Citation
Chemical engineering education

Material Information

Title:
Chemical engineering education
Alternate Title:
CEE
Abbreviated Title:
Chem. eng. educ.
Creator:
American Society for Engineering Education -- Chemical Engineering Division
Place of Publication:
Storrs, Conn
Publisher:
Chemical Engineering Division, American Society for Engineering Education
Publication Date:
Frequency:
Quarterly[1962-]
Annual[ FORMER 1960-1961]
quarterly
regular
Language:
English
Physical Description:
v. : ill. ; 22-28 cm.

Subjects

Subjects / Keywords:
Chemical engineering -- Study and teaching -- Periodicals ( lcsh )
Genre:
periodical ( marcgt )

Notes

Citation/Reference:
Chemical abstracts
Additional Physical Form:
Also issued online.
Dates or Sequential Designation:
1960-June 1964 ; v. 1, no. 1 (Oct. 1965)-
Numbering Peculiarities:
Publication suspended briefly: issue designated v. 1, no. 4 (June 1966) published Nov. 1967.
General Note:
Title from cover.
General Note:
Place of publication varies: Rochester, N.Y., 1965-1967; Gainesville, Fla., 1968-

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
01151209 ( OCLC )
70013732 ( LCCN )
0009-2479 ( ISSN )
Classification:
TP165 .C18 ( lcc )
660/.2/071 ( ddc )

UFDC Membership

Aggregations:
Chemical Engineering Documents

Downloads

This item has the following downloads:


Full Text







[en M teaching tips )


This one-page column will present practical teaching tips in sufficient detail that ChE educators can
adopt the tip. The focus should be on the teaching method, not content. With no tables or figures
the column should be approximately 450 words. If graphics are included, the length needs to be
reduced. Tips that are too long will be edited to fit on one page. Please submit a Word file to Phil
Wankat , subject: CEE Teaching Tip.



TEACHING TIP: ELEVATOR TALKS

PHIL WANKAT
Purdue University West Lafayette, IN 47907


Both industry and ABET require that engineering gradu-
ates can communicate. Clearly the best way to achieve this
is to have frequent assignments throughout the curriculum
requiring writing and oral presentations. Unfortunately, oral
presentations tend to require a significant amount of class
time. An alternative oral presentation is the "elevator talk."
The scenario: a student steps into an elevator with someone
she needs to persuade or sell. For example, the student may
want to convince the person to hire her. She has from one to
two minutes to do this.


I assigned the
topic to the stu-
dents (ask for a
job), gave them
the time (two
minutes), gave
them a copy of
the scoring ru-
bric (Table 1),
and told them
to prepare a
talk that they
will present
extemporane-
ously, without
visuals. There
was no written
assignment. In
class,I assigned
the "boss" for
each person.
Students were


told to assume that they knew the boss well enough to talk
to. Presenters and bosses went to the front of the room and
stood in the elevator. Talks were timed for a strict two min-
utes. Since two minutes is actually fairly long, most students
finished early and had to do something-perhaps just stand
there-for the remaining time. If they weren't finished at two


minutes, the elevator door opened anyway and they had to
summarize very quickly.

The students saw the relevance of elevator talks and were
well prepared. Grading the talks with the scoring rubric was
straightforward and I was able to finish the grading while
the next pair walked to the front. Since it takes less than 30
seconds to change speakers, 20 two-minute talks can be done
in a 50-minute period.

While not eliminating the need for more formal presen-


stations, eleva-
tor talks can
provide an
easy way to
include oral
communica-
tion in courses
that normally
would not
have time.
Grading all of
the talks with
the scoring ru-
bric and then
saving cop-
ies provides
evidence for
ABET that all
students have
been assessed
and can do
oral presenta-


tions, at least at the barely acceptable level.

REFERENCES
1. Mitchell, B.S., and V.J. Law, "Community-Based Presentations in the
Unit Ops Laboratory," Chem. Eng. Ed., 39(2), 160 (2003)


Copyright ChEDivision ofASEE 2006


TABLE 1
Scoring Rubric for Elevator Talks. Adapted from Mitchell and Law.1'
Attribute Not Barely Meets Exceeds
Acceptable Acceptable Expectations Expectations
Logical topic Disjointed; no Parts out of Organized by Superior;
order organization order guidelines enhances com-
munication
Appropriate Far too long or Somewhat Appropriate
time use too short long or short length
Objective Not stated Poorly stated Clearly stated
Background & Neither stated Only one Both stated Both clearly
Significance stated stated
Conclusions None Present, but Logical & Logical & supe-
not logical clearly stated rior explanation
Presentation Many Some No distractions Superior
mechanics* distractions distractions presentation
Response to Not responsive Incomplete Clear and Complete
questions (if any) direct
Focus on person Not focused; Some focus; Focused with Totally
speaking to distracted, no some eye good eye focused; excel-
eye contact contact contact lent eye contact
* voice, poise, mannerisms














Author Guidelines for the

LABORATORY

Feature

The laboratory experience in chemical engineering education has long been an integral part
of our curricula. CEE encourages the submission of manuscripts describing innovations in the
laboratory ranging from large-scale unit operations experiments to demonstrations appropriate
for the classroom. The following guidelines are offered to assist authors in the preparation of
manuscripts that are informative to our readership. These are only suggestions, based on the
comments of previous reviewers; authors should use their own judgment in presenting their
experiences. A set of general guidelines and advice to the author can be found at our Web site:
.


