Use of Advanced Educational Technologies in a

Process Simulation Course

Mordechai Shacham
Department of Chemical Engineering.
Ben-Gurion University of the Negev
Beer-Sheva, Israel
 Typical Process Simulation Course Characteristics

Commercial simulators (such as HYSIS, AspenPlus and PRO

II) are used to model the steady state or dynamic operations of
processes.
Benefits: The process simulator "provides a time-efficient and
effective way for students to examine cause-effect
relationships“* among various parameters of the process
Pedagogical drawbacks: "it is possible for students to
successfully construct and use models without really
understanding the physical phenomena within each unit
operation“*
"the majority of students see simulations merely as
sophisticated calculators that save time.Ӡ

*Dahm et al., CEE, 36, 192 (2002) †Rockstraw, CEE, 39, 68 (2005)
 Course Objectives

At the completion of this course the students should be able to:

1. Convert the mathematical models of the unit operations into
"unit modules" with consistent sets of input and output
variables.
2. Utilize up to date physical property databases for retrieving
property data and equations to be included in the unit modules;
3. Identify optimal computational sequence in a process flow-sheet
using various partitioning and tearing algorithm;
4. Identify the most efficient recycle convergence algorithm;
5. Identify suitable integration algorithms for various stages of a
dynamic process;
6. Use a commercial simulator to verify simulation results.
 New Techniques and Tools Utilized

1. An interface to the DIPPR physical property database, which

enables transferring the data and the correlations directly into a
computer code.
2. A new technique for grouping the equations and data
according to their role in the model instead of computational
sequencing requirements.
3. Automatic conversion of Polymath equation sets into MATLAB
programs.
4. Using the most recent tools available for numerical solution of
problems.
5. Self-recording of the lectures using a Tablet PC
 Semi-Batch Distillation of a Binary Organic Mixture* - An
Example

Steam at 99.2 ºC
2.31 mol/min flow-rate

10.991 mol n – octane

4.169 mol n - decane

738 Torr, ambient press.

Prenosil, J. E., Chem. Eng. J. 12, 59-68 (1976)

 Problem Statement

Initially M = 0.015 kmol of organics with composition x1

= 0.725 is charged into the still. The initial temperature in the
still is T0 = 25 °C. Starting at time t = 0 steam of temperature
Tsteam = 99.2 °C is bubbled continuously through the organic
phase at the rate of MS = 3.85e-5 kmol/s. All the steam is
assumed to condense during the heating period. The ambient
temperature is TE = 25 °C and the heat transfer coefficient
between the still and the surrounding is U = 1.05 J/s-K. The
ambient pressure is P = 9.839E+04 Pa.
Assumptions: 1) Ideal behavior of all components in
pure state or mixture; 2) Complete immiscibility of the water
and the organic phases; 3) Ideal mixing in the boiler and 4)
Equilibrium between the organic vapor and its liquid at all
times. For standard state for enthalpy calculations pure liquids
at 0 °C and 1 atm. can be used.
 Problem Statement

a) Calculate and plot the still temperature (T), component mole

fractions inside the still (x1,x2, y1 and y2) and the component
mole fractions in the distillate(x1dist and x2dist), as function of
time, using the data and the initial values provided.
b) The requirement is for the distillate to contain 90% of n-
octane. Determine the lowest n-octane mole fraction in the feed
that can yield the required distillate concentration. Compute
the percent recovery of n-octane in the distillate as function
of its concentration in the feed. Vary the feed concentration in
the range where the requirement for the n-octane concentration
in the distillate is attainable.
Semi-Batch Distillation - Heating Period Equations
Mass balance on the water phase yields
Water mass balance

Enthalpy balance

Phase equilibrium

Bubble point equation

Semi-Batch Distillation - Distillation Period Equations

Water mass balance

Organics mass balance

Vapor flow rate from

enthalpy balance

Temperature in still must follow the bubble point curve

 Semi-Batch Distillation - Physical Property Needs and
Sources for the Original Reference*

*Prenosil, J. E., Chem. Eng. J. 12, 59-68 (1976)

