Está en la página 1de 41

Accepted Manuscript

Title: The teaching of enhanced distillation processes using a


commercial simulator and a project-based learning approach

Author: Lourdes Calvo Cristina Prieto

PII: S1749-7728(15)30010-5
DOI: http://dx.doi.org/doi:10.1016/j.ece.2016.07.004
Reference: ECE 126

To appear in: Education for Chemical Engineers

Received date: 16-6-2015


Revised date: 5-4-2016
Accepted date: 15-7-2016

Please cite this article as: Calvo, L., Prieto, C.,The teaching of enhanced distillation
processes using a commercial simulator and a project-based learning approach,
Education for Chemical Engineers (2016), http://dx.doi.org/10.1016/j.ece.2016.07.004

This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.
HIGHLIGHTS

When a multicomponent fluid mixture is non-ideal, its


separation is difficult.

Enhanced separation processes with several equipments are


5 required.

Chemical processes commercial simulators must be used to

t
ip
design these processes.

The teaching method and outcomes of using computer labs

cr
with Aspen Plus are shown.

10 A survey done to the students showed their satisfaction with

us
the method.

an
M
15 The teaching of enhanced distillation processes using a commercial
simulator and a project-based learning approach
d
te

Lourdes Calvo*lcalvo@ucm.es, Cristina Prieto


p
ce

20 Department of Chemical Engineering, School of Chemical Sciences, Universidad

Complutense de Madrid, Av. Complutense s/n. 28040, Madrid.


Ac

Page 1 of 40
Phone: +34 91 394 4185

25 Abstract

t
ip
When a multicomponent fluid mixture is non-ideal, its separation is difficult;

thus, enhanced separation processes with several equipments are required.

cr
Commercial chemical processes simulators must be used to design these difficult

separations, so ChE students have to learn them. This work shows the teaching

us
30 methodology and outcomes of using computer labs with Aspen Plus to solve

advanced distillation processes within a fourth year course of the ChE degree.
an
Lectures on theoretical aspects are followed by seminars and then by computer

labs. During lab computers, the professor initially solves a case, step by step;
M
while the students do it at the same time. Then, they are asked to solve an

35 industry relevant case using a project-based learning approach in order to force


d

them to continue working with the simulator on their own, and to enhance the
te

comprehension of the process. They have to deliver a report summarizing and

critically analyzing the results. This assignment counts 25% the course total
p

grade. A survey done to the students showed their satisfaction with the method.
ce

40 They were very favorable to the use of Aspen Plus as a tool for a better

understanding of the enhanced distillation processes. Moreover, they found the


Ac

Page 2 of 40
tutorials very useful to improve their knowledge on this simulator. All students

passed the course, 80% obtained good grades.

45 Keywords

t
ip
Enhanced distillation, Aspen Plus, project-based learning, tutorials, survey

cr
1. Introduction

us
50 Separation processes could be defined as the processes in which a

multicomponent mixture is transformed into two or more different products. In


an
case of fluid mixtures, the most common separation operation is distillation. In

ideal binary mixtures, a single distillation column may suffice, but in case of
M
multicomponent mixtures, more than one distillation column is required.

55 Moreover, when the multicomponent fluid mixture is non-ideal, i.e. relative


d

volatilities are too low or azeotropes are formed, its separation by a sequence of
te

ordinary distillation columns will not be technically and/or economically feasible

(Henley et al., 2011). For such mixtures, enhanced separation processes have
p

been developed, including extractive distillation, homogeneous azeotropic


ce

60 distillation, heterogeneous azeotropic distillation, pressure-swing distillation,

reactive distillation, supercritical extraction and membrane separations.


Ac

Page 3 of 40
The analysis, design, control, and optimization of these separations involve

phase equilibrium relationships and material and energy balances. In case of

non-ideal mixtures these calculations often fail because of liquid-solution non-

65 idealities and/or the difficulty of specifying feasible separations (Henley et al.,

2011). In addition, calculations become too complex, due to the need of solving

a set of non-linear algebraic equations and in most cases, the requirement of

t
ip
several separation equipments.

cr
Commercial chemical processes simulators include accurate thermodynamic

70 packages coupled with sufficiently rigorous computational algorithms which

us
enable engineers to solve rapidly these difficult separations. Currently, there are

several commercial chemical process simulations available, such as Aspen Plus,

use these simulators.


an
CHEMCAD, SuperPro Designer, PROSIM, etc. Engineers at industrial practice

Thus, their learning needs to be covered in chemical


M
75 engineering education. An effective manner to learn them is through computer

laboratories introduced in order of growing complexity as shown by Rockstraw,


d

2005 or with the aid of interactive guides as demonstrated by Castrelln et al.,


te

2009. Furthermore, these tools can be added to a lecture course and used to

solve realistic problems to minimize the time necessary to train the students, as
p

80 successfully experienced by Wankat, 2002 and 2006 in his own courses. The use
ce

of process simulation in ChE courses was analyzed by Dahm et al. 2002

concluding that computing is generally accepted as an integral component in


Ac

teaching design but it has not significantly permeated the rest of the curriculum.

