`WELCOME TO PROCESS AND
EQUIPMENT DESIGN
Mass and heat transfer Equipment
(CHEM 405)
Learning Outcome 01:Perform material and energy
balance calculations and develop flow sheets.
Perform material balance calculations with special emphasis on
multiple-unit process with recycle and purging streams.
Material Balance:
Material out (output)
= Material in (input) + generation –
consumption – accumulation
or Input + Generation = Output + Consumption +
Accumulation
For a steady-state process:
Accumulation = 0.0
Also Generation = 0.0 (Except in nuclear processes)
Therefore the balance equation reduces to: input = output
Example 1.1
See Example 2.1 page 35 in your textbook Vol 6.
Exercise 1.1
Repeat the calculations using the same data accept that the
diluted slurry to be 12% instead of 5%.
Exercise 1.2
1000 kg/hr of a mixture of benzene and toluene contains 50%
benzene by mass is to be separated by distillation into two
fractions. The flow rate of benzene in the top stream (distillate) is
450 kg/hr, and that of toluene in the bottom stream (waste) is 475
kg/hr. Calculate:
1- The flow rate of the missing components in output streams in kg/s.
2- Mole fraction of the benzene in the bottom stream.
A multiple-unit process
Solve the following problem and define the boundary for each
material balance loop.
Exercise 1.3
Calculate the unknown flow rates in the continuous steady-state
distillation process shown below.
Conversion and Yield
Conversion is a measure of the fraction of the reagent that
reacted.
Conversion =
Amount of reagent consumed
Amount supplied in feed
Amount consumed = Amount in feed – Amount in product
See Example 2.10 in textbook Vol 6 page 47.
Exercise 1.4
Repeat the calculations for the Example 2.10 in the textbook Vol 6
using a conversion of 50% to produce 9500 lbm/h .
Stoichiometry equation for chemical reaction states the
number of molecules of the reactants and products that takes
part.
Example 1.2
See Example 2.3 page 36 in textbook Vol 6
Yield is a measure of the performance of a reactor or a plant, and
it concerns one of the product components. It is defined in
different ways, one definition is:
Reactor Yield =
(Moles of product produced) (Stoichiom etric factor)
(Moles of reagent converted)
Stoichiomm etric factor =
Stoichiomm etric number of moles of reactant
Stoichiomm etric number of moles of product
The plant yield is a measure of the overall plant performance and
includes all chemical and physical losses.
Plant Yield =
(Moles of product produced) (Stoichiom etric factor)
(Moles of reagent fed to the process)
See Example 2.11 page 48 in textbook Vol 6
Exercise 1.5
Calculate the yield of ether based on water in Example 2.11 of the
textbook Vol 6.
Exercise 1.6
Repeat the calculations of Example 2.12 in the textbook Vol6
using a conversion of ethylene equal to 98.5%.
Recycle and Purge Calculations
Recycle streams are unknown in quantity and therefore some
trial and error iterations may be required to perform the material
balance. With simple problems the calculations can be simplified
by careful selection of the basis and the boundaries.
Example 1.4
See Example 2.13 page 51 of the textbook Vol 6
Purge is the bleed off a portion of a recycle stream to prevent the
build-up of unwanted material.
Under steady-state conditions: Loss of inert in the purge = Rate of
feed of inert into the system
Purge rate: [Feed stream flow-rate] x [Feed stream inert
concentration] = [Purge stream flow – rate] x
[Specified (desired) recycle inert concentration]
Example 1.5
See Example 2.14 page 53 of the textbook. Vol 6
SO1: Evaluate Vapor-Liquid Equilibrium Data
When multi-component gas and liquid phases are in equilibrium, a limited
number of intensive system variables may be specified and the remaining
variables can be determined using equilibrium relationships for the
distribution of components between the two phases.
The best way to evaluate equilibrium compositions is from tabulated data,
Perry’s Chemical Engineers’ Handbook.
Example 1.6
See Example 6.41 page 255 Principles textbook
SO2: Apply Raoult’s law and Henry’s law to multi-component system
at equilibrium.
Raoult’s Law:
pA = yAP = xA p*A(T)
valid when xA is close to
1. Also for similar structure compounds like benzene & toluene.
where
p*A = the vapor pressure of pure liquid A at temperature T
yA = the mole fraction of A in the gas phase
pA = the partial pressure of A in the gas phase
P
= Equilibrium total pressure
T
= Equilibrium temperature
Henry’s Law
pA= yAP = xAHA(T)
valid when xA is close to 0
Where HA(T) is the Henry’s law constant for A in a specific solvent.
A gas-liquid system in which the vapor-liquid equilibrium relationship for every
volatile species is either Raoult’s law or Henry’s law – Exhibit ideal Solution
Ideal liquid solution is a mixture of liquids that exhibits ideal solution behavior
at equilibrium
Example 1.7
See Example 6.4-2 page 258 (Principles textbook)
SO3 Develop vapor-liquid equilibrium data for ideal
solutions.
When a liquid is heated slowly at constant pressure, the
temperature at which the first vapor bubble forms is the bubblepoint temperature of the liquid.
When a gas (vapor) is cooled slowly at constant pressure, the
temperature at which the first liquid droplet forms is the dew-point
temperature.
Suppose an ideal liquid solution follows Raoult’s law and contains
species A, B, C, …with known mole fractions xA, xB, xC,…If the
mixture is heated at a constant pressure P to its bubble-point
temperature Tbp, the further addition of a slight amount of heat will
lead to the formation of a vapor phase.
Since the vapor is in equilibrium with the liquid, and we now
assume for idea vapor, the partial pressures of the components
pi = xipi*(Tbp),
i = A, B, …..
Where pi* is the vapor pressure of component i at bubble-point
temperature. The total system pressure, P
P = xApA*(Tbp) + xBpB*(Tbp)+…
Tbp may be calculated by trial and error as the value of T bp that
satisfies this equation; all that is needed is a set of relationships
for pi*(T) as the Antoine equation.
Once Tbp is known, the composition of the vapor phase can
easily be determined by evaluating the partial pressures of
each component from Equation pi = xipi*(Tbp) and determine
each vapor – phase mole fraction as yi = pi/P.
The pressure at which the first vapor forms when a liquid is
decompressed at a constant temperature is the bubble-point
pressure of the liquid at the given temperature.
Equation P = xApA*(Tbp) + xBpB*(Tbp)+…can be used to determine
(1) pressure for an ideal liquid solution at a specific temperature,
(2) the mole fractions in the vapor in equilibrium with the liquid can
then be determined as
pi
xi pi* (T )
yi =
=
Pbp
Pbp
The dew point temperature of a gas (vapor):
Suppose a gas phase contains the condensable components A,
B, C, …. and a non-condensable component G at a fixed pressure
P.
Let yi be the mole fraction of component i in the gas.
If the gas mixture is cooled slowly to its dew point, Tdp, it will be in
equilibrium with the first liquid that forms.
y
xi =
Assuming that Raoult’s law applies,
i
*
i
p (Tdp )
i = A, B, C, …
At the dew point of the gas mixture, xA + xB + xC + …. =1
Liquid phase composition may be determined from Equation
yi
xi = *
p i (Tdp )
Using
Equation
Pdp =
1
yC
yA
yB
+
+
+ ..
p *A (T ) p B* (T ) pC* (T )
The value of Tdp can be found by trial and error once
expressions for pi*(T) have been substituted.
The dew-point pressure,
Liquid mole fractions may then be calculated
yi
x
=
from Equation
i
p * (T )
i
with Tdp replaced by the system
temperature, T.
Example 1.8
See example 6.4-3 page 260 in Principles textbook
dp
SO4 Illustrate the use of the graphical
representation of vapor-liquid equilibrium
method in binary systems calculation.
Txy & Pxy diagrams –Boiling point diagrams
If the liquid mixture behaves as an ideal solution then Raoult’s
law can be used to estimate the bubble and dew points.
The bubble and dew point calculations can be used to determine
the equilibrium vapor or liquid composition.
The vapor – liquid equilibrium calculations for a binary system
(two components system) can be simplified by using the Txy
diagram or using Pxy diagram.
These are known as boiling point diagrams.
In Figure the bubble point is determined from the saturated
liquid curve (bubble-point curve), and the dew point from the
saturated vapor curve (the dew-point curve). The two-phase
region is in the region between these two curves.
Figure Txy and Pxy diagram for benzene – toluene.
Example 1.9 / Determine the bubble point and the equilibrium
composition of a 40 mole% benzene - 60 mole % toluene liquid
mixture at 1 atm. If the mixture is steadily vaporized until the
remaining liquid contains 25% benzene, what is the final
temperature?
Solution/ With reference to Figure Txy Diagram, benzene is the
more volatile component.
for
When
xB = 0.40, Tbp= 95oC
xB = 0.25, Tbp= 101oC
See Principles textbook example 6.4-4 page 262
A similar phenomenon for single-component systems occurs for
liquid mixtures.
If a mixture is heated slowly in an open container, vapor bubbles
will form at the heated surface and emerge into the gas phase
when the vapor pressure of the liquid equals the pressure above
the liquid.
For an ideal liquid solution, the boiling point may therefore be
determined from equation:
XApA*(Tbp) + xBpB*(Tbp) + …. =P
See Principles textbook example 6.4-5 page 263
SO5: Equilibrium Between Two Liquid Phases
Miscibility and Distribution Coefficients
Suppose A and S are two nearly immiscible liquids and B is a
solute distributed between the phases of an A-S mixture.
The distribution coefficient (or partition ratio) of component B =
(mass fraction of B in the S phase)/(the mass fraction of B in the A)
Example 1.10 See Example 6.6-1 Page 272 Principles textbook
SO6: Phase Diagrams for Ternary Systems
Triangular phase diagram: Take the form of an equilateral triangle
or a right triangle.
Apex- represents a single component
Edges represent binary solutions.
Edge b on the Figure represents
solutions of water & acetone.
Region A is a single phase liquid
while region B separates into two
phases
The lines within region B-called tie
lines- connect compositions of the
two liquid phases in equilibrium.
Figure Triangular phase
diagram for water-acetonemethyl isobutyl ketone.
See Example 6.6-2 page 274
Principles textbook
SO7 Explain how to develop and present a process flow
sheet.
The data on the flow-rates of each component (composition),
temperature and pressure can be shown on the flow sheet. The
common method is to number each stream and then tabulate the data at
the bottom of the sheet. See P 133 – 135 in the textbook.
Essential information
1- The flow rate of each component (kg/h is preferred)
2- Total stream flow rate, kg/h.
3- Stream temperature, 0C.
4- Operating pressure.
Optional information
1- Molar composition.
2- Physical properties.
3- Stream enthalpy.
4- Basis used for calculations including operating hours per year,
reaction and overall yield, datum temperature used to perform the heat
balance, and assumptions used in calculations.
SO8 Identify a typical flow-sheet for a common chemical
processing plant.
The presentation of data on flow-sheets is shown in Figure 4.2 Page 136 in the
textbook. In this method each stream line is numbered and the data tabulated
at the bottom of the sheet.
Reference:
Chemical Engineering Design Vol 6
Coulson & Richardson’s Chemical Engineering Series. Fourth Edition
LO2 Explain the techniques for the prediction of physical
properties.
SO1
SO2
SO3
SO4
SO5
SO6
SO7
SO8
Explain the methods for calculation of liquids, gas and vapor densities.
Describe the methods for the calculation of viscosity of liquids and
gases.
Review the estimation methods of the thermal conductivity of Solids,
Liquids, gases, liquid mixtures, and gases mixtures.
Review the methods for estimate the specific capacity Solids, liquids.
and gases.
Estimate the enthalpy of vaporization for individual component and
mixtures.
Estimate the diffusion coefficients for gases and liquids.
Estimate the surface tension of liquids and hydrocarbon mixtures.
Estimate the K-values for hydrocarbons.
SO1 Explain the methods for calculation of liquids, gas and vapor
densities.
Liquids:
An approximate estimate of the liquid density at the normal boiling
point can be obtained from the molar volume (see Table 8.6 page 334)
b =
Where b = density, kg/m3,
M = molecular mass,
Vm= molar volume, m3/kmol.
M
Vm
For mixtures,
𝜌𝑚 = ෍ 𝑥𝑖 𝜌𝑖
The densities of many aqueous solutions are given by Perry et al.
handbook.
Gas and vapor density (specific volume)
Using the gas law to find the density of ideally behave real gases and
vapors.
PV = nRT
Where P = absolute pressure N/m2 (Pa),
V = volume m3,
n = mols of gas
T = absolute temperature, K.
R = universal gas constant.
Specific volume = (RT/P) =V/n
For more accurate include the compressibility factor z: PV = znRT
For mixture, the pseudocritical properties of the mixture should be used
to obtain the compressibility factor.
Pc,m = Pc,aya + Pc,byb + …..
Tc,m = Tc,aya +Tc,byb + ….
Where PC = critical pressure
TC = critical temperature
y = mol fraction. zm = compressibility factor for for gas mixture
Suffixes: m = mixture
a, b, etc. = components.
see principles page 211
SO2 Describe the methods for the calculation of viscosity of liquids
and gases.
Methods for the estimation of viscosity:
For liquids: rough estimate of µb = 0.01 ρb 0.5
where µb in mNs/m2 and ρb in kg/m3
Viscosity estimation: log (log 10) = (I/M) x 10-3 – 2.9
Where  = viscosity, mNs/m2
M = molecular mass,
I = Souders’ index, given in Table 8.1 page 317 in the textbook.
 = density at the required temperature, kg/m3.
Example 3.3 See example 8.2 page 316 in textbook
Variation of viscosity with temperature: Use Figure 8.1 page 318 in the
textbook.
Example 3.4 See Example 8.3 page 318 in the textbook.
For viscosities and densities of liquid see Table 8 page 800 Volume 1
Coulson & Richardson’s Chemical Engineering, sixth edition.
Viscosity of Liquid Mixtures: For binary organic liquid mixtures
 x I +x I 
log(log 10 m ) =  m  1 1 2 2  x10 −3 − 2.9
 x1 M 1 + x2 M 2 
Where m = viscosity of mixture
m = density of mixture
x1, x2 = mol fraction of components
M1, M2 = molecular masses of components.
I = calculated use Table 8.1 page 317
1
Estimate for organic mixtures:
m
=
w1
1
+
w2
2
Where w1, w2 = mass fractions of the components 1 and 2
1, 2 = viscosities of components 1 and 2.
For gases:
See Table 7 page 798 Volume 1 Coulson & Richardson’s Chemical
Engineering, sixth edition.
SO3 Outline the estimation methods of the thermal conductivity of
Solids, Liquids, gases, liquid mixtures, and gases mixtures.
The thermal conductivity of a material
is defined as the rate of heat transfer
through a unit thickness of the
material per unit area per unit
temperature difference. The thermal
conductivity of a solid is determined by
its form and structure, as well as
composition. See Table 9.1 page 389
Volume 1 Coulson & Richardson’s
Chemical Engineering, sixth edition for
the thermal conductivities of selected
materials.
The thermal conductivities of some
materials at room temperature:
Material
k, W/m.0C
Diamond
2300
Silver
429
Copper
401
Gold
317
Aluminum
237
Iron
80.2
Mercury (l)
8.54
Glass
0.78
Brick
0.72
Water (l)
0.613
Human skin
0.37
Wood (oak)
0.17
Helium (g)
0.152
Soft rubber
0.13
Glass fiber
0.043
Air (g)
0.026
Urethane, rigid foam 0.026
A rough estimate of the thermal conductivity of organic liquids.
 4 
−5