Polymath Interface to the DIPPR Database – Selection
of the Pertinent Compounds
Polymath Interface to the DIPPR Database – Property
Data Reports

Range of applicability
Steam Distillation –Heating Period - Polymath Model
Model Equations
dT WS  H S  H Lw   Q dmW
  WS
dt mw c pLw  m( x1c pL1  x2 c pL 2 ) dt
Q  UA(TE  T ) x1 P1 x2 P2
y1  y2 
P P
f (T )  1   y1  y 2  yW   0

Pure Component Properties

Mixture Properties

Problem Specific Data,

Initial and Final Values

Grouping Equations and Data According to their Role

 Steam Distillation –Heating Period - Results

Initial Minimal Maximal Final

Variable value value value value
16 t 0 0 181.72 181.72
19 Temp 25 25 90.65594 90.65594
2 fT 0.953485 2.64E-05 0.953485 2.64E-05
13 MW 0 0 0.006996 0.006996
23 VP_C10H22 181.0043 181.0043 6461.373 6461.373
24 VP_C8H18 1870.976 1870.976 3.42E+04 3.42E+04
25 VP_H2O 3170.386 3170.386 7.19E+04 7.19E+04
26 x1 0.725 0.725 0.725 0.725
27 x2 0.275 0.275 0.275 0.275
28 Y1 0.013787 0.013787 0.251608 0.251608
29 Y2 0.000506 0.000506 0.01806 0.01806
30 YW 0.032223 0.032223 0.730306 0.730306

Final time obtained by trial and error

 Steam Distillation – Distillation Period - Polymath Model
No.
1
Equation/ # Comment
# Distillation period model equations
Model Equations
2 d(MW)/d(t) = MS - V * YW # Eq. 6. Mass of water in the still (kmol) from mass balance dmW d  mx1  d  mx2 
 WS  Vy W  Vy 1  Vy 2
3 d(Mx1)/d(t) = -V * Y1 # Eq. 7. Mass of n-octane in the still (kmol) from mass balance dt dt dt
4 d(Mx2)/d(t) = -V * Y2 # Eq. 7. Mass of n-decane in the still (kmol) from mass balance WS  H S  H Lw   Q
5 V 
V =(MS*(HS-HL_H2O)+ Q) / (HV - (HL_H2O * YW + (Y1 * HL_C8H18 + Y2 * HL_C10H22)))
# Eq. 8. Vapor flow rate (kmol/s)
 