Page 4 of 40
So these authors recommend the implementation of mini-modules thorough the

85 courses.

Azeotropic distillation has been generally studied in Chemical Engineering

degrees succinctly, qualitatively, explaining the basics, without addressing the

rigorous resolution. Moreover, there are few universities that have courses in

t
Advanced Operations in Graduate Studies.

ip
90 On the other hand, there are not many textbooks that rigorously explain

cr
enhanced distillation. The latest edition of the book Separation Process

us
Engineering by Professor Wankat has a full chapter (number eight) dedicated to

these operations (Wankat, 2012). The sizing of the columns is made with the

95
an
approximate McCabe-Thiele method. An annex provides an introduction to the

resolution with Aspen Plus. However, it is not enough for teachers and students

who are not knowledgeable in this simulator. Other textbook that explains in
M
detail the azeotropic distillation is the latest version of Henley et al., chapter

eleven.
d
te

This work summarizes our experience in the design and put into practice a

100 course in advanced separation processes that included computer lab practices
p

with a commercial process simulator. We chose Aspen Plus because the


ce

Department of Chemical Engineering of the Complutense University of Madrid

had the license; but other commercial simulators could also being used.
Ac

1. The course

Page 5 of 40
105 Advanced separation processes is an optional six-credit (ECTS) course for the

fourth year students of the degree of Chemical Engineering. It covers enhanced

distillation operations, supercritical extraction, membrane separations and

advanced drying operations. The course is divided into 30 hours of lectures, 15

hours of seminars, 12 hours of laboratories, 3 hours of tutorship, and 3 hours of

110 exams.

t
ip
The student learning is evaluated through the projects they have to deliver after

cr
the lab practices and through several short exercises/solved problems that are

collected after the seminars. In addition, a written test is done at the end of the

us
classes to primarily evaluate the theoretical content of the subject.

115 an
The final exam counts 60% of the grade. The seminar exercises count 15% and

the lab assignment, the remaining 25%. Lab attendance is obligatory. The

grading scheme is the typical for the Universities in Spain: SS (fail) = <5; AP=5-
M
7; NT=7-9; SB= 9-10. The complete syllabus of the course is available in

http://quimicas.ucm.es/data/cont/media/www/pag-10534/2014-
d

120 15/GIQ_Guia%20docente%20Ampliacion%20Operaciones%20de%20Separacio
te

n_2014_FINAL.pdf and from the author at lcalvo@ucm.es.


p

The students get to this course after completing a course in Applied


ce

Thermodynamics in the second year and after finalizing Unit Operations

which is an obligatory subject for the third year students. In the latter, they learn
Ac

125 the fundamentals of thermodynamics, mass transfer and equilibrium stages as

Page 6 of 40
well as the separations based on phase addition or creation: distillation, liquid-

liquid extraction and absorption using approximated and rigorous computer

aided methods. During the second semester, they study separations processes by

barriers and solid agents as well those involving a solid phase.

130 The course is divided into two parts, each one leaded by a different professor.

t
We are in charge of Enhanced Distillation and Supercritical Extraction. The

ip
objective of the present document is to describe the method and the outcomes of

cr
our experience in teaching this complex subject with the aid of computer labs

using Aspen Plus and a project-based learning approach.

us
135 1. Teaching method

Lectures on theoretical aspects are followed by seminars and then by computer


an
labs. A project-based learning approach is used in the computer labs. Real

separation cases are presented to the students to oblige them to integrate the
M
theoretical knowledge previously given and to improve skills such as critical

140 thinking and problem solving. Mills and Treagust, 2003 studied the application
d

of project-based learning (PBL) to engineering education using several


te

examples. They concluded that students who participate in PBL are generally

motivated by it and demonstrate better teamwork and communication skills.


p

They have a better understanding of the application of their knowledge in


ce

145 practice and the complexities of other issues involved in professional practice.

Consequently, these authors recommend the use of PBL as a key component of


Ac

Page 7 of 40
engineering programs since it is also appreciated by industry and accreditors

alike. Engineering schools that have programs of a predominantly project-

organised curricula in Europe are: Aalborg and Roskilde in Denmark; Bremen,

150 TU Berlin, Dortmund and Oldenburg in Germany, Delft and Wageningen in

Netherlands. More recently, PBL has been implemented in the curriculum at the

department of Biotechnology and Chemical Engineering of the Helsinki

t
ip
Metropolia University (Fortelius et al., 2015). A positive response from the

students in the form of augmentation of motivation and activity has been found.

cr
155 Team skills have improved as well as independent study capacity.

us
PBL may be also applied in individual courses. Tom Joyce, 2009 found

excellent results on the use of PBL on an Engineering Design course at


an
Newcastle University. Students commented positively on the creativity allowed

them and the sense of ownership of their learning which this engendered.
M
160 1.1. Description of the course

The first lectures are devoted to show how to track a distillation following
d

residue and distillation curves, and how to define feasible separations and
te

product composition regions (bow-tie regions) for azeotropic mixtures,

respecting the distillation boundaries in triangular graphs. Seminars are used to


p

165 show the calculations to obtain a portion of residue and distillation curves and
ce

how to plot approximate residue curve maps, including approximate distillation

boundaries using the classification of the singular points (pure components and
Ac

Page 8 of 40
azeotropes) as stable, unstable nodes and saddles. The first computer lab

sessions with Aspen Plus are dedicated to obtaining the equilibrium data that are

170 necessary to compute these calculations and to plot binary and ternary diagrams.