k = 3.56  10 C P 
M 
1/ 3
Where k = thermal conductivity. W/m 0C
M = molecular mass
CP= specific heat capacity, kJ/kg 0C
ρ = density, kg/m3.
Example 3.5: See example 8.4 page 321 in the textbook.
Approximate values for the thermal conductivity of pure gases, up to
moderate pressures, can be estimated from values of the gas viscosity.
Where  = viscosity, mNs/m2
CP = specific heat capacity, kJ/kg 0C
M = molecular mass
Example 3.6
See example 8.5 page 321 in textbook.
For all non-polar a simple weighted average is usually sufficiently
accurate for design purposes.
km = k1w1 + k2w1 + …
where km
= thermal conductivity of mixture
k1, k2 = thermal conductivity of components
w1, w2= component mass fractions.
[Polar – a result of an unsymmetrical distribution of electrons, the
bond or molecule contains a positive and a negative pole and is
therefore a dipole. Covalent bonds between unlike atoms are always
polar; e.g. HF, and H–C]
[Nonpolar – a symmetrical distribution of electrons leads to a bond or
molecule with no positive or negative poles. e.g. H2 , and F2,]
SO4 Outline the methods for estimate the specific heat capacity Solids,
liquids and gases.
The heat capacity of a compound is taken as the sum of the heat capacities of
the individual elements of which it is composed. See table 8.2 in page 322 of
the text gives the values of each element.
Example 3.7 See Example 8.6 page 323 in the textbook.
For organic liquids, the group contribution method gives accurate predictions.
The contributions to be assigned to each molecular group are given in Table
8.3 and method illustrated in Examples 3.8 & 3.9 see Examples 8.7 & 8.8 in
pages 323 – 325 in the textbook.
The specific heats of liquid mixtures can be estimated by taking heat capacities
of the components as additive. For dilute aqueous solutions it is usually
sufficient to take the specific heat of the solution as that of water.
For gas in the ideal state the specific heat capacity at constant pressure is given
by:
C0P = a + bT + cT2 + dT3
To estimate the constants for organic compounds use the values of the
molecular group in Table 8.4 page 326 in the textbook with the
exception of acetylenic compounds (
).
Example 3.10
For illustration see example 8.9 in page 328 in the textbook.
SO5 Estimate the enthalpy of vaporization for individual component
and mixtures.
The latent heats of vaporization can be found in Appendix C page 937 in
the textbook. A very rough estimate can be predicted: Lv
= cons tan t
Tb
Where LV = latent heat of vaporization, kJ/kmol
Tb = normal boiling point, K
For organic liquids the constant can be taken as 100. More accurate
estimate can be made from a knowledge of the vapor pressuretemperature relationship for the substance. Using Antoine vapor
pressure equation, the latent heat can be predicted from this equation:
8.32 BT 2 z
LV =
(T + C ) 2
Where LV = latent heat at the required temperature, kJ/kmol
T = temperature, K. B, C = coefficients in the Antoine equation.
∆z = zgas – zliquid ( where z is the compressibility constant),
calculated from the equation:
If an experimental value of the
latent heat at boiling point is
known. It is used to estimate
the LV at other temperature

P 
z = 1 − r3 
 Tr 
0.5
Where LV = latent heat at temperature T, kJ/kmol
LV,b= latent heat at the normal boiling point, kJ/kmol
Tb = boiling point, K
TC = critical temperature, K
Tr = T/Tc Pr = P/Pc
T = temperature, K.
Another formula that provides roughly 2% accuracy is Chen’s equation:
LV (kJ / mol) =
Tb 0.0331(Tb / Tc ) − 0.0327 + 0.0297 log 10 Pc 
1.07 − (Tb / Tc )
Where Tb and Tc are the normal boiling point and critical temperature in Kelvin
and Pc is the critical pressure in atmospheres.
If phase change occurs at temperatures other than the temperature for which
the latent heat is tabulated, select a hypothetical process path that permits the
available data to be used.
For example, a substance is to be vaporized isothermally at 1300C, but the only
available value of heat of vaporization is at 800C.
A process path from the liquid at 1300C to vapor at the same temperature:
Cool the liquid from 1300C to 800C, vaporize the liquid at 800C, and then heat
the vapor back to 1300C. Summing the changes in enthalpy for each of these
steps yields the change in enthalpy for the given process. The calculated value
is the latent heat of vaporization at 1300C.
Detailed method can be found in section 8.4a page 379 of the textbook
“Elementary Principles of Chemical Processes” by Felder & Rousseau, third
edition.
Example 3.11 Estimation of a heat of Vaporization
The normal boiling point of methanol is 337.9 K, and the critical
temperature of this substance is 513.2 K. Estimate the heat of
vaporization of methanol at 2000C. Given that the heat of vaporization
of methanol at 337.90C =36.8 kJ/mol.
0.38
Solution:
T −T 
2

LV (T2 ) = LV (T1 ) c
 Tc − T1 
 513.2 − 473 
LV (473K ) = 36.8 

513
.
2
−
337
.
9


0.38
= 21.0 kJ / mol
The latent heat of mixtures: LV, mixture = Lv1x1 + Lv2x2 +……
Where
LV1, Lv2 = latent heats of the components kJ/kmol.
x1 , x2 = mol fractions of components.
Example 3.12
See Example 8.10 page 329 in textbook.
SO6 Estimate the diffusion coefficients for gases and liquids.
Diffusivities of gases and vapors in air at 298 K and atmospheric
pressure are given in Table 10.2 page 581 of Coulson & Richardson’s
Chemical Engineering volume 1 sixth edition.
Diffusivities of gases can be predicted by the following equation:
Dv =
 1   1 
 + 

1.013  10 −7 T 1.75 
 M a   M b 
1/ 3
1/ 3



 
P   vi  +   vi  
 b  
 a 
0. 5
2
Dv
= diffusivity in cm2/s
T
= temperature in K
Ma, Mb = molecular mass of components a & b
a vi , b vi = the summation of the special diffusion volume coefficients for
components a and b, given in the Table 8.5 page 332 in the
textbook.
P
= total pressure, bar.
Example 3.14
See Example 8.11 page 332 in the textbook
Diffusion of solutes in liquids is very important in many industrial processes,
especially in such separation operations as liquid-liquid extraction or solvent
extraction, gas absorption, and distillation.
The diffusivities in liquids are generally four to five orders of magnitude
smaller than in gases at atmospheric pressure.
Diffusivities of in liquids at 293 K are given in Table 10.7 page 598 of Coulson &
Richardson’s Chemical Engineering volume 1 sixth edition.
Diffusivities for dilute liquid solutions of non-electrolytes can be calculated
approximately from the equation
0.5
(

M
)
T
−13
DL = 1.173x10
Vm0.6
where
DL = Liquid diffusivity, m2/s
T = absolute temperature, K
 = viscosity of solvent, mN s/m2
Vm =molar volume of solute as liquid at its normal boiling point, m 3/kmol.
This can be estimated from the group contributions given in Table 8.6 page 334 in textbook
 = association factor for the solvent
Recommended values of 
M = molecular mass of solvent
Example 3.15
See Example 8.12 page 333
in the textbook.
SO7 Estimate the surface tension of liquids and hydrocarbon mixtures.
If reliable values of liquid and vapor density are available, the surface
tension can be estimated by a group contribution method.
 P ( −  ) 
 =  ch L v   10 −12
M


4
Where  = surface tension, mJ/m2 (dyne/cm)
Pch = Sugden’s parachor
L = liquid density of the saturated vapor, kg/m3
v = density of the saturated vapor, kg/m3
M = molecular mass.
, L, v evaluated at the system temperature. The Sugden’s parachor
for organic compounds are given in Table 8.7 page 335 of the textbook.
The surface tension of a mixture is rarely a simple function of
composition. A rough estimation for hydrocarbons:
m = 1x1 + 2x2+ ….
Where m
= surface tension of mixture
1, 2 = surface tension of components
x1, x2 = component mol fractions
Example 3.16 See Example 8.13 page 336 in textbook.
SO8 Estimate the K-values for hydrocarbons.
One way to represent equilibrium data is to define a distribution
coefficient or K value as
KA = yA / xA
In general, the K values depend on temperature, pressure, and
composition. But for many systems the K values are
approximately independent of composition. Thus,
K = K(T, p)
(approximate)
For light hydrocarbons, the approximate K values can be
determined from the monographs prepared by DePriester.
An approximate K values can be determined from using the following equation that fits
DePriester charts for K values:
ln K =
aT1
aT2
T
T
+
2
+ aT6 + a p1 ln p + +
a p2
a p3
p
p
+
2
where T is in 0R and p is in psia. The constants aT1, aT2, aT6, ap1, ap2, and ap3 are given in
Table 3.1. If K and p are known, the equation can be solved fort.
The K values are used along with the stoichiometric equations:
,
N
 y = 1.0
i =1
i
N
 x = 1.0
i =1
i
Example 3.17 / Find the boiling point of isobutene at p =150 kPa and
K=1.
Solution:
ln K =
aT1
T
2
+
aT2
T
+ aT6 + a p1 ln p + +
a p2
p
2
+
a p3
p
In addition to the equations arising from the material and energy balances
over a stage, and the equilibrium relationships, there will be a fourth
relationship, the summation equation for the liquid and vapor compositions:
Dew points and bubble points can be calculated from a knowledge of the
vapor-liquid equilibrium for the system.
Bubble point:
yi =  Kixi = 1.0
and
dew point:
xi =  =1.0
Equilibrium flash calculations
Flash calculations are often needed to determine the condition of the feed to a
distillation column and occasionally, to determine the flow of vapor from the
reboiler, or condenser if a partial condenser is used.
The following figure shows a typical equilibrium flash process. The equations
describing this process are:
Material balance, for any component, I Fzi = V yi + Lxi
Energy balance, total stream enthalpies: Fhf= V H + Lh
For
yi = Kixi
Hence
F zi = VKi xi + Lxi
From which
and similarly
V