HV  y w hLw  ( y1hL1  y 2 hL 2 )
6 d(Temp)/d(t) = 1000 * eps # Eq. 9. Still temperature by controlled integration
7 M = Mx1 + Mx2 # Organic mass in the still (kmol) Controlled Integration
8 x1 = Mx1 / M # n-octane organic liquid mole fraction
dT
9 x2 = Mx2 / M # n-decane organic liquid mole fraction K c
10
11
Q = U* (Ta - Temp) # Eq. 3. Heat removed from the still (J/s)
Y1= VP_C8H18* x1 / P # Eq. 4. n-octane vapor mole fraction
dt
12 Y2 = VP_C10H22* x2 / P # Eq. 4. n-decane vapor mole fraction
13 YW =VP_H2O / P # Water vapor mole fraction
14 eps = 1 - (Y1 + Y2 + YW) # Eq. 5A. Error used in controlled integration 
  1  y1  y2  yW 
15 M1dist = M0 * x01 - Mx1 # Mass of n-octane in the distillate (kmol)
16 M2dist = M0 * x02 - Mx2 # Mass of n-decane in the distillate (kmol)
17 MWdist = MS * t - MW # Mass of water in the distillate (kmol)
18 Mdist = M1dist + M2dist # Distilled organic phase (kmol)
19 x1dist = If (Mdist > 0) Then (M1dist / Mdist) Else (0) # n-octane distillate mole fraction
20 x2dist = If (Mdist > 0) Then (M2dist / Mdist) Else (0) # n-decane distillate mole fraction
21 TK = Temp + 273.15 # Absolute temperature (K)
22 #
23
-
#Pure compound property equations
lines 13 - 22 in Table 3
Pure Component Properties
24 HV1 = (135540*TK + 443100 *1635.6* (coth(1635.6 / TK)) - 305400 *746.4* (tanh(746.4 / TK))-
4.928E+08) # n-octane vapor enthalpy (J/kmol)
25 HV2= (167200*TK+ 535300 * 1614.1*(coth(1614.1 / TK)) - 378200 *742* (tanh(742 / TK)) -
5.791E+08) # n-decane vapor enthalpy (J/kmol)
26 HVW = (33363*TK + 26790 *2610.5* (coth(2610.5 / TK)) ^ 2 + 8896 * 1169*(tanh(1169 / TK)) -
4.471E+07) # Water vapor enthalpy (J/kmol)
27 #
28 #Mixture property equations Mixture Properties
- lines 25 - 26 in Table 3
29 HV = YW * HVW + Y1 * HV1 + Y2 * HV2 # Vapor phase enthalpy (J/kmol)
30 #
 Steam Distillation – Distillation Period - Results
Initial Minimal Maximal Final
Variable value value value value
25 t 181.7 181.7 2000 2000
21 Mx1 0.010875 2.59E-05 0.010875 2.59E-05
22 Mx2 0.004125 0.001271 0.004125 0.001271
19 MW 0.006995 0.006995 0.022793 0.022793
28 Temp 90.64995 90.64995 96.56251 96.56251
1 eps 0.000249 1.01E-06 0.000249 1.73E-06
38 x1 0.725 0.019947 0.725 0.019947
40 x2 0.275 0.275 0.980053 0.980053
42 Y1 0.251556 0.008462 0.251608 0.008462
43 Y2 0.018055 0.018055 0.082287 0.082287
44 YW 0.730141 0.730141 0.909249 0.909249
39 x1dist 0 0 0.933033 0.791744
41 x2dist 0 0 0.208256 0.208256
15 M1dist 0 0 0.010849 0.010849
16 M2dist 0 0 0.002854 0.002854
17 Mdist 0 0 0.013703 0.013703
DAE system solution
Stop the distillation when x1dist <= 0.9   1   y1  y2  yW 
Heating Period – Automatic Conversion of the Polymath
Model into a MATLAB Program

The model is converted into a MATLAB

function
Equations reordered into a computational
sequence
MATLAB syntax is used
Heating Period – Secant Method Iterations on the Final
Time when fT = 0

The matrix yd stores the values of the

differential variables for both periods of
the distillation
Distillation Period – Use of Fully Implicit Integration Algorithm
and Secant Method Iterations on tfinal until x1dist <=0.9

Use the end points of the heating period as

starting points of the distillation period

res(1,1) = 1 - (Y1 + Y2 + YW);

res(2,1) = - (V * Y1)-dMx1dt;
res(3,1)= - (V * Y2)-dMx2dt;
res(4,1) = MS-V*YW-dMWdt ;
 Temperature and Mole Fractions Variation during Batch
Distillation (x1 = 0.725)
100
90
Temperature (Deg. C)

80 Distillation Period
70
60
50
40
30
20
0 5 10 15 20 25 30
Time (min)
x1 x2 y1 y2

1
0.9
Heating Period 0.8
Mole fraction

0.7
0.6
0.5
0.4
0.3
Results are compared with 0.2
0.1
results of Aspen 0
0 5 10 15 20 25 30
Time (min)
 Percent Recovery of n-octane in Distillate (90 % Purity)
as Function of its Initial Mole Fraction in the Feed

100
90
Percent recovery of n-octane

80
70

60

50
40

30
20
10

0
0.6 0.65 0.7 0.75 0.8 0.85 0.9 0.95
n-octane mol fraction in feed
 Conclusions
1) The broad content and wide diversity of the course material
require that the students review, enhance, update and make
practical use of their knowledge of programming, material and
energy balances, thermodynamics, numerical methods and
reaction engineering.

2) Covering such a broad content is made possible by the

introduction of advanced instructional technologies and new,
time efficient problem solving techniques and tools

3) Optimal utilization of state of the art problem solving tools in

education requires continuous development of new problems
and modification of existing ones so that their level of
complexity, detail and the required precision follow the
advancement of the capabilities of the problem solving tools.