Following lectures show the fundamentals of the enhanced distillation operations

(extractive, salt, pressure-swing, homogeneous, heterogeneous and reactive

t
distillation), the criteria to select the most adequate type of distillation for a given

ip
mixture and the proper separation sequence. Opportunities and limitations for

cr
175 each operation are discussed using examples of commercial or feasible

processes. Multiplicity of solutions in steady-state operation is discussed.

us
During the seminars, real separations are solved by manual calculation, using the

an
approximate multicomponent methods of Fenske-Underwood-Gilliland to obtain

the reflux ratio and equilibrium stages of the distillation columns involved in

180 each process. Mass balances are also solved. These results are used as the initial
M
values to deduce the dimension of the columns with rigorous methods using

Aspen Plus during the computer labs. As basic references, we use chapter 11
d

dedicated to Enhanced Distillation of the book by Henley et al., 2011 and


te

chapter 7 entitled Synthesis of Separation Trains of the book by Seader et al.,

185 2004.
p
ce

Supercritical fluids and their properties as solvents for multicomponent fluid

mixtures separation are the objective of the following lectures. The


Ac

fundamentals of multicomponent mixtures fractionation using supercritical CO2,

Page 9 of 40
commercial applications, the installations and an analysis of the costs of these

190 processes are given. The height of the packed columns to perform such

separations is deduced using the short-cut method of Ponchon-Savarit in a

Jnecke diagram. Diameter of the columns is determined from specifically

developed flooding diagrams. The book on the technology (Brunner, 1994) and

the review of counter-current separations in supercritical conditions (Brunner,

t
ip
195 2009) are used as references. During seminars, the design and cost evaluation of

a specific multicomponent mixture fractionation using dense CO2 is fully solved

cr
based on the paper by lvarez et al., 2009.

us
1.1. Computer laboratory

200
an
When the students get to this course they have already received a course in

computer simulation (Process Control and Simulation), where they are taught the

basic functionalities of Aspen Plus to solve simple blocks, heat exchangers and
M
ideal distillation columns.
d

However, enhanced distillation systems usually require two or three columns


te

and/or other separation equipment (for example, an intermediate decanter or a

205 liquid-liquid extraction unit is needed in the case of homogeneous distillation).


p

Therefore, it is necessary not only to design each apparatus but to connect them
ce

and achieve convergence.


Ac

10

Page 10 of 40
In Reactive Distillation, it is also necessary to introduce the reaction zone into

the distillation column. Other important configuration factors are the feed entry

210 and the product-removal stages, the possible need for intercoolers and

interheaters when the heat of reaction is appreciable and obtaining required

residence time for the liquid phase, where reaction takes place.

t
Since the simulation of these systems is complex, during the computer labs, the

ip
professor initially solves a case, step by step, in a classroom equipped with

cr
215 computers for each student. The students follow the simulation and do it at the

same time. Results are then critically discussed in group. Besides, the students

us
are given written instructions in a recipe so they are able to repeat the whole

procedure on their own. They were developed for the Aspen Plus 8.6 version of

220
an
2014. Annex I shows an example of the tutorial to solve a reactive distillation

corresponding to the forth computer lab. To learn the basics of Aspen Plus we
M
recommend the students to read the Aspen Plus documentation and the book by

Schefflan, 2011. For the simulation of distillation columns we recommend the


d

book by Luyben, 2006 and the Appendixes on computer simulation of binary and
te

complex distillation systems of Wankat, 2012.

225 There are four sessions of computer labs with the presence of the professor and
p

the assistant professor. Each one lasts approximately two hours.


ce
Ac

11

Page 11 of 40
1.1.1. Comparison of thermodynamical models for the prediction of the
singular points of the mixture. Construction of binary diagrams as
230 function of composition and pressure. Construction of residue and
distillation curves maps in triangular diagrams.

During the first computer lab session, students learn on how to obtain and fit

t
ip
different equilibrium data. Due to the non-ideality of the liquid solutions of

235 azeotropic mixtures, it is crucial to choose the most appropriate

cr
thermodynamical model. In general, liquid activity-coefficient models as

UNIQUAC, NRTL or Wilson should be the first election. In case of doubts and

us
for other mixtures, the students are introduced on the use of the Property

Method Selection Assistant of Aspen Plus.