= Lxi  K i + 1
L

Fzi
L=

i VK i
+
1
 L



Fzi
V =

i  L
+ 1

VKi

Where L/VKi is known as the
absorption factor Ai
VKi/L is called the
stripping factor Si
Example 3.18
A feed to a column has the composition given in the table below, and is at a
pressure of 14 bar and a temperature of 600C. Calculate the flow and
composition of the liquid and vapor phases. Take the equilibrium data from the
Depriester charts.
Solution
y i  i f i OL
Ki =
=
xi
Pi
The phase equilibria
Where Ki is the distribution coefficient (the K value)
 i vapor fugacity coefficient
P total systems pressure
yi concentration of component i in the vapor phase
xi concentration of component i in the liquid phase
 i liquid-phase activity coefficient
f i OL standard state fugacity of the pure liquid.
For systems in which the vapor phase imperfections are not significant, the above
equation reduces to the familiar Raoult’s law equation
 P0
Ki =
i i
P
Where the pure component vapor pressure (calculated from Antoine
equation), N/m2. The relative volatility of two components can be
expressed as the ratio of their K values:  ij = K i
Kj
Pi 0
For ideal mixtures (obeying Raoult’s law)
Ki =
P
K i0 Pi 0
o
and
 ij = 0 = 0
where K i and K 0j are the ideal K values
Kj
Pj
for components i and j.
Learning Outcome 3
Separation Processes
Chemical Engineering Design
Separation Processes
• Separation processes are needed for feed pretreatment,
product recovery and waste processing
• Most separations are based on moving a component from
one phase to another and then segregating the two phases
– Driven by activity gradient as phases try to reach equilibrium
– Affected by rates of mass transfer and heat transfer
• Ch 16 (this lecture): separations involving gases or liquids
• Ch 17: multistage vapor-liquid separations (distillation,
absorption, stripping, extraction)
Chemical Engineering Design
Separation Specifications
P, yA, yB
Product enriched in A
F, zA, zB
R, xA, xB
• Recovery: How much of the desired component made it
to the stream it was supposed to be in:
P yA
P yA
Recovery of A =
=
F z A (P y A + R x A )
Chemical Engineering Design
Separation Specifications
P, yA, yB
Product enriched in A
F, zA, zB
R, xA, xB
• Purity: The concentration of desired component in the
stream it was supposed to be in:
Purity of A in product = yA
Chemical Engineering Design
Impact of Separation Specifications
Cost
• Tighter specifications lead to higher cost:
90
99
99.9
99.99
Purity or Recovery (%)
• Final product must meet purity specifications
– Set by ASTM, USP, etc.
• Recycles sometimes have purity specifications
– e.g. to protect catalyst from contaminants or poisons
• Product that is not recovered is lost profit and also increased
waste cost: separation recovery factors into plant yield
Chemical Engineering Design
Vapor-Vapor Separations
Membrane
Based on differences in relative
permeability of gases
Used for H2/CH4, CO2 removal,
air separation
Absorption
Using a liquid solvent
in an absorberstripper loop
Used for acid gases,
drying, water wash
Adsorption
Adsorb components selectively on a solid
Regenerate sorbent by temperature swing
(TSA) or pressure swing (PSA)
Used for air separation, H2/CH4, most
separations involving low concentrations
Chemical Engineering Design
Adsorption
• One component from vapor phase preferentially adsorbs
onto the surface of a solid adsorbent
• Two types of adsorption:
– Reversible:
• Usually physisorption
• Adsorbed component can be released by decreasing pressure or increasing
temperature
• Sorbent can be regenerated and used in multiple cycles, hence temperatureswing adsorption (TSA) and pressure-swing adsorption (PSA)
– Irreversible:
• Usually chemisorption
• Adsorbed component usually reacts irreversibly with solid
• Low concentrations can be achieved, but solid is difficult to regenerate
• Used for contaminant removal guard beds
Chemical Engineering Design
Concentration Profiles During
Adsorption
Gas mixture A + B
Concentration of B on sorbent
• Concentration
profile moves
down the bed
during adsorption
t1
t2
tB
Distance down sorbent bed
Purified gas A
• At time tB
breakthrough of
the adsorbed
component
occurs and it
begins to appear
in the outlet gas
Chemical Engineering Design
Irreversible Adsorption
Feed
= open valve
= closed valve
Product
• Two guard beds can be used in parallel so that when Bed 1
nears breakthrough the process flow can be switched to
Bed 2
• Some adsorbent will be wasted, as beds cannot be run
close to breakthrough for fear of contaminant slippage
Chemical Engineering Design
Irreversible Adsorption: Lead-Lag
Guard Bed System
Feed
= open valve
= closed valve
Product
• Alternative arrangement has beds in series
• When upstream bed reaches breakthrough, downstream
bed is still OK. Upstream bed can be taken offline, reloaded
and brought back into downstream service, etc.
• Makes more efficient use of adsorbent
Chemical Engineering Design
Guard Beds for Mercury Capture
• Mercury occurs in
natural gas and
light oils
• It must be
removed to
protect
equipment and
catalysts
Source: UOP
Chemical Engineering Design
Reversible Adsorption: Isotherms
T2
T2 > T1
Partial pressure
T1
p1
p2
m2
m1
Mass adsorbed (g/g sorbent)
• Reversible adsorption exploits changes in loading with
pressure or temperature
Chemical Engineering Design
Reversible Adsorption: Isotherms
T2
T2 > T1
Partial pressure
T1
Adsorb at (p1, T1) gives
loading m1
p1
Pressure Swing:
Decrease pressure to p2 and
loading decreases to m2
p2
m2
m1
Mass adsorbed (g/g sorbent)
Delta loading = m1 – m2
(kg/kg sorbent)
• PSA: cycle between high and low pressure to load and
regenerate the adsorbent
Chemical Engineering Design
Reversible Adsorption: Isotherms
T2
T2 > T1
Partial pressure
T1
Adsorb at (p1, T1) gives
loading m1
p1
Temperature Swing:
Increase Temperature to T2
and loading decreases to
m2
p2
m2
m1
Mass adsorbed (g/g sorbent)
Delta loading = m1 – m2
(kg/kg sorbent)
• TSA: cycle between low and high temperature to load
and regenerate the adsorbent
Chemical Engineering Design
PSA and TSA Systems
PSA
• Shorter cycle time (no
heating or cooling)
– Typically 5 – 60 mins
• Multiple beds needed for
high recovery, purity
– Use pressure balancing
and purge to get better
recovery and purity
– 8, 10, 12, 16 bed plants
TSA
• Longer cycle time for heating
and cooling of bed and vessel
– Typically 60 – 200 mins
• Additional equipment needed
for heating & cooling
– Often use a purge gas for regen,
e.g. steam, N2 or a slip-stream of
product
• Fewer beds (no need to
pressure equalize)
• Applications: hydrogen
purification, air separation • Applications: gas drying, VOC
capture, CO2 removal in cryo
plants
Chemical Engineering Design
12-Bed PSA Unit
Surge Tank
Adsorber Vessels
Valve Skid
Source: UOP
Chemical Engineering Design
PSA Cycle
Bed
A
1,2,3
PP
D
P
R
Time
absorb
equalize
provide purge
desorb
purge
repressure
From U.S. 4,381,189 “Pressure Swing Adsorption System and Process”
• Pressure equalization steps reduce the amount of gas lost during
depressurization and hence improve recovery
• Repressurization is done using product gas to improve purity
• Some steps are co-current, some counter-current, to exploit
concentration profiles in the bed
• Many different cycles have been invented – see patent literature for
examples
Chemical Engineering Design
Preliminary Design of PSA Units
1. Delta loading across cycle depends on the adsorbent
selected, the temperature of operation and the pressure
cycle – use isotherms to determine delta loading
2. Select number of beds (more beds = more equalization
steps, higher recovery, higher purity)
3. Select cycle time and time in adsorption step, ta
4.
Mass of adsorbent per bed Mass flow rate of adsorbed component
ta
=
delta loading  bed loading factor
Bed loading factor = fraction of bed loaded at end of adsorption stage ~ 0.8 to 0.9
5. Size each bed as a cylindrical pressure vessel
6. Add costs for valve skids, surge tank
(Detailed design – need to consider mass transfer rates and
dynamics – much more complex analysis)
Chemical Engineering Design
Adsorption Equipment Design
• (F1y1 − F2y2)Mw ta = 1000(m1 − m2) MafL
•
•
•
•
•
•
•
•
•
•
F1= feed molar flow rate (mol/s)
F2= product molar flow rate (mol/s)
y1= feed mole fraction of adsorbed component
y2= product mole fraction of adsorbed component
Mw = molecular weight of adsorbed component (g/mol)
ta= time the bed is in the adsorption stage of the cycle (s)
m1= maximum adsorbent loading (g/g adsorbent)
m2= minimum adsorbent loading (g/g adsorbent)
Ma= mass of adsorbent per bed (kg)
fL =fraction of bed that is fully loaded at end of adsorption phase of cycle
Chemical Engineering Design
Design Consideration
• fL less than 0.7 for 2 beds
For 4 or more bed this is almost equal to 1
• Preliminary analysis: cycle time 5 to 60 minutes for PSA
and 60 to 200 minutes for TSA
• See example 16.1 p 760 text
Chemical Engineering Design
Membrane Separation
• Thin membranes of polymer or metal can be used to
separate gases:
– Different species diffuse through a thin membrane at different rates:
– Different gases have different solubility in metal or polymer
• Permeate passes through the membrane and becomes
enriched in faster or more soluble species
• Retentate does not pass through membrane and becomes
enriched in slower or less soluble species
• Membranes have relatively low cost, but cannot obtain high
purity or recovery
• Membranes are therefore widely used for bulk separation of
gases
Chemical Engineering Design
Asymmetric Membrane
Dense
layer
0.1 to 1.0 m
Porous
support
0.1 to 1.0
mm
(not to scale)
• Polymer membranes are usually cast as asymmetric
membranes
• Thin, dense, active layer is supported on a thicker
stronger porous layer
• Backing cloth is used in some cases as support for
active layer
Chemical Engineering Design
Hollow Fiber Membranes
Feed
One hollow fiber (of
thousands)
Permeate
Potting
Retentate
Membrane cross
section
• Membranes are cast as long thin fibers
• A bundle of fibers is set into a resin (potting) that effectively
forms a tubesheet
• Feed is fed shell-side and permeate withdrawn from inside
the fibers
Chemical Engineering Design
Flat Sheet (Spiral Wound) Membranes
Enlarged cross section
Membrane envelopes
Dense
Porous
Porous
Dense
Feed
Retentate
Perforated tube
Permeate
Permeate
• Membranes are cast as sheets
• Sheets are glued back-to-back along edges to form an
envelope and attached to a perforated tube
• The assembly is then rolled up into a spiral-wound module
Chemical Engineering Design
Gas Separation Membranes
• SEM, TEM, STEM can be used for
microscopic analysis
• Note asymmetric structure
– Thin selective skin
– Porous support layer
UOP 5565M-73
Chemical Engineering Design
Membrane Flux and Permeability
• Flux of species i through the membrane is proportional to
partial pressure gradient:
Mi =
Pi

