.
No heat diffusion in the axial direction

No heat diffusion in the radial direction

Constant profile temperature in each section but different from a section to

another.
 Flow direction

Reference volume between z

e z+z
z z+z
T
Temperature axial profile
for an adiabatic PFR in
which an endothermic
reaction occurs.
z

E – U =A– G + Q General structure of the

energy balance

outgoing generated
incoming accumul. exchanged
 ENERGY BALANCE

Fiflusso molare della specie i;

Niquantità della specie i presente nel volume di riferimento;
T temperatura nel volume di riferimento;
Totemperatura di riferimento;
Tatemperatura dell’ambiente;
S area della sezione di passaggio supposta costante;
A area di scambio termico.
 Dividing by Sz , the energy balance becomes

Simplifying:
SF
  i c  dT = R (-) ΔH
N
 (T
)a-t T
N
Uπd
 pi  R j
rj -
 i=1 S  dz j=1 S

NR is the number of the reactions in the system and Nc is the number of the
species.
For an adiabatic system (U=0) and
supposing that only one reaction occurs
(j=1), is:

The performance equation of a PFR in which the

reaction A→B occurs, so A=-1 , is:

so
Considering T vs. X the curve
that describes the
temperature variation with
the conversion is a straight
line (se Finon dipende da X)

The increase of the temperature is maximum when the system

is adiabatic with a complete conversion.

Integrating you can get the

maximum increase of temperature
(with total conversion and in an
adiabatic system)
 MOMENTUM BALANCE
It provides an equation that gives the pressure drop. You can
replace it with empiric equations that give the pressure drop
with a good approximation.

dP ρ 2
=-2 u f Pressure drop in a tube
dz d
16
Per Re (numero di Reynolds)<2000
Re Tubi
f=
-0,25 vuoti
0,059  Re Per Re>5000

Friction factor (o di 1 - ε 150 + 1,75 

  Ergun equation (filled tubes with
Fanning) 2ε  d p ρu µ
3
 particles with the dp diameter)
R = recycle ratio= (recycled fluid volume) / (outgoing volume)

Reactor with recycle

When R increases, the behaviour changes from piston flow (R=0) to
mixing flow (R=∞).

X Af
V dXA
= (R +1) 
FA0
 X
− rA
R  f
R+1 A
 

The recycle is useful to give a mixing to a PFR.

 PFR WITH RECYCLE I
veCeA Xe vuCuA Xu
1 PFR 2 v
v CoA CuA
r vR CuA XuA
Xu
rECY
CLE
SCHEME
R recycle ratio; vRvolumetric recycle flow rate; v incoming flow rate;
CeA/XeAconcentration/conversion in the incoming flow after the recycle;
CuA/XuA concentration/conversion in the outgoing flow and in the
recycled flow.

Recycle ratio
definition
 The PFR
reactor
with
recycle
is aPoiché
PFR
ve vu that si tratta di un PFR che opera fra C e
eA

disposes
C se supponiamo che esso operi su un
CuA of the
uA,
sistema a volume costante ,potremo
CeA
XuA flow
PFR applicare il modello tradizionale e scrivere

rate ve
XeA
Essendo ’il tempo di
and works riempimento basato
sulla portata ve
between the
starting and
Suppose to have a first orderfinal
kinetic:

integrate: concentratio
-rA=kCA
ns CeAand
CuA

e
r
 0 10 20 30 40
50
r

The outgoing flow concentration increases (at the same k) with the increse of the
recycle ratio. If r --> ∞ (r> 40) the reactor is like a CSTR. With low r, the reactor is
completely a PFR..
THE OPTIMAL CHOICE OF “r”
The optimal value of the recycle ratio is the value that makes minimum the
volume of the reactor (or its filling time ) i.e :
 LAYOUT OF IDEAL
REACTORS
Outlet flow from nth reactor = inlet flow of (n+1) reactor IN SERIES REACTORS
Indipendent feeds PARALLEL REACTORS

In series layout

Parallel layout
It is better to define the conversion with respect to only
one stream (also artificial), considering conversion zero in
correspondence of the network, rather than with respect
to each reactor feed.

Thanks to this choice it is possible to simplify many stages:

• calculations are easier;
• the number which indicates the conversion in each network point is the
conversion fraction at that point.