240
an
M
Using these models, azeotrope search and classification of the singular points is

done with the Find Azeotropes Tool. Comparison between the reported results
d

and the experimental data (when available) allow choosing the best model for the
te

mixture. Once it is fixed, the Ternary diagram window is subsequently used to

245 plot ternary diagrams with residual and distillation curves. Next, the students are
p

taught on how to build binary diagrams with the Binary Analysis tool.
ce

Emphasis on the azeotrope composition variation as a function of pressure is

given since this is the key for pressure swing distillation.


Ac

12

Page 12 of 40
250 If experimental equilibrium data are available, the students are trained on how to

fit them using the Data Regression system of Aspen Plus in order to customize

the thermodynamic model that will be further used for the simulation. At the end

of this session, the way of visualizing, copying and exporting the results is

t
shown.

ip
255

cr
us
1.1.1. Determination of dew and bubble points. Calculation of K-values

During the second session, students are trained on how to estimate bubble/dew
an
points of the mixtures as well as equilibrium constants (K-values) required for

260 approximate multicomponent methods that are used in the seminars. For that
M
purpose, the students are trained on how to define the variable in Property Sets

Folder/Properties tab, and then on how to fix the conditions for the examination
d

in Analysis Folder/Properties. Bubble points can be alternatively determined


te

with the Stream Analysis/Bubble and Dew Point tool.


p

265
ce

1.1.1. Simulation of an extractive distillation process and a pressure-


swing process to solve same separation.
Ac

13

Page 13 of 40
During the third session, students are trained to solve a separation of an

270 azeotropic mixture using an extractive and a pressure-swing distillation process.

We use a different system each year; some of them come from the published

work of Prof. Luyben as the acetone-methanol (Luyben, 2008) and acetone-

chloroform (Luyben, 2013), since optimized processes are reported, and

comparison can be made. Other nice example can be found in Muoz et al.,

t
ip
275 2006.

cr
us
For extractive distillation they have to compare several solvents. The best

solvent should not form azeotropes or create additional distillation boundaries

280 key components.


an
(based on the ternary diagram) and should increase the relative volatility of the

For pressure-swing distillation, they have to check if the

azeotrope composition significantly changes with pressure so the process could


M
be feasible. They also have to solve the mass balances to have accurate values

for the products flow rates and to use the approximate Fenske-Underwood
d

Gilliland method to obtain the initial values for the Aspen Plus simulation. The
te

285 students are warned that the DSTWU (Winn-Underwood-Gilliland) block of

Aspen Plus can alternatively be used to estimate the reflux ratio and number of
p

stages.
ce
Ac

14

Page 14 of 40
The specific objective of the computer lab session is to show the students how to

290 conduct a steady-state design of the required equipment: the columns for each set

up (two columns in each case) plus the auxiliary equipment, and to achieve

convergence of the whole installation. The distillation columns are dimensioned

using RADFRAC, in which it is necessary to define the following parameters:

number of plates and the reflux ratio (as deduced by the approximate methods),

t
ip
295 the entry of the feed and recycle (care should be taken since Aspen Plus calls the

condenser stage 1 and the reboiler stage N), the type of condenser (Total) and

cr
reboiler (Kettle), the type of calculation (Equilibrium) and convergence

us
(Azeotropic). It is also necessary to input an approximate value for the operating

pressure (from the algorithm of distillation column pressure shown in Henley et

300 an
al., 2011) and the distillate rate (from mass balances). Once this is done, the

students are trained on how to use the Design Specifications function to meet

the target composition of the products varying the reflux ratio and the distillate
M
rate and on how to use Sizing and Rating/Tray Sizing to find the diameter of

the columns. Sieve trays are usually chosen. The next step is to determine the
d

305 pressure drop in the column in the Design/Pdrop tab. When done, the students
te

are shown how to visualize the results such as the heat duties (Results) and

composition profiles (Profiles) in the Blocks folder. This later visualization


p

is very important to check that no pinch points are present.


ce
Ac

15

Page 15 of 40
310 Once this procedure is made for the first column, the students are shown to

connect the second column that is solved similarly to the first one. In both

processes, an outlet stream from the first column serves as the feed to the second,

while an outlet stream of the second is recycled to the first. If the linking

streams are not at the same conditions, intermediate equipment is needed. For

315 example, since pressure is different in the columns of a pressure swing process, a

t
ip
pump (in Pressure changers) may be required. In extraction distillation,

interheaters or intercoolers (in Exchangers) are often needed as well as solvent

cr
make-up to account for the losses in the product streams. These intermediate

us
apparatuses are designed in the Specifications window.