(p − p )
i, f
i, p
Mi = molar flux of component i (mol/m2.s),
Pi = permeability of membrane for omponent i (mol/m.s.bar),
 = membrane thickness (m),
pi,f = local partial pressure of component i on feed side (bar),
pi,p = local partial pressure of component i on permeate side (bar)
• Proportionality constant is the permeability divided by
membrane thickness
• Ratio of permeabilities of two species is the selectivity of
the membrane for species i relative to species j
Pi
S ij =
Pj
Chemical Engineering Design
Membrane Flow Pattern
Feed
• Integration of the flux
equation along the
membrane depends on
the flow pattern
• Note that only flat sheet
membranes can be used
in cross-flow mode
• Neither flat sheet nor
hollow fiber membranes
can use a sweep gas
Countercurrent
Permeate
Retentate
Feed
Cocurrent
Permeate
Retentate
Feed
Cross-flow
Permeate
Permeate
Retentate
© 2012 G.P. Towler / UOP. For educational use in conjunction with
Towler & Sinnott Chemical Engineering Design only. Do not copy
Chemical Engineering Design
Membrane Process Performance
• Membranes usually give low recovery of permeate
species (<95%, often <90%)
– Need to maintain a high enough partial pressure on retentate
side to give a reasonable flux at the outlet of the unit
– If outlet partial pressure is low, flux is low and area required
becomes large and costly
• Unless the selectivity is very high, membranes usually
give low purity on permeate side (<98%, often <95%)
• Hence membranes are used for bulk separations:
– Air separation (hollow fiber)
– CO2 rejection from natural gas (spiral wound)
– H2 recovery from mixtures with methane (hollow fiber)
Chemical Engineering Design
Permeate Recycle
Retentate
Feed
Permeate
• Permeate from 2nd module is recycled to feed of 1st module
• First module can now run under conditions that maximize permeate
purity (high selectivity) and we don’t have to worry about recovery
• Second module can run under conditions that maximize recovery (high
flux) and we don’t have to worry about purity
• Partial pressure of desired component in 1st module is increased
Chemical Engineering Design
Retentate Recycle
Feed
Retentate
Permeate
• Retentate from 2nd module is recycled to feed of 1st module
• Permeate now goes through two membranes in series, so final permeate
purity is increased
• First module can run at higher flux, lower selectivity as it makes a rough
separation
• An extra compressor is needed between the membrane stages
Chemical Engineering Design
Membrane Modules
UOP Separex modules for rejecting CO2 from natural gas
Chemical Engineering Design
Vapor-Liquid Separations
Flash
Single stage thermal
& phase eqbm
Evaporation
Single stage removal of
volatile solute or solvent
Distillation
Multiple stage separation
between identified light key
& heavy key components
Fractionation
Separation of multicomponent
mixture into fractions by
boiling ranges (e.g. in oil
refining)
Absorption
Removal of vapor
component using
non-volatile solvent
Stripping
Multi-stage removal
of volatile solute
from solvent
Multi-stage: see next lecture
See heat exchange lectures
Chemical Engineering Design
Vapor-Liquid Flash Drums
• Flash or knockout drums are widely used in chemical plants:
– Downstream of condensers and coolers
– Upstream of compressors and between compressor stages
– As reflux drums on columns
– In relief systems
• Design function is to separate liquid drops from vapor and
prevent vapor blowing out into liquid-filled lines by
maintaining liquid level control
• There will usually be ~1 to 2% liquid entrainment in the
vapor from a knockout drum unless a demister is used.
Chemical Engineering Design
Vertical Flash Drum
• Vessel diameter is chosen to give
vapor velocity that is less than
terminal velocity of drops
ut = 0.07[(L − v)/v]1/2
• Use 0.15 ut if there is no demister
• Allow 1 diameter above feed and
at least 0.6 diameters below feed
for settling, also allow 0.4
diameters for demister
• Height of liquid depends on level
control
Chemical Engineering Design
Liquid Level Control Bands
• Level control needs to allow for some natural variation
in liquid level due to splashing, etc.
• Alarms must not be set too close to normal operating
level or they will be a nuisance (& will be ignored)
• Operators need time to respond to alarms before
shutdown
• A typical assumption is about 2
LAHH shutdown trip
minutes between alarm and trip,
LAH alarm
so allow 10 minutes of liquid
residence time below feed
Normal operating band
• But note: midpoint of normal
operating band should be > Dv/2
LAL alarm
below feed point, so if 5 mins of
LALL shutdown trip
liquid holdup gives height < Dv/2,
use half a diameter to the
midpoint and 5 min holdup below.
5 min of
liquid or Dv/2
5 min of
liquid
Chemical Engineering Design
Design Vertical Separator
Chemical Engineering Design
Horizontal Flash Drum
• Bigger area for settling + more space for liquid holdup
• Trade-off is higher plot space and stronger foundations needed to
support vessel
• Often used when process control demands some liquid inventory, e.g.
reflux drums
• Design is more complex than vertical drum – see Example 16.3 p.772
Chemical Engineering Design
Liquid-Liquid Separations
Decanting
Single stage thermal
& phase eqbm
Extraction
Multi-stage
contacting of two
liquid phases
Mixer-Settler
Single theoretical stage
extraction process
Often 2 or 3 stages are still
cheaper than a column
Membrane
Based on differences in relative
permeability of components
Membrane can be used to keep
two solvents from mixing
Chemical Engineering Design
Horizontal Decanter
Light
liquid
Vent
Heavy
liquid
Feed
Dispersion
band
Drain
• Design is similar to knockout drum: allow droplets to settle and
provide adequate holdup for level control – see Example 16.4 p776
• Siphon take-off can control level without instruments if densities are
constant
Chemical Engineering Design
Separation Columns
Chemical Engineering Design
Separation Columns
• Distillation Basics
• Absorption & Stripping
• Column Internals & Hydraulics: Trays
• Column Internals & Hydraulics: Packing
• Column Design & Optimization
• Complex Columns
Chemical Engineering Design
Distillation
• Distillation is the most
commonly used separation
process in the chemical
industry
– Relatively cheap
– Reliable and low
maintenance
• Distillation costs can be
80% of capital and utilities
in some processes
Chemical Engineering Design
Distillation Column Nomenclature
Condenser
Vapor
Reflux drum (receiver)
(External) reflux
Distillate, D
• Feed is separated into distillate
enriched in light components (high
volatility) and bottoms enriched in
heavy components (low volatility)
Feed, F
Trays
Reboiler
Bottoms, B
• Vapor is generated in reboiler and
passes up through many trays to
overhead condenser
• Liquid reflux flows down over trays,
giving counter-current contacting
Chemical Engineering Design
Distillation Specifications
• Recovery
• The fraction of a component’s flow rate in the feed that is recovered in either
the distillate of bottoms
• Purity
• The mole fraction of a component in either the distillate or the bottoms
• A distillation column only has two degrees of freedom
(split fraction and energy balance), so only two things
can be specified
• e.g. recovery and purity of one component in the distillate
• Specifications must be feasible: be careful when
specifying multicomponent distillation
Chemical Engineering Design
Feasible Specifications Example:
Three Component System
A
Composition
diagram
Relative volatility A > B > C
F
B
C
• If we want to separate feed F and achieve a distillate
product that is >99% A (purity specification), which of the
following specifications are feasible
– 30% B in bottoms?
– 99% recovery of B in bottoms?
– 98% recovery of A in distillate?
– 95% C in bottoms?
Chemical Engineering Design
Feasible Specifications Example
A
Composition
diagram
D
Relative volatility A > B > C
F
B
B
C
• Feed, distillate and bottoms must mass balance!
• For multicomponent distillation, you should always do a
quick mass balance to make sure specifications are
feasible
Chemical Engineering Design
Distillation Concepts: Theoretical Stage
Liquid
flow L
xn-1
Liquid and vapor leaving the
stage are in equilibrium
yn
Tray n
xn
yi,n = Ki xi,n
for all components i
yn+1
Vapor
flow V
• Theoretical stages are useful to understand distillation
• Real trays do not usually achieve theoretical stage
performance
Chemical Engineering Design
What Really Happens on a Tray
• Liquid comes down a
downcomer from tray
above and is distributed
across the active area of
the tray
• Vapor bubbles up through
holes creating a froth on
the tray
• Froth or liquid overflows
into next downcomer
• Some columns use packing
instead (see below)
Chemical Engineering Design
Binary Distillation: McCabe-Thiele
Diagrams
• With two components, we can
plot equilibrium vapor mole
fraction y vs. liquid mole
fraction x
• We can also plot operating lines
on the y-x diagram (see
distillation Ch 17 for equations)
• Hence step off between OL and
EL to find number of trays
• Useful for conceptual
understanding, but not used in
practice since the 1960s
Chemical Engineering Design
Non-Constant Relative Volatility
• For non-ideal mixtures,
Ki varies with
composition
• Volatility order can
change: azeotropes
Example: Acetone water (see Ch17 for details)
• McCabe Thiele
diagrams illustrate that
azeotropes and near
azeotropes pinch the
EL and OL and cause
difficult separations
Chemical Engineering Design
Azeotrope Formation
T
y
x or y
x
• Volatility order changes
• Azeotrope can be minimum (shown) or maximum boiling
• Common for mixtures with high polarity or hydrogen
bonding (water, oxygenates)
Chemical Engineering Design
Multicomponent Distillation
• In industrial problems, there are almost never just two
components
– Usually start from a mixture with many compounds
– There are always trace components present that need to be
tracked
• Multicomponent distillation cannot be solved graphically
– Shortcut correlations (1930s) can be used for initial estimates
• Fenske, Underwood, Gilliland, Smoker, Kremser, etc: see Ch 17
– Rigorous tray-to-tray calculations (1960s) are used for detailed
simulation
• Multicomponent distillation is always designed using
commercial process simulators
Chemical Engineering Design
• See effect of reflux ration on Number of stages Example
17.3 p.836
• Estimate feed location example 17.4 page 838
Chemical Engineering Design
Stage Efficiency
• Real trays do not usually achieve theoretical stage
performance
Overall stage efficiency =
Number of theoretical stages
Number of actual trays for same separation
Liquid droplets can be
entrained to tray above
Vapor can flow
up downcomer
Liquid can weep
to tray below
Vapor concentration
varies along the tray
Chemical Engineering Design
Stage Efficiency
• Depends on tray design, fluid properties, relative volatility
• O’Connell’s
correlation (1946)
for hydrocarbons,
bubble cap trays
•
Eo = 51 – 32.5 log ()
• See Ch 17 for
details
Molar average liquid
viscosity mMs/m2
Relative volatility
of light key
• For initial estimate
of stage efficiency
usually use 0.6 to
0.7
Chemical Engineering Design
Separation Columns
• Distillation Basics
• Absorption & Stripping
• Column Internals & Hydraulics: Trays
• Column Internals & Hydraulics: Packing
• Column Design & Optimization
• Complex Columns
Chemical Engineering Design
Absorption and Stripping
• Selectively absorb a component from the vapor phase
into a solvent
• Regenerate the solvent by stripping using steam, air,
nitrogen or reboiled solvent
• Two types of solvent:
– Physical solvent (physisorption): physical interactions only
• Examples: Methanol (Rectisol process), PEG DME (Selexol process)
– Chemical solvent (chemisorption): chemical + physical
interactions
• Examples: Alkanolamines (MEA, DEA, MDEA) for acid gases
Chemical Engineering Design
Example: Acid Gas Scrubbing
Treated Gas
Water
Filter
Flash Gas
Water
Rich Flash Drum
•
Acid Gas
Knock-Out
Drum
Amine
Stripper
Amine
Absorber
Feed Gas
Cooler
Lean Amine
Cooler
Lean/Rich
Exchanger
Reboiler
For removal of H2S &/or CO2 from gases using solvents such as
methyldiethanolamine (MDEA)
Chemical Engineering Design
Absorption
Liquid
flow L
xT
• Mass balance around any tray:
yT
Gy + LxT = GyT + Lx
•
x
xB
Gy – Lx = GyT – LxT = GyB – LxB = constant
y
yB
Hence
Gas
flow G
• This is the operating line for
absorption or stripping
Liquid and vapor leaving the
stage are in equilibrium
yi,n = Hi xi,n
for all components i,
where Hi is the
Henry’s Law constant
Chemical Engineering Design
Absorption & Stripping
Liquid
flow L
y=Hx
y
xT
yT
x
y
x
xB
yB
Gas
flow G
If liquid and vapor leaving the stage
are in equilibrium
yi,n = Hi xi,n
for all components i, where Hi
is the Henry’s Law constant
Chemisorption systems tend to be
more curved
Chemical Engineering Design
Absorption
Liquid
flow L
G (yT – y)= L (xT – x)
y
xT
yT
x
y
yT
y=Hx
yB
x
xB
yB
Gas
flow G
• In absorption the operating line plots
above the equilibrium line
• xT, yT from overall mass balance
• Step off between OL & EL to find
number of stages needed
Chemical Engineering Design
Stripping
Liquid
flow L
y
xT
y=Hx
yT
yT
x
y
yB
G (yT – y)= L (xT – x)
x
xB
yB
Gas
flow G
• In stripping, the OL plots below
the EL
Chemical Engineering Design
Absorption & Stripping: Heat Effects
y
Increasing T
x
y
• Increasing temperature decreases H
and makes stripping easier
• For chemisorption, the EL is
concave upwards. The stripper must
provide heat of reaction as well as
stripping vapor flow
• If EL has high curvature then it is
difficult to strip to low concentrations
x
• But concentration in the stripped
liquid is what determines the
concentration in the gas leaving the
absorber!
Chemical Engineering Design
Absorption and Stripping
• In industrial practice, absorbers and strippers are
designed using commercial simulation programs
– rigorous fractionation models can be used to solve absorbers
or strippers with appropriate reboiler/condenser options
• It is worth doing quick calculations at either end of the
column to make sure that L/G > H for an absorber or
L/G < H for a stripper. If not, then column could be
pinched or infeasible
• Optimization usually requires trading off solvent
circulation (hence energy costs for stripping) vs.
number of trays in absorber and stripper (capital cost)
Chemical Engineering Design
Separation Columns
• Distillation Basics
• Absorption & Stripping
• Column Internals & Hydraulics: Trays
• Column Internals & Hydraulics: Packing
• Column Design & Optimization
• Complex Columns
Chemical Engineering Design
Tray Dimensions
Weir length
Tray spacing
Active area
Height over weir
Downcomer
backup
Weir height
Downcomer clearance
Active area
Chemical Engineering Design
Tray Types
Sieve Tray
Valve Tray
•
Tray has holes punched in it
•
•
Liquid is retained by vapor upflow
Moving valves close when vapor
flow is low
•
Cheapest design
•
If valves fall off then it turns into a
sieve tray
•
Resilient to solids, corrosion,
particularly if large holes used
•
Many proprietary designs
•
Better turndown performance (5:1
range)
•
Cost about 1.2 times sieve trays
•
Lowest pressure drop
•
Turndown performance is poor
due to weeping (2:1 range)
Chemical Engineering Design
Tray Types
Fixed Valves
Bubble Caps
•
Valves are stamped into the tray,
so always open
•
Ensure that liquid level is
maintained on each tray
•
Less prone to corrosion or fouling
than moving valves, but poorer
turndown performance
•
Turndown performance is
better, but most prone to fouling
and corrosion damage
•
One side is always open, so
orientation of tray is important
•
Cost can be double sieve trays
•
Pressure drop is highest
•
Widely used before 1970s, but
seldom now
•
Cost about 1.1 times sieve trays
Chemical Engineering Design
Plate Hydraulics
• Plate hydraulic behavior is best understood by watching
the video “Film A: Design and Performance of
Fractionating Devices – Part 1 – Hydraulics of trays with
downcomers” from Fractionation Research Institute
– Available from www.FRI.org
– $50 for FRI members, $400 for non-members
• Widely recognized as the classic distillation movie!
Chemical Engineering Design
Column Operating Window
Vapor rate
Excessive
entrainment
Coning
Jet flooding
Area of satisfactory
operation
Downcomer flooding
Weeping
Liquid rate
• Plate design must ensure good contacting between phases
• Coning: vapor bypasses liquid
• Weeping: liquid drains through to tray below
• Usually design to operate near (~ 70 to 80% of) flooding limits so as
to allow for turn-down
Chemical Engineering Design
Jet Flooding
• Excessive liquid entrained in rising vapor
• Effected by
– Capacity factor
– Liquid rate (weir rate)
– Tray spacing
– Number of passes
– Fluid properties
– Perforation type and shape
Capacity factor
 V