This simplification can be done when there are not lateral feed streams
or sample streams.
 F1A F2A
C1A C2A FeA , CeA , XA (artificial)feed
X1A X2A to reactor N° 1;
FeA
F1A , C1A , X1A (real) feed to
CeA
XA=0 reactor N° 2;
1 2 3
F3A F2A , C2A , X2A (real) feed to
C3A reactor N° 3;
X3A
artificial F3A , C3A , X3A features of the
product stream

If we suppose the existence of an artificial reactor (1), downstream to the real reactors (2
and 3) and the conversions X1A , X2A , X3A will be referred to the artificial feed, the number
corresponding to the real conversions indicate the fractional transformation of the generic
reagent (in that case A).
 CSTR in
series
CeA C1A
XeA X1A

CSTR 1 CSTR 2 Layout with two CSTR

C2A in series
X2A

Reaction: A → B
1st order kinetic, irreversible: –rA =kCA =kC°A (1-XA )
 One
CSTR
E - U = -G
Balance equation of a
CSTR vC eA vC eA (1 X 1 A ) V1k C eA (1 −X
= A 1)

By solving the equation with regards to the conversion XA, remembering

that =V1 /v is the filling time of the only one reactor, occurs the following
relation:
 Two
CSTR
E - U = -G
Balance on the 2nd
CSTR
v−C−−(1 X
eA 1A
) vC
eA
(1 X
2A
) Vk C
2 eA
(1 −=
X )
A2

Determining X2A is:

 k CSTR

If all the k CSTR in series

are identical: τk=τ, and
occurs:
Graphical meaning
Two CSTR in series have a volume inferior than the one of
a single CSTR, but they also have a volume much more
larger than a PFR.

Increasing the number of CSTR in series, the volume tend to

equal the one of a PFR.
CSTR with different volumes
 System with more than one PFR
PFR in series and/or in parallel
Let’s consider a number N of PFR in series:

Xi
V dX
FA0
=  − rA
X i−1

XiN
V1 + V 2 + V3 + ...
V
FA0
=  N Vi
i =1
F 0
=
F 0
=  ... +  ... +  ... = 
dX
− rA
0

We can conclude that N PFR in series, having a total volume V, permit the same
conversion of a single PFR having the same volume V.
When PFR are in parallel, we can consider the system as a single reactor only if the
feed is distributed in a way that the composition of the fluid streams encountered is
always the same.
 Comparing CSTR-PFR

For a generic reaction,

characterized by a formation
rate R, occurs:

For the reaction having –r=kC it is possible

to determine the curves reported besides;
observing the picture it is possible to state
that, when conversion is equal, an higher
reaction order corresponds to an higher ratio
CSTR /PFR
 Conclusions

VCSTR →VPFR
1. if  (reaction order) is >0, then is always VCSTR >VPFR
2. if X → 0
2. When >0, the ratio VCSTR /VPFR increases
 When <0, the ratio VCSTR /VPFR diminishes
6. if  >0, is better to work with the higher C (PFR is more suitable)
7. per  <0, is better to work with the lower C (CSTR is more suitable)
 Comparing CSTR –
PFR: Multiple
Let’s consider the two reactionreactions
which evolves at rates r and r
1 2

𝜎1𝐴𝐴 → 𝜎1𝐵𝐵 𝑟1 = 𝑘1𝐶𝐴 𝜎2𝐴𝐴 → 𝜎2𝐶𝐶 𝑟2 = 𝑘2𝐶𝐴

If the two reactions occur in a CSTR

and in a PFR, the following
expressions are valid:

In this graph is reported the possible

conversion achievable with a PFR and
with a CSTR. Even if multiple
reactions evolve, at equal conversions,
it will be always VCSTR <VPFR
Optimization of the total
volume
1st CSTR, 2nd CSTR, nth CSTR,
volume VT total volume V2 Total volume Vn

Each of them permit the same

conversion X
 C1 and V1 represent cost and
C1  V1 
0,6
The cost varies with
the empirical law: =   volume in the situation 1;
C2  V2  C2 and V2 represent cost and
volume in the situation 2

If, shifting from situation 1 to

V/C situation 2, the volume become the
half (V2 =0,5V1 ), the cost C2 will
V1 /C1 be:
C1 C1
C2 C2 = =  0,66  C1
 V1 
0,6 0,6
2
V2  
 V2 

n
n=1 n=2
For the reaction A -> B with k= 2,5 h-1, v= 5 m3/h, X=
0,99, it’s possible to obtain

By considering as
reference the situation
in which n=1

n VT CT
1 198 25
2 36 12
The two graphs report the cost curve and the curve
of the total volumes. The minimum cost (so the 4 17 10
optimal one) can be obtained for 4 different CSTR. 10 12 12

250 160
VOLUMI TOTALI

140

COSTI
200
120
150 100
80
100 60
40
50
20
0 0
0 10 20 30 40 0 10 20 30 40

n n
 Comparing CSTR-PFR:
Auto-catalytic reactions
Auto-catalytic reaction:
A+ B → B + C
The presence of one of
the product is necessary
for permitting the
chemical reaction.