320
an
Sometimes the program does not converge; guidance on how to achieve it is

given; for example, by increasing the number of iterations in Blocks


M
Convergence or by changing the columns specifications. Additionally,

checking the .HIS file that contains the commands of each simulation may help
d

325 to find the problem. The troubleshooting guide in Appendix A of the book
te

Separation Process Engineering by Wankat, 2012 is very useful.


p
ce

1.1.1. Simulation of a reactive distillation process.


Ac

16

Page 16 of 40
The purpose of this latter session is to conduct the steady-state design of a

330 reactive distillation process. Real chemical systems as the acetic acid

esterification and the TAME and MTBE distillation columns are used as

examples. Deep analysis of these systems is done in the book by Luyben and

Yu, 2008.

t
By this point, the students are capable of solving an azeotropic distillation

ip
335 column, so the emphasis of this session is given to the introduction of a reaction

cr
zone in a distillation column. For that purpose they are trained on the use of

Reactions where the reaction type: kinetic, conversion or equilibrium must be

us
defined. Next the stoichiometric coefficients for the forward (positives) and

backward reaction (negatives) are introduced. Aspen Plus does not permit the

340
an
use of activities in the reaction rate expressions. User subroutines should be

used to incorporate this feature when necessary. Additionally, kinetic reactions


M
can only be used in the power law form. Consequently, the following step is to

introduce the exponent for each reactant and the activation energy. Instead, in
d

equilibrium reactions, the natural logarithm of the equilibrium constant k is


te

345 required.

The reaction zone is located using the window Blocks/Specifications/


p

Reactions. Next, liquid holdup needs to be provided because the net reaction
ce

rate on a reactive tray depends on the liquid holdup on that tray. Aspen Plus

allows introduction of this value in mols or in kilograms. In case of


Ac

17

Page 17 of 40
350 heterogeneous catalytic reactions, the weight of catalyst has to be introduced as

holdup. The simulation proceeds as shown in the previous labs. See Annex I for

more details about this simulation.

1.1. The lab assignment

t
355 After completing these practical sessions, the students are asked what if

ip
questions in order to force them to enhance the comprehension of the process

cr
and to continue working with the simulator on their own.

us
For example, after the first two sessions we ask the students to find the best

thermodynamical model for a ternary mixture that forms binary as well as

360 ternary azeotropes and to use it to calculate the bubble and dew point at certain
an
composition. They have to plot the ternary diagram with residue and distillation

curves, and to draw a residue curve starting from a specific composition.


M
After the third session, students are challenged to solve an industrial case. They
d

are given the specifications for the products, the feed composition and flow rate.

365 A critical comparison of the two processes is requested based on the dimension
te

of the columns and the energy requirements. We also ask them to model the
p

extractive distillation process with increasing number of stages on the azeotropic


ce

distillation column and to discuss the impact on the reflux ratio and on the heat

duties. In addition, they have to check what happens if feed location or the

370
Ac

solvent flow rate changes and the influence of pressure variation on the relative

18

Page 18 of 40
volatility. This work is quite demanding, so students do it in groups of two or

three members.

After the computer lab on reactive distillation, we ask them to analyze the

changes in various parameters that are specific for this process such as the liquid

375 holdup, the reactants feed location or the number of reactive trays. The impact

t
of this latter parameter is not straightforward and requires deep analysis of the

ip
composition profiles. In addition, they have to discuss the results if the reaction

cr
is equilibrium limited instead of kinetically limited or viceversa depending on

the case. We recommend the students to read chapter two of the book Reactive

us
380 Distillation Design and Control (Luyben and Yu, 2008) where all these

parameters are profoundly discussed.


an
Aspen Plus should not be used as black box. Thus, we request the students to

deliver a report summarizing and critically analyzing the results of their work.
M
We pretend the students apply their improved knowledge and their judgment to

385 do it that means to achieve the maximum evaluation level of Blooms taxonomy.
d
te

1. Teaching outcomes and conclusions


p

This course has been given during four academic years, 2012-2013, 2013-2014,
ce

2014-2015 and 2015-2016. At the end of the last three periods, we distributed a

390 survey with several questions shown in Table 1. The purpose was to know if the
Ac

computer labs and the tutorials were useful to understand the complex enhanced

19

Page 19 of 40
distillation processes as well as to improve the skills on the use of Aspen Plus.

We also asked the students if the proportion of the total grade assigned to these

practices was adequate based on their relative effort. In addition, they were

395 questioned about their satisfaction with the methodology.

The students completed the survey and returned it anonymously. They had to

t
rank each question from 1 (strongly disagree) to 5 (strongly agree). Table 1

ip
shows the results. On average, the answers were very favorable to the use of

cr
Aspen Plus as a tool for a better understanding of the enhanced distillation

400 processes. They also agreed on the temporary integration of the computer labs

us
during the course. Moreover, they found the labs and the tutorials very useful to

improve their knowledge on Aspen Plus. So they strongly recommended to keep


an
these labs and to extend their use to other advanced separation processes.

Consequently, we are preparing a practice on the use of Aspen Plus to design a


M
405 supercritical fractionation installation. They were less happy with the personal

work load that they have to do after the computer labs to complete the
d

assignments; some of them thought that it should count more on the total course
te

grade.
p
ce

410 We allowed the students to write comments at the end of the survey. Only few

students used this option, but all of them were related to the same issue. They
Ac

requested faster computers and easier access to them. The Aspen Plus license

20

Page 20 of 40
can be only used at the school computers; so the students had some problems to

access to them at certain moments. We transferred this request to the academic

415 authorities of our school.