C B = V 
  L − V 
0 .5
V = vapor velocity
• For hydrocarbons, jet flooding occurs if CB is
– > ~ 0.25 with 18-24” tray spacing
– > ~ 0.35 with 24-36” tray spacing
– > ~ 0.45 ultimate capacity
Chemical Engineering Design
Downcomer Flooding
• Downcomer is unable to move liquid down the column
• Static head to move liquid is limited to a fraction of a tray
spacing and must balance
– Exit loss
– Liquid head on tray
– Tray pressure drop
• If there is too much liquid velocity then vapor is entrained in
the downcomer and exit head is lost
• Downcomer flooding usually occurs around 0.2 to 0.6 ft/s
downcomer velocity
Chemical Engineering Design
Tray Design
• Design correlations for tray design and prediction of flooding
are in the literature:
•
•
•
•
Perry’s Chemical Engineer’s Handbook (McGraw-Hill)
Kister, H. “Distillation Design” & “Distillation Operation” (McGraw-Hill)
FRI publications
Ch 17 has example procedure
• Tray design is automated in most process simulation
programs
• Tool may need to be “switched on” or activated
• Beware of “rating” vs. “sizing” – rating calcs are to test performance of a given
existing column, e.g. for revamp
• Tray designs should always be confirmed by a column
internals vendor
• They use more sophisticated proprietary programs
• They can provide much better estimates of pressure drop and stage efficiency
Chemical Engineering Design
Typical Tray Design Procedure
See Example 17.6 p878
1. Calculate maximum and minimum vapor and liquid flow rates
2. Select trial tray spacing (24” typical default)
3. Estimate diameter based on flooding correlations
4. Make trial tray layout (active area, hole size, hole area, weir height)
5. Check weeping, pressure drop, downcomer flooding & entrainment
(return to 4 if necessary)
6. Finalize tray details (calming zones, hole pitch)
7. Confirm percentage flooding
8. Optimize design
Usually automated in software!
Chemical Engineering Design
Separation Columns
• Distillation Basics
• Absorption & Stripping
• Column Internals & Hydraulics: Trays
• Column Internals & Hydraulics: Packing
• Column Design & Optimization
• Complex Columns
Chemical Engineering Design
Packed Columns
• Instead of trays, liquid is distributed
over a packing inside the column
• Packing requirements
•
•
•
•
Provide high surface area
Minimal resistance to gas flow
Give uniform gas flow across cross-section
Promote liquid distribution and re-mixing
• Two types of packing:
– Random (a.k.a. dumped)
• Formed particles made of metal or ceramics
– Structured
• Metal sheets with holes and patterns punched
in them, assembled into blocks
Chemical Engineering Design
Random Packing
•
Raschig rings
•
Pall rings
•
Berl saddles
•
Intalox saddles
•
Hypac
•
Super Intalox
•
Many other variants available from vendors
Chemical Engineering Design
Structured Packing
• Folded or stamped perforated sheets assembled into a
block, with high voidage
• Give better efficiency (lower HETP: height equivalent to
a theoretical plate), but higher cost than random
packings
Chemical Engineering Design
Typical Packing Efficiencies
For Pall Rings: See Ch 17 for Other Packings
System
Pressure
(kPa)
Column
diameter (m)
Packing size
(mm)
HETP
(m)
Hydrocarbon absorption
6000
0.9
50
0.85
Water absorption of acetone
101
0.6
50
0.75
Pentane-propane distillation
101
0.46
25
0.46
Acetone-water distillation
101
0.46
25
0.37
Acetone-water distillation
101
0.38
38
0.45
MEK-toluene
101
0.38
25
0.35
• Vendors should be contacted for proprietary types and for
structured packing
Chemical Engineering Design
Trayed vs. Packed Columns
Trayed Columns
Packed Columns
•
Wider range of liquid and vapor
flow rates
•
Not good for low liquid rates
•
Easier to design and predict stage
efficiency
•
Cheaper for corrosive systems
•
•
Easier to withdraw side streams,
use side columns or provide
intermediate heating or cooling
Reduced liquid hold-up, so safer
for systems where inventory must
be minimized
•
Good for high liquid throughputs
(no downcomers)
•
Lower pressure drop, so better for
vacuum columns
•
Cheaper for small columns where
trays are difficult to install (< 2ft
diameter)
•
Handle foaming systems better
•
Easier to clean and handle fouling
•
Don’t have to worry about liquid
distribution, so better for large
diameters
•
Usually cheaper (sieve trays cost
~20% of packing)
Chemical Engineering Design
Packed Column Design
• Simulation programs have correlations for pressure drop
and HETP for some of the older and more common types of
random packing
– Column diameter is set by operating at typically 80% of flooding limit
– Column height is set by HETP and number of stages needed
• Some vendors have provided approximate correlations of
structured packing performance to simulation companies
• For conceptual design, costs can be estimated in Aspen
ICARUS
• Vendors should always be contacted to confirm packing
performance
Chemical Engineering Design
Column Internals Exercise
• What column internals would you specify for the following
applications?
• Gasoline debutanizer
• Sieve trays
• Ethanol-water (azeotropic distillation)
• Sieve trays
• Acid gas scrubbing using alkanolamine
(corrosive, high liquid loading)
• Packing (or sieve
trays)
• Styrene monomer re-run column
(polymerizes easily)
• Packing or sieve
trays
• Fermentation broth stripper (high fouling)
• Sieve trays
• VOC recovery fractionation (high turndown) • Bubble caps
Chemical Engineering Design
Separation Columns
• Distillation Basics
• Absorption & Stripping
• Column Internals & Hydraulics: Trays
• Column Internals & Hydraulics: Packing
• Column Design & Optimization
• Complex Columns
Chemical Engineering Design
Column Design Procedure
• Define feed
• Define product specifications
• Set column pressure
• Design and select internals
• Optimize column design
Chemical Engineering Design
Column Pressure Considerations
Higher pressure
Lower pressure
•
•
Fewer trays, bigger diameter
•
Try to avoid dropping condenser
temperature to point where
(expensive) refrigeration is needed
•
Try to maintain total condenser to
obviate net gas compression and
minimize vent gas loss
•
Try to avoid vacuum operation for
safety, quality and reliability reasons
Lower relative volatility so need more
trays
•
Higher vapor density, so reduced
column diameter
•
Higher boiling and dew points, so
hotter reboiler and condenser
–
–
•
May mean that reboiler utility becomes
more expensive
Try to avoid needing a fired heater if
possible
In some cases, high reboiler
temperatures can cause product
degradation
–
–
Monomers can polymerize
Foods and flavors can degrade
Usually set column pressure
to allow condenser to run on
cooling water or air cooling
Chemical Engineering Design
Column Optimization
• It is not trivial to formulate and solve distillation column
optimization problems
– Discrete variables: number of trays, feed tray
– Continuous variables: pressure, reflux rate
– Many options for complex columns
– Column sequencing
• Operational constraints and design margins may make
the mathematically optimal design impractical
• Most designs are “close to optimal” rather than rigorously
optimized
Chemical Engineering Design
Column Optimization
• How do you expect the following parameters to vary with
reflux ratio?
Diameter
Reflux
V = (R+1)D
D = distillate
Reflux
Reflux
Total Costs
Condenser Duty
Reflux
Reboiler Duty
d2  R
Nmin
Height
Number of trays
Rmin
Reflux
Optimum
Reflux
Chemical Engineering Design
Total Costs
Column Optimization
Optimum
Reflux
• Calculated optimum reflux ratio is usually between 1.05 and
1.10 times Rmin
• Most designers choose to use a somewhat larger reflux
ratio to allow some design margin: typically 1.15 R min
• Reflux and pressure are main optimization parameters,
number of trays and feed tray usually have lower impact on
cost
Chemical Engineering Design
Separation Columns
• Distillation Basics
• Absorption & Stripping
• Column Internals & Hydraulics: Trays
• Column Internals & Hydraulics: Packing
• Column Design & Optimization
• Complex Columns
Chemical Engineering Design
Side Streams
Distillate
• Withdrawing a side stream allows a
component to be removed that has
accumulated in the middle of the
column
• OK if low purity is needed for
middle product and middle product
has low flow rate
Feed
Side
product
• Often used to remove small
amounts of intermediate boilers –
cheaper than adding a column if
product loss can be tolerated or
stream is recycled
• Side stream can be above or below
feed
Bottoms
Chemical Engineering Design
Side Streams
D
F
Side
product
Side product
composition
B
• Side stream purity is usually low and some amount of
desired product is usually lost
Chemical Engineering Design
Side Strippers and Rectifiers
• Use of side strippers and rectifiers
allows higher purity in a side stream
Distillate •
Side
stripper
For intermediates withdrawn above the
feed tray use a side stripper to strip out
light components
• For intermediates withdrawn below the
feed tray, use a side rectifier to knock
down heavy components:
Feed
Bottoms
• Side strippers are widely used in
petroleum refining
Chemical Engineering Design
Intermediate Reboiling or Condensing
Distillate
• A side reboiler allows part of the heat
requirement to be added at a lower
temperature using cheaper hot utility
(or less area)
• A side condenser allows part of the
heat to be removed at a higher
temperature using cheaper cold utility
(or less area)
Feed
Qr2
Side
reboiler
Qr1
Bottoms
• These options are considered for
columns with high heat duties,
refrigeration, or wide temperature
difference between reboiler and
condenser
• Another way to add or remove heat at
intermediate points is with a liquid
“pumparound” circuit that can be heat
exchanged with other process streams
Chemical Engineering Design
Dividing Wall Distillation
A
A
B
C
B
C
A
A
B
C
B
C
• Vertical wall separates column sections
• Three products can be made using a single shell (middle product purity
is much better than simple side draw)
• Up to 30% savings in capital and energy costs
• BASF have built 30 since 1984, BP 1, SASOL 2
Chemical Engineering Design
UOP PEP Splitter Dividing Wall Column
Mechanical Pictures
Source: UOP
Chemical Engineering Design
Complex Column Simulation
• Most simulators allow specification of side streams, side
columns and intermediate exchangers
• Adding an additional stream or exchanger adds a degree
of freedom, so one more column specification is required
– Usually side-stream draw rate or exchanger duty
• DWC is usually not a standard option in the simulation
program, but can be built up from component sections
Chemical Engineering Design
High Capacity Trays
• Closed downcomers
• Maximize active area
• Directional valves
• Enhanced phase separation
• Reduced tray spacing
UOP MD Tray: Source UOP
• Examples:
– UOP: MD , ECMD , PFMD , SimulFlow
– Shell: HiFi ,ConSep
– Koch-Glitsch: SuperFrac®, Ultra-Frac®
– Sulzer: VGPlus
Chemical Engineering Design
High Capacity Trays
• Increased active area or enhanced phase separation
enable much higher throughput
– E.g.: UOP SimulFlow
has 230% capacity of sieve trays
• Efficiency is usually similar to sieve or valve trays, but
some designs have enhanced efficiency as well as
capacity
• Enhanced capacity trays are used a lot in revamps to save
buying a new shell, also for very large diameter new
columns
• Details must be obtained from vendors
Chemical Engineering Design
Separation Columns
Chemical Engineering Design
Separation Columns
• Distillation Basics
• Absorption & Stripping
• Column Internals & Hydraulics: Trays
• Column Internals & Hydraulics: Packing
• Column Design & Optimization
• Complex Columns
Chemical Engineering Design
Distillation
• Distillation is the most
commonly used separation
process in the chemical
industry
– Relatively cheap
– Reliable and low
maintenance
• Distillation costs can be
80% of capital and utilities
in some processes
Chemical Engineering Design
Distillation Column Nomenclature
Condenser
Vapor
Reflux drum (receiver)
(External) reflux
Distillate, D
• Feed is separated into distillate
enriched in light components (high
volatility) and bottoms enriched in
heavy components (low volatility)
Feed, F
Trays
Reboiler
Bottoms, B
• Vapor is generated in reboiler and
passes up through many trays to
overhead condenser
• Liquid reflux flows down over trays,
giving counter-current contacting
Chemical Engineering Design
Distillation Specifications
• Recovery
• The fraction of a component’s flow rate in the feed that is recovered in either
the distillate of bottoms
• Purity
• The mole fraction of a component in either the distillate or the bottoms
• A distillation column only has two degrees of freedom
(split fraction and energy balance), so only two things
can be specified
• e.g. recovery and purity of one component in the distillate
• Specifications must be feasible: be careful when
specifying multicomponent distillation
Chemical Engineering Design
Feasible Specifications Example:
Three Component System
A
Composition
diagram
Relative volatility A > B > C
F
B
C
• If we want to separate feed F and achieve a distillate
product that is >99% A (purity specification), which of the
following specifications are feasible
– 30% B in bottoms?
– 99% recovery of B in bottoms?
– 98% recovery of A in distillate?
– 95% C in bottoms?
Chemical Engineering Design
Feasible Specifications Example
A
Composition
diagram
D
Relative volatility A > B > C
F
B
B
C
• Feed, distillate and bottoms must mass balance!
• For multicomponent distillation, you should always do a
quick mass balance to make sure specifications are
feasible
Chemical Engineering Design
Distillation Concepts: Theoretical Stage
Liquid
flow L
xn-1
Liquid and vapor leaving the
stage are in equilibrium
yn
Tray n
xn
yi,n = Ki xi,n
for all components i
yn+1
Vapor
flow V
• Theoretical stages are useful to understand distillation
• Real trays do not usually achieve theoretical stage
performance
Chemical Engineering Design
What Really Happens on a Tray
• Liquid comes down a
downcomer from tray
above and is distributed
across the active area of
the tray
• Vapor bubbles up through
holes creating a froth on
the tray
• Froth or liquid overflows
into next downcomer
• Some columns use packing
instead (see below)
Chemical Engineering Design
Binary Distillation: McCabe-Thiele
Diagrams
• With two components, we can
plot equilibrium vapor mole
fraction y vs. liquid mole
fraction x
• We can also plot operating lines
on the y-x diagram (see
distillation class or Ch 17 for
equations)
• Hence step off between OL and
EL to find number of trays
• Useful for conceptual
understanding, but not used in
practice since the 1960s
Chemical Engineering Design
Non-Constant Relative Volatility
• For non-ideal mixtures,
Ki varies with
composition
• Volatility order can
change: azeotropes
Example: Acetone water (see Ch17 for details)
• McCabe Thiele
diagrams illustrate that
azeotropes and near
azeotropes pinch the
EL and OL and cause
difficult separations
Chemical Engineering Design
Azeotrope Formation
T
y
x or y
x
• Volatility order changes
• Azeotrope can be minimum (shown) or maximum boiling
• Common for mixtures with high polarity or hydrogen
bonding (water, oxygenates)
Chemical Engineering Design
Multicomponent Distillation
• In industrial problems, there are almost never just two
components
– Usually start from a mixture with many compounds
– There are always trace components present that need to be
tracked
• Multicomponent distillation cannot be solved graphically
– Shortcut correlations (1930s) can be used for initial estimates
• Fenske, Underwood, Gilliland, Smoker, Kremser, etc: see Ch 17
– Rigorous tray-to-tray calculations (1960s) are used for detailed
simulation
• Multicomponent distillation is always designed using
commercial process simulators
Chemical Engineering Design
Stage Efficiency
• Real trays do not usually achieve theoretical stage
performance
Overall stage efficiency =
Number of theoretical stages
Number of actual trays for same separation
Liquid droplets can be
entrained to tray above
Vapor can flow
up downcomer
Liquid can weep
to tray below
Vapor concentration
varies along the tray
Chemical Engineering Design
Stage Efficiency
• Depends on tray design, fluid properties, relative volatility
• O’Connell’s
correlation (1946)
for hydrocarbons,
bubble cap trays
•
Eo = 51 – 32.5 log ()
• See Ch 17 for
details
Molar average liquid
viscosity mMs/m2
Relative volatility
of light key
• For initial estimate
of stage efficiency
usually use 0.6 to
0.7
Chemical Engineering Design
Separation Columns
• Distillation Basics
• Absorption & Stripping
• Column Internals & Hydraulics: Trays
• Column Internals & Hydraulics: Packing
• Column Design & Optimization
• Complex Columns
Chemical Engineering Design
Absorption and Stripping
• Selectively absorb a component from the vapor phase
into a solvent
• Regenerate the solvent by stripping using steam, air,
nitrogen or reboiled solvent
• Two types of solvent:
– Physical solvent (physisorption): physical interactions only
• Examples: Methanol (Rectisol process), PEG DME (Selexol process)
– Chemical solvent (chemisorption): chemical + physical
interactions
• Examples: Alkanolamines (MEA, DEA, MDEA) for acid gases
Chemical Engineering Design
Example: Acid Gas Scrubbing
Treated Gas
Water
Filter
Flash Gas
Water
Rich Flash Drum
•
Acid Gas
Knock-Out
Drum
Amine
Stripper
Amine
Absorber
Feed Gas
Cooler
Lean Amine
Cooler
Lean/Rich
Exchanger
Reboiler
For removal of H2S &/or CO2 from gases using solvents such as
methyldiethanolamine (MDEA)
Chemical Engineering Design
Absorption
Liquid
flow L
xT
• Mass balance around any tray:
yT
Gy + LxT = GyT + Lx
•
x
xB
Gy – Lx = GyT – LxT = GyB – LxB = constant
y
yB
Hence
Gas
flow G
• This is the operating line for
absorption or stripping
Liquid and vapor leaving the
stage are in equilibrium
yi,n = Hi xi,n
for all components i,
where Hi is the
Henry’s Law constant
Chemical Engineering Design
Absorption & Stripping
Liquid
flow L
y=Hx
y
xT
yT
x
y
x
xB
yB
Gas
flow G
If liquid and vapor leaving the stage
are in equilibrium
yi,n = Hi xi,n
for all components i, where Hi
is the Henry’s Law constant
Chemisorption systems tend to be
more curved
Chemical Engineering Design
Absorption
Liquid
flow L
G (yT – y)= L (xT – x)
y
xT
yT
x
y
yT
y=Hx
yB
x
xB
yB
Gas
flow G
• In absorption the operating line plots
above the equilibrium line
• xT, yT from overall mass balance
• Step off between OL & EL to find
number of stages needed
Chemical Engineering Design
Stripping
Liquid
flow L
y
xT
y=Hx
yT
yT
x
y
yB
G (yT – y)= L (xT – x)
x
xB
yB
Gas
flow G
• In stripping, the OL plots below
the EL
Chemical Engineering Design
Absorption & Stripping: Heat Effects
y
Increasing T
x
y
• Increasing temperature decreases H
and makes stripping easier
• For chemisorption, the EL is
concave upwards. The stripper must
provide heat of reaction as well as
stripping vapor flow
• If EL has high curvature then it is
difficult to strip to low concentrations
x
• But concentration in the stripped
liquid is what determines the
concentration in the gas leaving the
absorber!
Chemical Engineering Design
Absorption and Stripping
• In industrial practice, absorbers and strippers are
designed using commercial simulation programs
– rigorous fractionation models can be used to solve absorbers
or strippers with appropriate reboiler/condenser options
• It is worth doing quick calculations at either end of the
column to make sure that L/G > H for an absorber or
L/G < H for a stripper. If not, then column could be
pinched or infeasible
• Optimization usually requires trading off solvent
circulation (hence energy costs for stripping) vs.
number of trays in absorber and stripper (capital cost)
Chemical Engineering Design
Separation Columns
• Distillation Basics
• Absorption & Stripping
• Column Internals & Hydraulics: Trays
• Column Internals & Hydraulics: Packing
• Column Design & Optimization
• Complex Columns
Chemical Engineering Design
Tray Dimensions
Weir length
Tray spacing
Active area
Height over weir
Downcomer
backup
Weir height
Downcomer clearance
Active area
Chemical Engineering Design
Tray Types
Sieve Tray
Valve Tray
•
Tray has holes punched in it
•
•
Liquid is retained by vapor upflow
Moving valves close when vapor
flow is low
•
Cheapest design
•
If valves fall off then it turns into a
sieve tray
•
Resilient to solids, corrosion,
particularly if large holes used
•
Many proprietary designs
•
Better turndown performance (5:1
range)
•
Cost about 1.2 times sieve trays
•
Lowest pressure drop
•
Turndown performance is poor
due to weeping (2:1 range)
Chemical Engineering Design
Tray Types
Fixed Valves
Bubble Caps
•
Valves are stamped into the tray,
so always open
•
Ensure that liquid level is
maintained on each tray
•
Less prone to corrosion or fouling
than moving valves, but poorer
turndown performance
•
Turndown performance is
better, but most prone to fouling
and corrosion damage
•
One side is always open, so
orientation of tray is important
•
Cost can be double sieve trays
•
Pressure drop is highest
•
Widely used before 1970s, but
seldom now
•
Cost about 1.1 times sieve trays
© 2012 G.P. Towler / UOP. For educational use in conjunction with
Towler & Sinnott Chemical Engineering Design only. Do not copy
Chemical Engineering Design
Plate Hydraulics
• Plate hydraulic behavior is best understood by watching
the video “Film A: Design and Performance of
Fractionating Devices – Part 1 – Hydraulics of trays with
downcomers” from Fractionation Research Institute
– Available from www.FRI.org
– $50 for FRI members, $400 for non-members
• Widely recognized as the classic distillation movie!
Chemical Engineering Design
Column Operating Window
Vapor rate
Excessive
entrainment
Coning
Jet flooding
Area of satisfactory
operation
Downcomer flooding
Weeping
Liquid rate
• Plate design must ensure good contacting between phases
• Coning: vapor bypasses liquid
• Weeping: liquid drains through to tray below
• Usually design to operate near (~ 70 to 80% of) flooding limits so as
to allow for turn-down
Chemical Engineering Design
Jet Flooding
• Excessive liquid entrained in rising vapor
• Effected by
– Capacity factor
– Liquid rate (weir rate)
– Tray spacing
– Number of passes
– Fluid properties
– Perforation type and shape
Capacity factor
 V