The correct design expects:

1. A CSTR dimension corresponding to a final conversion in the outlet
flow which corresponds to the minimum of the curve 1/(-R) vs. X;
2. To put in series a PFR, which is able to convert the outlet flow from
CSTR in order to reach the conversion value desired.
 Auto-catalytic reaction A+B → B+C

SITUATION I: low conversion required (the curve: 1/-RA vs. X, has a

minimum)
1
+
−RA PFR

CSTR

Considering the graph is evident that the

0 Xf 1
XA volume required for using a PFR (it
corresponds to the sum of the green area
and the red area) is higher than the volume
required for using a CSTR (corresponding
A CSTR is more suitable to the red area).
 SITUATION II: very high conversion required (the curve
1/-RA vs.
X has a minimum)
1
Area proportional to the
−RA volume required for a PFR in
order to obtain a conversion
Xf

Area proportional to the

+ volume required for a
CSTR in order to obtain a
0 1Xf conversion Xf
XA

is evident that the volume required for using a CSTR (it

corresponds to the sum of the purple area and the red area) is
higher than the volume required for using a PFR (corresponding to
the red area).
SITUAZIONE III high conversion, but not too much (the
curve 1/-RA vs. X has a minimum)

In the graph it is highlighted the limit for which the sum of the red area and the blu
area (proportional to the volume of a PFR, required for obtaining conversion Xf ) is
equal to the sum of the green area and the blu area (proportional to the volume of a
CSTR, required for obtaining conversion Xf )

1
The equivalence VCSTR =VPFR
−RA can be obtained when the green
area is equal to the red area.
The X value corresponding to
that equivalence identifies the
border between the
advantageous use of a PFR
instead of a CSTR, and vice-
0 Xf 1 versa.
XA
 Multiple reactors
The choice to use a multiple reactor system can permit to operate in optimal
conditions when chemical reactions characterized by very complex kinetics
occur.
Often optimal conditions correspond to the maximum utilization of the
reaction volume available.
The reaction volume can be used at the best if in each reactor the
concentrations permit the maximum reaction rate.
The main problem of multiple reactors is the check, once the reactors and
the kinetic are provided, it is difficult to determine the correct position of the
reactors themselves in order to reach the maximum conversion.
 In the case of an auto-catalytic reaction, if we have to
reactors, one CSTR (volume: VCSTR ) and one PFR (volume: VPFR
), the most important thing is to decide correctly the layout,
in order to get the best one.

La disposizione che consente l’evoluzione della reazione alla velocità massima

è la prima.
In fact by choosing the CSTR volume in such a way that:
− CSTR /Ce permits to reach the minimum in the curve 1/-R vs. X
- Use the PFR in order to reach higher conversion than the one
corresponding to the minimum

The system is managed in the optimal way, so we can obtain the

best performance with the minimum total volume.
 Example: multiple reactors
We can use number: 3 PFR and number: 3 CSTR for a reaction, whose kinetic was
experimentally determined (kinetic data are reported in figure);

60

50
the curve 1/-R vs X is characterized
by a maximum point (when X=0,30)
40
and a minimum point (when
!/-

30 X=0,65).
R

20

10 In the increasing part of the curve

0
is better to use the PFRs, while in
0 0,2 0,4 0,6 0,8 1 the decreasing part of the curve the
X
CSTRs.
 60

50
Is better to put the PFRs in the area in which
the curve 1-R vs. X is increasing. Let's
40
suppose that two PFRs (for example N° 1
and N° 2) have /Ce values able to cover the
!/-

30
R

20 2 /Ce red areas in the picture, permitting to reach

conversions X1 and X2 in the outlet streams,
10
1 /Ce respectively of the PFR 1 and PFR 2.
0
0
X10,2 X2 0,4 0,6 0,8 1
X

60
The CSTRs should be put in the area in which
50
the curve is decreasing. In such a way it is
40 possible to cover an area equal to 3 /Ce and 4
/Ce , the N° 3 CSTR is just after positioned, the
!/-

30
R

area covered by this reactor corresponds to

3 /Ce
5 /Ce. Even if this layout is not perfect, it is
20

10
4 /Ce the best one of the possible layouts.
5/Ce
0 X5 The conversions characterizing the outlet
X3 X4
0 0,2 0,4 0,6 0,8 1
streams of these reactor are, respectively, X 3 ,
X
X4 and X5 .
 60

50

The last PFR aviable has been

40
positioned at the end; the corresponding
 6 /Ce and the
!/-R

30 covered area is
conversion of the outlet stream is X6 .
20
6 /Ce
The optimal structure is reported in
10 the following picture.

0 X6
0 0,2 0,4 0,6 0,8 1
X

CSTR N° 2 PFR N° 3
Xo =0 X1 X6
X2 X3 X4 X
5

PFR N° 1 PFR N° 2
CSTR N° 1 CSTR N° 3