The satisfaction of the students was also reflected into the teaching quality

questionnaire that our university does every year within the DOCENTIA

t
program. It includes sixteen questions with five levels of satisfaction. An

ip
analysis in three dimensions and three sub-dimensions is further conducted,

cr
420 comparing the answers of students with the maximum possible values. Table 2

summarizes the results of the academic year 2014-15. The evaluation was very

us
positive.

an
But what was more important, the resolution of the assignments was pretty good.
M
425 For example, the grades of the lab projects ranged from 7.2 to 9.4 for the

academic year 2013-14, and from 7.6 to 10 for the academic year 2014-15.
d

Moreover, the dropout rate was null in the four years we taught the subject. All
te

students passed the subject in the first call (except one in the academic year

2013-14). The marks were very good as shown in Fig. 1. Near 80% obtained
p

430 superior grades (NT or SB).


ce

References
Ac

21

Page 21 of 40
Alvarez, L., Sanjun, G., Martn, A., Calvo L., 2009. Design and cost evaluation

of a separation process for a multicomponent mixture using dense CO2. Ind.

435 Eng. Chem. Res. 48, 5779-5788.

Aspen Plus version 8.6 documentation, 2014, Aspen Technology, Inc.,

Burlington.

t
ip
Brunner, G., 1994. Gas extraction: an introduction to fundamentals of

supercritical fluids and the application to separation processes, Darmstadt:

cr
440 Steinkopff; New York: Springer.

us
Brunner, G., 2009. Counter-current separations. J. Supercritic. Fluids, 47, 574-

582.
an
Castrelln, T., Bota, D.C., Gmez, R., Orozco, G., Gil, D., 2009. Using process

simulators in the study, design and control of distillation columns for


M
445 undergraduate chemical engineering courses. Comput. Appl. Eng. Educ. 19,

621-630.
d
te

Dahm, K.D., Hesketh, R.P., Savelski, M.J., 2002. Is process simulation used

effectively in ChE courses?. Chem. Eng. Edu., 192-198.


p

Fortelius, C., Yli-Pentti, A., Halsas, M., Turunen, H., Seuranen, T., Akerman, M-
ce

450 J., 2015. A cooperative project based learning course for engineer students in
Ac

22

Page 22 of 40
biotechnology and chemical engineering. 3rd Int. Eng. Technol. Edu. Conference,

De Gruyter Open, DOI 10.1515/cplbu-2015-0002.

Henley, E.J., Seader, J.D., Roper, D.K., 2011. Enhanced distillation and

supercritical extraction, in: Separation process principles. John Wiley & Sons,

455 Inc., New York, pp. 448-495.

t
ip
Joyce, T., 2009. A project-based learning Design course: experience,

developments and assessment. Higher Education Academy,

cr
https://www.heacademy.ac.uk/resource/project-based-learning-design-course-

us
experience-developments-and-assessment.

460 Luyben, W.L., 2006. Distillation design and control using ASPEN simulation,

John Wiley & Sons, Inc., New Jersey.


an
Luyben, W.L., Yu, Ch-Ch., 2008. Reactive distillation design and control, John
M
Wiley & Sons, Inc., New Jersey.
d

Luyben, W.L., 2008. Comparison of extractive distillation and pressure-swing


te

465 distillation for acetone-methanol separation. Ind. Eng. Chem. Res. 47, 2696-

2707.
p

Luyben, W.L., 2013. Comparison of extractive distillation and pressure-swing


ce

distillation for acetone/chloroform separation. Comput. Chem. Eng. 50, 1-7.


Ac

23

Page 23 of 40
Mills, J.E., Treagust, D.F., 2003. Engineering education-is problem based or

470 project-based learning the answer?. Australas. J Eng. Edu., online publication,

2003-04. Retrieved from

https://www.researchgate.net/publication/246069451_Engineering_Education_Is

_Problem-Based_or_Project-Based_Learning_the_Answer

t
Muoz, R., Montn, J.B., Burguet, M.C., de la Torre J., 2006. Separation of

ip
475 isobutyl alcohol and isobutyl acetate by extractive distillation and pressure-swing

cr
distillation: Simulation and optimization. Sep. Purif. Technol. 50, 175183.

us
Rehfinger, A., Hoffmann, U., 1990. Kinetics of methyl tertiary butyl ether liquid

phase synthesis catalyzed by ion exchange resin I. Intrinsic rate expression in

an
liquid phase activities. Chem. Eng. Sci. 45, 1605-1617.

480 Rockstraw, D.A., 2005. Aspen Plus in the ChE curriculum, suitable course
M
content and teaching methodology. Chem. Eng. Edu. Winter, 69-75.