C B = V 
  L − V 
0 .5
V = vapor velocity
• For hydrocarbons, jet flooding occurs if CB is
– > ~ 0.25 with 18-24” tray spacing
– > ~ 0.35 with 24-36” tray spacing
– > ~ 0.45 ultimate capacity
Chemical Engineering Design
Downcomer Flooding
• Downcomer is unable to move liquid down the column
• Static head to move liquid is limited to a fraction of a tray
spacing and must balance
– Exit loss
– Liquid head on tray
– Tray pressure drop
• If there is too much liquid velocity then vapor is entrained in
the downcomer and exit head is lost
• Downcomer flooding usually occurs around 0.2 to 0.6 ft/s
downcomer velocity
Chemical Engineering Design
Tray Design
• Design correlations for tray design and prediction of flooding
are in the literature:
•
•
•
•
Perry’s Chemical Engineer’s Handbook (McGraw-Hill)
Kister, H. “Distillation Design” & “Distillation Operation” (McGraw-Hill)
FRI publications
Ch 17 has example procedure
• Tray design is automated in most process simulation
programs
• Tool may need to be “switched on” or activated
• Beware of “rating” vs. “sizing” – rating calcs are to test performance of a given
existing column, e.g. for revamp
• Tray designs should always be confirmed by a column
internals vendor
• They use more sophisticated proprietary programs
• They can provide much better estimates of pressure drop and stage efficiency
Chemical Engineering Design
Typical Tray Design Procedure
1. Calculate maximum and minimum vapor and liquid flow rates
2. Select trial tray spacing (24” typical default)
3. Estimate diameter based on flooding correlations
4. Make trial tray layout (active area, hole size, hole area, weir height)
5. Check weeping, pressure drop, downcomer flooding & entrainment
(return to 4 if necessary)
6. Finalize tray details (calming zones, hole pitch)
7. Confirm percentage flooding
8. Optimize design
Usually automated in software, but designer must be
vigilant about warnings!
Chemical Engineering Design
Separation Columns
• Distillation Basics
• Absorption & Stripping
• Column Internals & Hydraulics: Trays
• Column Internals & Hydraulics: Packing
• Column Design & Optimization
• Complex Columns
Chemical Engineering Design
Packed Columns
• Instead of trays, liquid is distributed
over a packing inside the column
• Packing requirements
•
•
•
•
Provide high surface area
Minimal resistance to gas flow
Give uniform gas flow across cross-section
Promote liquid distribution and re-mixing
• Two types of packing:
– Random (a.k.a. dumped)
• Formed particles made of metal or ceramics
– Structured
• Metal sheets with holes and patterns punched
in them, assembled into blocks
Chemical Engineering Design
Random Packing
•
Raschig rings
•
Pall rings
•
Berl saddles
•
Intalox saddles
•
Hypac
•
Super Intalox
•
Many other variants available from vendors
Chemical Engineering Design
Structured Packing
• Folded or stamped perforated sheets assembled into a
block, with high voidage
• Give better efficiency (lower HETP: height equivalent to
a theoretical plate), but higher cost than random
packings
Chemical Engineering Design
Typical Packing Efficiencies
For Pall Rings: See Ch 17 for Other Packings
System
Pressure
(kPa)
Column
diameter (m)
Packing size
(mm)
HETP
(m)
Hydrocarbon absorption
6000
0.9
50
0.85
Water absorption of acetone
101
0.6
50
0.75
Pentane-propane distillation
101
0.46
25
0.46
Acetone-water distillation
101
0.46
25
0.37
Acetone-water distillation
101
0.38
38
0.45
MEK-toluene
101
0.38
25
0.35
• Vendors should be contacted for proprietary types and for
structured packing
Chemical Engineering Design
Trayed vs. Packed Columns
Trayed Columns
Packed Columns
•
Wider range of liquid and vapor
flow rates
•
Not good for low liquid rates
•
Easier to design and predict stage
efficiency
•
Cheaper for corrosive systems
•
•
Easier to withdraw side streams,
use side columns or provide
intermediate heating or cooling
Reduced liquid hold-up, so safer
for systems where inventory must
be minimized
•
Good for high liquid throughputs
(no downcomers)
•
Lower pressure drop, so better for
vacuum columns
•
Cheaper for small columns where
trays are difficult to install (< 2ft
diameter)
•
Handle foaming systems better
•
Easier to clean and handle fouling
•
Don’t have to worry about liquid
distribution, so better for large
diameters
•
Usually cheaper (sieve trays cost
~20% of packing)
© 2012 G.P. Towler / UOP. For educational use in conjunction with
Towler & Sinnott Chemical Engineering Design only. Do not copy
Chemical Engineering Design
Packed Column Design
• Simulation programs have correlations for pressure drop
and HETP for some of the older and more common types of
random packing
– Column diameter is set by operating at typically 80% of flooding limit
– Column height is set by HETP and number of stages needed
• Some vendors have provided approximate correlations of
structured packing performance to simulation companies
• For conceptual design, costs can be estimated in Aspen
ICARUS
• Vendors should always be contacted to confirm packing
performance
© 2012 G.P. Towler / UOP. For educational use in conjunction with
Towler & Sinnott Chemical Engineering Design only. Do not copy
Chemical Engineering Design
Column Internals Exercise
• What column internals would you specify for the following
applications?
• Gasoline debutanizer
• Sieve trays
• Ethanol-water (azeotropic distillation)
• Sieve trays
• Acid gas scrubbing using alkanolamine
(corrosive, high liquid loading)
• Packing (or sieve
trays)
• Styrene monomer re-run column
(polymerizes easily)
• Packing or sieve
trays
• Fermentation broth stripper (high fouling)
• Sieve trays
• VOC recovery fractionation (high turndown) • Bubble caps
© 2012 G.P. Towler / UOP. For educational use in conjunction with
Towler & Sinnott Chemical Engineering Design only. Do not copy
Chemical Engineering Design
Separation Columns
• Distillation Basics
• Absorption & Stripping
• Column Internals & Hydraulics: Trays
• Column Internals & Hydraulics: Packing
• Column Design & Optimization
• Complex Columns
© 2012 G.P. Towler / UOP. For educational use in conjunction with
Towler & Sinnott Chemical Engineering Design only. Do not copy
Chemical Engineering Design
Column Design Procedure
• Define feed
• Define product specifications
• Set column pressure
• Design and select internals
• Optimize column design
© 2012 G.P. Towler / UOP. For educational use in conjunction with
Towler & Sinnott Chemical Engineering Design only. Do not copy
Chemical Engineering Design
Column Pressure Considerations
Higher pressure
Lower pressure
•
•
Fewer trays, bigger diameter
•
Try to avoid dropping condenser
temperature to point where
(expensive) refrigeration is needed
•
Try to maintain total condenser to
obviate net gas compression and
minimize vent gas loss
•
Try to avoid vacuum operation for
safety, quality and reliability reasons
Lower relative volatility so need more
trays
•
Higher vapor density, so reduced
column diameter
•
Higher boiling and dew points, so
hotter reboiler and condenser
–
–
•
May mean that reboiler utility becomes
more expensive
Try to avoid needing a fired heater if
possible
In some cases, high reboiler
temperatures can cause product
degradation
–
–
Monomers can polymerize
Foods and flavors can degrade
© 2012 G.P. Towler / UOP. For educational use in conjunction with
Towler & Sinnott Chemical Engineering Design only. Do not copy
Usually set column pressure
to allow condenser to run on
cooling water or air cooling
Chemical Engineering Design
Column Optimization
• It is not trivial to formulate and solve distillation column
optimization problems
– Discrete variables: number of trays, feed tray
– Continuous variables: pressure, reflux rate
– Many options for complex columns
– Column sequencing
• Operational constraints and design margins may make
the mathematically optimal design impractical
• Most designs are “close to optimal” rather than rigorously
optimized
© 2012 G.P. Towler / UOP. For educational use in conjunction with
Towler & Sinnott Chemical Engineering Design only. Do not copy
Chemical Engineering Design
Column Optimization
• How do you expect the following parameters to vary with
reflux ratio?
Diameter
Reflux
V = (R+1)D
D = distillate
Reflux
© 2012 G.P. Towler / UOP. For educational use in conjunction with
Towler & Sinnott Chemical Engineering Design only. Do not copy
Reflux
Total Costs
Condenser Duty
Reflux
Reboiler Duty
d2  R
Nmin
Height
Number of trays
Rmin
Reflux
Optimum
Reflux
Chemical Engineering Design
Total Costs
Column Optimization
Optimum
Reflux
• Calculated optimum reflux ratio is usually between 1.05 and
1.10 times Rmin
• Most designers choose to use a somewhat larger reflux
ratio to allow some design margin: typically 1.15 R min
• Reflux and pressure are main optimization parameters,
number of trays and feed tray usually have lower impact on
cost
© 2012 G.P. Towler / UOP. For educational use in conjunction with
Towler & Sinnott Chemical Engineering Design only. Do not copy
Chemical Engineering Design
Separation Columns
• Distillation Basics
• Absorption & Stripping
• Column Internals & Hydraulics: Trays
• Column Internals & Hydraulics: Packing
• Column Design & Optimization
• Complex Columns
© 2012 G.P. Towler / UOP. For educational use in conjunction with
Towler & Sinnott Chemical Engineering Design only. Do not copy
Chemical Engineering Design
Side Streams
Distillate
• Withdrawing a side stream allows a
component to be removed that has
accumulated in the middle of the
column
• OK if low purity is needed for
middle product and middle product
has low flow rate
Feed
Side
product
• Often used to remove small
amounts of intermediate boilers –
cheaper than adding a column if
product loss can be tolerated or
stream is recycled
• Side stream can be above or below
feed
Bottoms
© 2012 G.P. Towler / UOP. For educational use in conjunction with
Towler & Sinnott Chemical Engineering Design only. Do not copy
Chemical Engineering Design
Side Streams
D
F
Side
product
Side product
composition
B
• Side stream purity is usually low and some amount of
desired product is usually lost
© 2012 G.P. Towler / UOP. For educational use in conjunction with
Towler & Sinnott Chemical Engineering Design only. Do not copy
Chemical Engineering Design
Side Strippers and Rectifiers
• Use of side strippers and rectifiers
allows higher purity in a side stream
Distillate •
Side
stripper
For intermediates withdrawn above the
feed tray use a side stripper to strip out
light components
• For intermediates withdrawn below the
feed tray, use a side rectifier to knock
down heavy components:
Feed
Bottoms
© 2012 G.P. Towler / UOP. For educational use in conjunction with
Towler & Sinnott Chemical Engineering Design only. Do not copy
• Side strippers are widely used in
petroleum refining
Chemical Engineering Design
Intermediate Reboiling or Condensing
Distillate
• A side reboiler allows part of the heat
requirement to be added at a lower
temperature using cheaper hot utility
(or less area)
• A side condenser allows part of the
heat to be removed at a higher
temperature using cheaper cold utility
(or less area)
Feed
Qr2
Side
reboiler
Qr1
Bottoms
© 2012 G.P. Towler / UOP. For educational use in conjunction with
Towler & Sinnott Chemical Engineering Design only. Do not copy
• These options are considered for
columns with high heat duties,
refrigeration, or wide temperature
difference between reboiler and
condenser
• Another way to add or remove heat at
intermediate points is with a liquid
“pumparound” circuit that can be heat
exchanged with other process streams
Chemical Engineering Design
Dividing Wall Distillation
A
A
B
C
B
C
A
A
B
C
B
C
• Vertical wall separates column sections
• Three products can be made using a single shell (middle product purity
is much better than simple side draw)
• Up to 30% savings in capital and energy costs
• BASF have built 30 since 1984, BP 1, SASOL 2
© 2012 G.P. Towler / UOP. For educational use in conjunction with
Towler & Sinnott Chemical Engineering Design only. Do not copy
Chemical Engineering Design
UOP PEP Splitter Dividing Wall Column
Mechanical Pictures
Source: UOP
© 2012 G.P. Towler / UOP. For educational use in conjunction with
Towler & Sinnott Chemical Engineering Design only. Do not copy
Chemical Engineering Design
Complex Column Simulation
• Most simulators allow specification of side streams, side
columns and intermediate exchangers
• Adding an additional stream or exchanger adds a degree
of freedom, so one more column specification is required
– Usually side-stream draw rate or exchanger duty
• DWC is usually not a standard option in the simulation
program, but can be built up from component sections
© 2012 G.P. Towler / UOP. For educational use in conjunction with
Towler & Sinnott Chemical Engineering Design only. Do not copy
Chemical Engineering Design
High Capacity Trays
• Closed downcomers
• Maximize active area
• Directional valves
• Enhanced phase separation
• Reduced tray spacing
UOP MD Tray: Source UOP
• Examples:
– UOP: MD , ECMD , PFMD , SimulFlow
– Shell: HiFi ,ConSep
– Koch-Glitsch: SuperFrac®, Ultra-Frac®
– Sulzer: VGPlus
© 2012 G.P. Towler / UOP. For educational use in conjunction with
Towler & Sinnott Chemical Engineering Design only. Do not copy
Chemical Engineering Design
High Capacity Trays
• Increased active area or enhanced phase separation
enable much higher throughput
– E.g.: UOP SimulFlow
has 230% capacity of sieve trays
• Efficiency is usually similar to sieve or valve trays, but
some designs have enhanced efficiency as well as
capacity
• Enhanced capacity trays are used a lot in revamps to save
buying a new shell, also for very large diameter new
columns
• Details must be obtained from vendors
© 2012 G.P. Towler / UOP. For educational use in conjunction with
Towler & Sinnott Chemical Engineering Design only. Do not copy
Chemical Engineering Design
Questions ?
© 2012 G.P. Towler / UOP. For educational use in conjunction with
Towler & Sinnott Chemical Engineering Design only. Do not copy
Chemical Engineering Design
Heat Transfer Equipment
Heat Exchangers
Chemical Engineering Design
Heat Exchangers
• Heat Transfer Basics
• Tubular Exchangers
• Heat Exchanger Design
• Compact Heat Exchangers
Chemical Engineering Design
Three Mechanisms of Heat Transfer
• Conduction
Q = A k dT
dx
• Affects wall resistances, which are usually negligible for heat transfer equipment
• Convection
Q = U A T
• Usually the governing mechanism in most process applications
• Radiation
Q = A σ ε (ΔT4)
• Important in fired heaters
Q = heat duty
A = area
T = absolute temperature
k = thermal conductivity
U = heat transfer coefficient
σ = Stefan Boltzman const.
ε = transmission factor
Chemical Engineering Design
Convective Heat Transfer in a Tube
Dittus-Boelter Equation:
h i di
Nu =
k
(for inside h.t.c. hi based on
inside diameter di)
Collect variables:
hi
=
1/3
 Cp μ 

= 0.023 
k


 ρ v di 


μ


0.8
k 2/3
1/3
0.8
0.023 0.2 C p μ −0.467 (v ρ )
di
So if we increase
then hi will:
Fluid thermal conductivity, k
Fluid heat capacity, Cp
Fluid density, ρ
Velocity, v
Fluid viscosity, μ
Pipe diameter, di
increase
increase
increase
increase
decrease
decrease
Chemical Engineering Design
Combined Conduction and
Convection
L
• For a flat plate, overall resistance is the
sum of the individual resistances
h1
h2
hot
cold
• Hence overall heat transfer coefficient, U is
given by
k
T
Q 1 L 1 
 + +  = Th − Tc
A  h1 k h 2 
Th
Tc
Q
= U A (Th − Tc )
1
U
=
1 L 1
+ +
h1 k h 2
x
Chemical Engineering Design
Cylindrical Geometry (Tubes)
• By convention, U is based on outside diameter
= U A (Th − Tc )
Q
ro
ri
1
ro U
=
ln (ro / ri )
1
1
+
+
ro h o
k
ri h i
• Add terms for fouling:
1
ro U
=
ln (ro / ri ) 1
1
1
1
+
+
+
+
ro h o ro f o
k
ri f i ri h i
Outside h.t.c. ho depends strongly on equipment type: see Chapter 19
for correlations
Chemical Engineering Design
Counter Current Heat Transfer
Tc, in
Cold End
Hot End
ΔTh = Th,in – Tc,out
Th, in
Th, out
ΔTc = Th,out – Tc,in
Tc, out
• Q = U A ΔTm
• For perfect counter current flow, ΔTm is the log mean
temperature difference:
Tlm
• Easiest to remember as:
Tlm
(
(
T
=
h,in
=
− Tc,out ) − (Th,out − Tc,in ))
 (Th,in − Tc,out ) 