Schefflan R., 2011. Teach yourself the basics of Aspen Plus, AIChE and John
d

Wiley & Sons, Inc., New Jersey.


te

Seader, W.D., Seader, J.D., Lewin, D.R., 2004. Synthesis of separation trains,
p

485 in: Product and process design principles, John Wiley & Sons, Inc., New York,
ce

pp. 231-299.
Ac

24

Page 24 of 40
Venimadhavan, G. Buzad, G. Doherty, M.F. Malone. M.F., 1994. Effect of

kinetics on residue curve maps for reactive distillation. AIChE J. 10, 1814-1824.

Wankat, P.C., 2002. Integrating the use of commercial simulators into lecture

490 courses. J. Eng. Edu. 91, 19-23.

t
Wankat, P.C., 2006. Using a commercial simulator to teach sorption separations.

ip
Chem. Eng. Edu. Summer, 165-172.

cr
Wankat, P.C., 2012. Separation process engineering: includes mass transfer

analysis, third ed. Prentice Hall, New Jersey.

us
495

an
Fig. 1 Comparison of exam marks for academic years 2012-13, 2013-14 and

2014-15 (n=37).
M
d
p te
ce
Ac

25

Page 25 of 40
70
2014-15
60 2013-14
2012-13
50
Percentage

40

t
30

ip
20

cr
10

us
SB NT AP SS
500

an
Table 1 Results of the survey on the computer labs with Aspen Plus for
academic years 2013-14, 2014-15 and 2015-16 (n=55). 1= strongly
M
disagree, 2= disagree, 3= neutral, 4= agree, 5= strongly agree.

QUESTION
d

1 The labs with Aspen Plus helped me to understand the lecture material
te

2 The labs were adequately planned in the time schedule (after lectures and seminars)

3 The labs helped me to improve the use of Aspen Plus


p

4 The tutorials adequately describe the procedure to solve the practice


ce

5 The tutorials allowed solving the practice on your own

6 The resolution of a similar practice by the Professor beforehand is useful


Ac

26

Page 26 of 40
7 The proportion of the project assingments to the global course grade is fair

8 The personal work load of the project assingments is appropriate

9 The labs should be retained as part of the course

10 This type of practices with Aspen Plus should be used in other Advance Separation courses

505

t
ip
cr
Table 2 Results of the Teaching Quality questionnaire for the academic
year 2014-15 (n=20).

us
DIMENSIONS SUBDIMENSIONS

Planification,
information,
Informationan
Participation in coordination activities
organization and
M
coordination Organization of the subject

Formal fulfillment
d

Development Adequation of the methodology to reach the objetives


te

Teaching quality
p

Results Satisfaction of the students with the teaching


ce

Performance
Ac

27

Page 27 of 40
Review and reflexion for the enhancement of the teaching

510

t
ip
cr
us
an
M
d
te
p
ce
Ac

28

Page 28 of 40
Annex I. Tutorial for the simulation of a reactive distillation process.

MTBE production and purification by reactive distillation

The MTBE (methyl-tert butyl ether) is a gasoline component whose function is to


increase the octane number in unleaded gasoline. If its purification takes place through
5 simple distillation, the sequence of columns would be very complex since its mixture
with the reactants forms three minimum boiling and one maximum boiling binary

t
azeotropes. However, only one column is necessary by reaction distillation.

ip
The problem to be solved is as laid out in Henley et al., 2011:

cr
The reactive distillation column to produce MTBE is fed by one stream entering stage
10 10 consisting of 215.5 mol/s of methanol at 320 K and 11 bar, and another stream at
stage 11 of a mixture of 195.44 mol/s of isobutene and 353.56 mol/s of n-butene at 350

us
K and 11 bar. The column has 15 stages of liquid-vapor equilibrium, a total condenser,
a partial reboiler, a reflux ratio equal to 7 and a bottom flow of 197 mol/s. The kinectic
reaction between isobutene and methanol to produce MTBE according to Rehfinger and
15
an
Hoffman, 1990 takes place in liquid phase, in a temperature range between 40 and
100C, using a strong-acid ion-exchange resin as catalyst (4.9 eq/kg and 204.1
kg/stage). The column section where the reaction takes place is between stages 4 and
M
11.

a) Do the simulation of this scenario and discuss the results.


d

20 b) Compare the results with those obtained if the reaction is equilibrium limited
te

when sulfuric acid is used as catalyst according to the data by Venimadhavan et


al., 1994.
p

Lab. 4. Part a) Tutorial of the simulation of a reactive distillation column with a


ce

25 kinetically controlled reaction.

1. In Setup tab in Properties window, type a title for the simulation, and change units
Ac

to MET.

2. Do the first 9 steps of lab 3 to introduce components (methanol, isobutene, 1-butene,


MTBE) and properties. In this simulation the thermodynamic model used is UNIQ-RK.
30 Binary interaction parameters are introduced by the user. To do that, go to Data
Browser > Properties > Parameters > Binary interaction > UNIQ-1.

1
Page 29 of 40
t
ip
cr
us
3. Select a RADFRAC column in the available blocks menu, Columns tab. Pick it to
the flow diagram.

35 4. Add FEED, METANOL, D1 and B1 streams. an


M
d
te

5. Add the specifications of FEED stream: temperature, 350 k, pressure, 11 bar, molar
p

flow of isobutene, 195.44 mol/s and of 1-butene, 353.56 mol/s.


ce
Ac

40 6. Add the specifications of the METHANOL stream: temperature, 350 k, pressure, 11


bar, molar flow of methanol, 215.5 mol/s.

2
Page 30 of 40
t
ip
cr
7. Add the reaction. Go to Reactions > New. Name it as R-1, and select the type of

us
reaction, in this case REAC-DIST. Click on New. Select the type of reaction as:
45 kinetic, conversion, equilibrium and click on OK.

an
M
d
p te

(1)
ce

For the forward reaction, the rate law is formulated in terms of mole-fraction
concentrations (Rehfinger and Hoffmann, 1994).
Ac

50 (2)

The corresponding backward rate law, consistent with chemical equilibrium, is

(3)

, where r is in moles per second per equivalent of acid groups, R=8.314 J/mol-K, T is in
K, and Ci is liquid mole fraction.

55 8. In the following window, introduce the type of reaction, in this case Kinetic. In
reactants introduce isobutene and methanol. The stoichiometric coefficient for
isobutene and methanol is 1 (negative stoichiometric coefficient for reactive and

3
Page 31 of 40
positive for products), and the exponent is 1 for the isobutene and -1 for the methanol.
Next, introduce the product: MTBE, with coefficient 1 and exponent 0.

t
ip
cr
60

9. Do the same for the backward reaction.

us
an
M
d
te

10. Now, it is time to introduce the kinetic parameters. In Reactions > R-1, introduce
p

the parameters in the Kinetic tab. For each reaction, select Use built-in Power Law,
ce

65 select the reaction 1, in Reacting phase select Liquid, introduce k value, and the value
of the activation energy in kJ/kmol. Finally, select the [Ci] basis in Mole fraction.
Ac

4
Page 32 of 40
11. Do the same for reaction 2.

t
ip
cr
us
70 12. Define the RADFRAC column, in Data Browser > Blocks > B1.

Calculation type: Equilibrium.

Number of stages: 17.


an
Condensador: Total.
M
Convergence: Strongly non-ideal liquid.

75 In Operating specifications, indicate that Bottoms rate, 197 mol/s, and the
d

Reflux ratio, 7.
p te
ce
Ac

13. In Streams tab indicate that the entering stage for FEED stream, 11 and for
Methanol, 10.

5
Page 33 of 40
t
80

ip
14. In Pressure tab, indicate that the condenser pressure, 11 bar.

cr
us
an
M
15. In Blocks > C1 > Specifications > Reactions, indicate that the reaction that takes
place in the column R-1, and between 4 and 11 stages.
d
p te
ce
Ac

85

16. Next, define the Holdup in Holdups tab. A holdup of 8000 kg between stages 4 and
11 is used.

6
Page 34 of 40
t
ip
17. Click on Next. The results of the simulation appear as shown in the following

cr
90 screenshot.

us
an
M
d
p te
ce

18. Obtain the composition profile in Blocks > C1 > Profiles > Plot > Composition.
Ac

7
Page 35 of 40
t
ip
cr
us
19. Represent the generation rate of MTBE in Blocks > C1 > Profiles > Reactions >
95 Plot > Custom.
an
M
d
p te
ce
Ac

8
Page 36 of 40
t
ip
cr
us
20. Finally, obtain the energetic consumption of the condenser and the reboiler for the
C1 column, in Blocks > C1 > Results.
an
M
d
p te
ce
Ac

100

Lab. 4. Part b) Tutorial of the simulation of a reactive distillation process with an


equilibrium controlled reaction.

1. Do again the same steps until number 7.

2. Add the reaction that takes place in Reactions > New. Name it as R-1, and select the
105 type of reaction, in this case REAC-DIST. Click on New.

(1)

9
Page 37 of 40
(2)

3. In the following window, you have to introduce the reaction, in this case
Equilibrium. In reactants introduce isobutene and methanol; the stoichiometric
110 coefficient of isobutene and methanol is -1 (negative stoichiometric coefficient for
reactive and positive for products) and for MTBE is 1.

t
ip
cr
us
an
4. Next, you have to introduce the equilibrium parameters. In Data Browser >
Reactions > Reactions > R-1, introduce the parameters in Equilibrium tab. Indicate
M
115 that the Keq basis is mole gamma and the equilibrium temperature, 350 K. Tick
Compute Keq from built-in expression. Introduce the value of A equal to 16.33
and B equal to -6820.
d
p te
ce
Ac

5. Click on Next to start the simulation. To visualize the results do same procedure as
120 described in stages 18 to 20 of the previous simulation.

10
Page 38 of 40
Ac
ce
pte
d

11
M
an
us
cr
ip
t

Page 39 of 40
t
ip
cr
us
an
M
d
pte
ce

125
Ac

12
Page 40 of 40