ln 

 (Th,out − Tc,in ) 
(ΔTh − ΔTc )
 ΔT 
ln  h 
 ΔTc 
Chemical Engineering Design
Counter Current Heat Transfer
Tc, in
Cold End
Hot End
ΔTh = Th,in – Tc,out
Th, in
ΔTc = Th,out – Tc,in
Th, out
Tc, out
Tlm
=
(ΔTh − ΔTc )
 ΔT 
ln  h 
 ΔTc 
A useful shortcut to know:
Th/Tc
1.0
1.5
2.0
3.0
4.0
Tlm / Tc
1.0
1.2
1.4
1.8
2.2
Tgeom mean = √(Tc x Th)
Tgeom mean /Tc
1.0
1.22
1.41
1.73
2.0
So Tlm  Tgeom mean if ratio is 3 or less!
Chemical Engineering Design
Counter-current Flow in Real Exchangers
• Most real heat exchangers do not have pure countercurrent flow
• We apply a correction factor for this in design
Q = U A F ΔTlm
• F is usually > 0.8 unless the design is poor (see later)
• F is sometimes called Ft
hot out
hot in
cold out
cold in
Chemical Engineering Design
Heat Exchangers
• Heat Transfer Basics
• Tubular Exchangers
• Heat Exchanger Design
• Compact Heat Exchangers
Chemical Engineering Design
Shell and Tube Heat Exchangers
How do we turn this -
Into this -
Chemical Engineering Design
Shell and Tube Heat Exchangers
Source: Perry’s Chemical Engineers Handbook, McGraw-Hill
Chemical Engineering Design
Shell and Tube Exchangers
Source: Riggins Company:
www.rigginscompany.com
Chemical Engineering Design
S&T Exchanger Construction
Baffle assembly
Inserting
tubes
Welding the shell
Tubesheet
Final
product
Source: Bos-Hatten Inc.: www.Bos-Hatten.com
Chemical Engineering Design
Tube Bundles
U-tubes
Tubesheet
Baffles
Source: UOP
Chemical Engineering Design
Heat Exchanger Design
• Heat exchange design must:
– Provide required area
– Contain process pressure
– Prevent leaks from shell to tubes or tubes to shell
– Allow for thermal expansion
– Allow for cleaning if fouling occurs
– Allow for phase change (some cases)
– Have reasonable pressure drop
• S&T heat exchangers are built to standards set by the
Thermal Exchanger Manufacturers Association (TEMA)
Chemical Engineering Design
TEMA Nomenclature: Front Heads
• A Type
Easy to open for tubeside access
• Extra tube side joint
•
A
B
•
B Type
Must break piping connections to open
exchanger
• Single tube side joint
•
C
N
• C Type
•
•
Channel to tubesheet joint eliminated
Bundle integral with front head
• N Type
D
•
Fixed tubesheet with removable cover
plate
• D Type
•
Special closures for high pressure
applications
Chemical Engineering Design
TEMA Nomenclature: Shells
E
• E Type
•
F
• F Type
•
G
H
Most common configuration without phase change
Counter current flow obtained. Baffle leakage problems.
• G Type
•
Lower pressure drop
• H Type
•
Horizontal thermosyphon reboilers
• J Type
J
K
•
Older reboiler designs
• K Type
•
Phase separation integral to exchanger
• X Type
X
•
Lowest pressure drop, low F factor
Chemical Engineering Design
TEMA Nomenclature: Rear Heads
• L Type
•
•
L
•
N
P
M Type
•
M
Same as A type front head
Same as B type front head
N Type
•
Same as N type front head
• P & W Types
S
T
•
Rarely used
• S Type
•
U
W
Floating head with backing ring
• T Type
•
Floating head pulls through shell
• U Type
•
Removable bundle without floating head
Chemical Engineering Design
Selection of Exchanger Type: Examples
1) Feed preheater
2) Crude preheat train
Low pressure
Low pressure
Tubeside - Steam
Tubeside – Vacuum Residue
Shellside – Naphtha
Shellside – Crude oil
BEU
3) Reboiler
AES or AET
4) Sterilizer Preheat
Medium pressure
Low pressure
Tubeside - Steam
Tubeside - Milk
Shellside - Kerosene
Shellside - Steam
BKU or BHM
AET
Chemical Engineering Design
LMTD Correction Factors
T1
t2
t1
T2
R = (T1 - T2)
(t2 – t1)
Source: Perry’s Chemical Engineers Handbook, McGraw-Hill
E Shell - 1 Shell Pass: Similar correlations exist for other shell arrangements
Chemical Engineering Design
Temperature Cross
• When Th, out < Tc, out we have a
“temperature cross”
T
• Temperature cross causes problems
if exchanger is not counter-current
and gives low F factors
• If Tc,out – Th,out > 5% of Lmtd then F <
0.8 and it is usually best to split the
exchanger into multiple shells in
series
Duty, or Enthalpy, H
T
• Number of shells can be estimated
by stepping off on T-H diagram
Note:
H
• Real exchangers can have non-constant Cp
• Size & duty of HX in series is not necessarily the same
• Large number of shells in series approximates pure counter-current exchange
Chemical Engineering Design
Temperature Cross in Simulation
• Most simulators
show an error if
there is a low F
factor
• For example, in
UniSim Design
the exchanger
shows up
yellow
•
Details of the example are in Ch 4
Chemical Engineering Design
Temperature Cross in Simulation
• Opening the exchanger shows the low F factor
Chemical Engineering Design
Temperature Cross in Simulation
• Can use the “plots” tab to plot temperature against heat
flow and visualize the temperature cross
Chemical Engineering Design
Temperature Cross in Simulation
3
2
1
• Stepping off between profiles suggests we need three
exchangers and gives target inlet and outlet temperatures
Chemical Engineering Design
New design with temperature cross
eliminated
Chemical Engineering Design
Profiles for the New Exchangers
(b) E101
(a) E100
(c) E102
Chemical Engineering Design
Tube Pitch
Pitch
30
60
90
45
• Triangular or square pitch, each with two orientations
• TEMA minimum pitch is 1.25 x tube outside diameter
• Sometimes use larger pitch for easier cleaning (but bigger shell,
lower shellside h.t.c.)
FEATURE
USE PATTERN:
lower P on shellside
square (effective only at low Re number)
shellside fouling
square - easier cleaning
horizontal shellside
boiling
square - prevent vapor blanketing
decrease shell size
fit 15% more tubes if triangular pitch used
Chemical Engineering Design
Baffle Types & Shell Flow Patterns
max unsupported
tube span
FULL
TUBEFIELD
max unsupported
tube span
FULL
TUBEFIELD
max unsupported
tube span
empty space
empty space
PARTIAL
TUBEFIELD
empty space
empty space
Chemical Engineering Design
Exercise: Selection of Sides
Process Fluid
Side Selection
Reason
Fouling fluid
Tube
Easier to clean
Viscous fluid
Shell
Lower Δp
Suspended solids
Tube
No dead spots for settling
Highest T
Tube
Cheaper, mechanically stronger
Highest pressure
Tube
Cheaper, mechanically stronger
Cooling water
Tube
Easier to clean
Corrosive fluid
Tube
Cheaper, easier to replace tubes
Much larger flow
Shell
Lower Δp
Condensing fluid
Shell
Drains better
Chemical Engineering Design
Heat Exchangers
• Heat Transfer Basics
• Tubular Exchangers
• Heat Exchanger Design
• Compact Heat Exchangers
Chemical Engineering Design
Heat Exchanger Design
1. Determine duty, check for temp cross
Q = (m Cp ΔT) + (δm ΔHvap)
2. Estimate U and hence calculate area
Q = U A F ΔTlm
3. Determine exchanger type and tube layout
4. Pick d, L and calculate number of tubes, hence shell diameter
5. Calculate hi, ho and confirm U. Return to 2 if needed.
6. Calculate Δp. Return to 3 if needed.
Chemical Engineering Design
Approximate Heat Transfer Coefficients
Fluid
More examples (in metric units) in Chapter 19
h (Btu/(hr.ft2.F))
Shell-side
Tube-side
Liquids
Water solutions, 50% water or more
Alcohols, organic solvents
Light Hydrocarbons (naphtha, gasoline)
Medium Hydrocarbons (kerosene, diesel)
Heavy oils (gas oils, crude oil)
300
200
190
130
30
300
200
190
120
20
Vapors
Air, 10 psig
Hydrogen, 50 psig
Hydrogen, 500 psig
Hydrocarbon vapor, 50 psig
Noncondensible gas, 2 psig
10
100
400
60
5
10
100
400
60
5
Note: Coefficients are based on 3/4 inch diameter tubes. For Tube side flows, correct by
multiplying by 0.75/Actual OD. Estimated accuracy is 25%. For 50% hydrogen in
vapor, reduce h to 2/3 of pure H 2 value.
Chemical Engineering Design
Approximate Fouling Factors
Fluid
f (Btu/(hr.ft2.F))
River water
Sea water
Cooling tower water
Town water (soft)
Town water (hard)
Flue gas
Steam
Steam condensate
Light & medium hydrocarbons
Heavy oils
Boiling organics
Aqueous salt solutions
Fermentation broths
600
500
600
700
300
800
1000
1000
800
200
400
600
300
Chemical Engineering Design
Hydraulics & Pressure Drop
• Heat exchanger design is a trade off between better heat
transfer (high velocity, low diameter) and pressure drop
• In early stages of design, we usually allow for a “typical”
pressure drop:
• 5 psi shell-side
• 10 psi tube-side
– But we have to calculate Δp rigorously where it is critical to
performance, e.g. thermosyphon reboilers
• In detailed design, use correlations or simulation programs
to more rigorously optimize if pressure drop is important to
process performance – see Chapter 19 for examples
Chemical Engineering Design
Example: S&T HX Design
• What size of exchanger is needed to heat 375,000 lb/h of naphtha from
150F to 300F using medium pressure steam at 360F?
•
•
•
•
•
•
•
•
•
•
Q = m.Cp.ΔT = 375 x 103 x 0.5 x 150 = 28.125 MMBtu/h
Put steam shell-side, oil tube-side, TEMA BEU
Estimate hi = 190 Btu/(h.ft2.F), ho = 1500 Btu/(h.ft2.F)
Estimate fi = 400 Btu/(h.ft2.F), fo = 1000 Btu/(h.ft2.F)
Estimate 1/U = 1/190 + 1/400 + 1/1000 + 1/1500
U = 106 Btu/(h.ft 2.F)
R ≈ 0 (negligible ΔT on shell-side) so F = 1.0
Lmtd = (60 – 210)/ln(60/210) = 119.7°F
Hence A = Q/U.F.Lmtd = 28.1 x 106 / (106 x 1.0 x 119.7) = 2215 ft2
For 20’ long 3/4” tubes area/tube = 3.9 ft 2, hence need 564 tubes
From table in Perry’s Hbk, 1” triangular pitch TEMA U, nearest shell size is 29”
i.d., with 648 tube count.
• We can now use correlations to confirm design and check hydraulics:
see Ch 19 for details of using commercial design software
Chemical Engineering Design
Heat Exchangers
• Heat Transfer Basics
• Tubular Exchangers
• Heat Exchanger Design
• Compact Heat Exchangers
Chemical Engineering Design
Hairpin Exchangers
•
When small duties are required, hairpin exchangers are specified:
–
–
–
–
cheaper than very small shell and tube
highly effective (single pass, true countercurrent)
75 → 1500 ft2 surface area
4 → 16" shell diameter, 20 ft long
U
tubes
split flange/
split ring
arrangement
tube
sheets
removable
bonnet
This design is used for double-pipe and multi-tube exchangers.
Chemical Engineering Design
Plate & Frame Exchangers
Plates
Source: Alfa-Laval, www.AlfaLaval.com
Gasket
Chemical Engineering Design
Gasket Layout of Alternating Plates
Chemical Engineering Design
Plate & Frame Exchangers
• Advantages
• Close to counter-current heat transfer, so high F factor
allows temperature cross and close temperature
approach
• Easy to add area
• Compact size
• Relatively inexpensive for high alloy
• Can be designed for quick cleaning in place
• Disadvantages
• Lots of gaskets
• Lower design pressure, temperature
• External leakage if gaskets fail
• Applications
• Food processing, brewing, biochemicals, etc.
• Design method: see Chapter 19
Source: Alfa-Laval,
Alfa-Laval www.AlfaLaval.com
Chemical Engineering Design
Plate & Frame Exchangers
Source: Alfa-Laval,
Alfa-Laval www.AlfaLaval.com
Chemical Engineering Design
Welded Plate Heat Exchangers
Feed Outlet
Effluent Inlet
Vent
•
–
Hot End Bellows
Source: Alfa-Laval Packinox
Advantages
–
Manhole
Feed Header
Effluent Header
Bundle Support
–
Pressure Vessel
–
Welded
Plate Bundle
–
–
•
Disadvantages
–
–
Bracket
(or Skirt)
–
Effluent Header
Venturi
Spray Bar
Liquid Feed
Cold End Bellows
Recycled Gas Inlet
Effluent Outlet
•
Higher thermal efficiency
Single unit can replace multiple shell &
tube units
Closer approach to hot inlet temperature
Low pressure drop
Little chance of vibration problems
Excellent distribution of two phase flows
Single alloy material for plates
Difficult to clean
Few manufacturers at large scale (Alfa
Laval Packinox)
Used in large scale clean services that
need close temperature approach
Chemical Engineering Design