> Manuscripts should describe the results of original and laboratory-tested ideas.
The ideas should be broadly applicable and described in sufficient detail to
allow and motivate others to adapt the ideas to their own curricula. It is noted
that the readership of CEE is largely faculty and instructors. Manuscripts must
contain an abstract and often include an Introduction, Laboratory Description,
Data Analysis, Summary of Experiences, Conclusions, and References.
An Introduction should establish the context of the laboratory experi-
ence (e.g., relation to curriculum, review of literature), state the learning
objectives, and describe the rationale and approach.
The Laboratory Description section should describe the experiment in
sufficient detail to allow the reader to judge the scope of effort required
to implement a similar experiment on his or her campus. Schematic dia-
grams or photos, cost information, and references to previous publica-
tions and Web sites, etc., are usually of benefit. Issues related to safety
should be addressed as well as any special operating procedures.
If appropriate, a Data Analysis section should be included that concisely
describes the method of data analysis. Recognizing that the audience
is primarily faculty, the description of the underlying theory should be
referenced or brief. The purpose of this section is to communicate to the
reader specific student-learning opportunities (e.g., treatment of reac-
tion-rate data in a temperature range that includes two mechanisms).
The purpose of the Summary of Experiences section is to convey the
results of laboratory or classroom testing. The section can enumerate,
for example, best practices, pitfalls, student survey results, or anecdotal
material.
A concise statement of the Conclusions (as opposed to a summary) of
your experiences should be the last section of the paper prior to listing
References.













EDITORIAL AND BUSINESS ADDRESS:
Chemical Engineering Education
Department of Chemical Engineering
University of Florida Gainesville, FL 32611
PHONE and FAX: 352-392-0861
e-mail: cee@che.ufl.edu


EDITOR
Tim Anderson

ASSOCIATE EDITOR
Phillip C. Wankat

MANAGING EDITOR
Lynn Heasley

PROBLEM EDITOR
James O. Wilkes, U. Michigan

LEARNING IN INDUSTRY EDITOR
William J. Koros, Georgia Institute of Technology


-PUBLICATIONS BOARD
CHAIRMAN *
E. Dendy Sloan, Jr.
Colorado School of Mines
VICE CHAIRMAN *
John P. O'Connell
University of Virginia

MEMBERS
KristiAnseth
University of Colorado
Pablo Debenedetti
Princeton University
Dianne Dorland
Rowan University
Thomas F. Edgar
University of Texas at Austin
Richard M. Felder
North Carolina State University
Bruce A. Finlayson
University of Washington
H. Scott Fogler
University of Michigan
Carol K. Hall
North Carolina State University
William J. Koros
Georgia Institute of Technology
Steve LeBlanc
University of Toledo
Ronald W. Rousseau
Georgia Institute of Technology
Stanley I. Sandler
University of Delaware
C. Stewart Slater
Rowan University
Donald R. Woods
McMaster University


Chemical Engineering Education
Volume 40 Number 3 Summer 2006


> DEPARTMENT
146 The University of Sherbrooke
J. Peter Jones, Bernard Marcos, Gervais Soucy

> EDUCATOR
154 Susan Montgomery of the University of Michigan
Scott Fogler, Lara Zielin

> CLASSROOM
165 Using A Commercial Simulator To Teach Sorption Separations
Phillip C. Wankat
203 A Tire Gasification Senior Design Project That Integrates Laboratory
Experiments and Computer Simulation
Brian Weiss, Marco J. Castaldi

> RANDOM THOUGHTS
173 How To Teach (Almost) Anybody (Almost) Anything
Richard M. Felder, Rebecca Brent

> CURRICULUM
175 Hyper-TVT: an Interactive Learning Environment
Marina Santoro, Marco Mazzotti
181 Integrating Biological Systems in Process Dynamics and Control Curriculum
Robert S. Parker, Francis J. Doyle III, Michael A. Henson
231 Enhancing the Undergraduate Computing Experience
Thomas F Edgar

> LEARNING IN INDUSTRY
189 The Role of Industrial Training in Chemical Engineering Education
Mamdouh T. Ghannam

> LABORATORY
159 An Agitation Experiment with Multiple Aspects
Jordan L. Spencer
195 Validating The Equilibrium Stage Model for an Azeotropic System in a
Laboratorial Distillation Column
B.P.M. Duarte, M.N. Coelho Pinheiro, D.C.M. da Silva, M.J. Moura
215 Plant Design Project: Biodiesel Production Using Acid-Catalyzed Trans-
esterification of Yellow Grease
Rafael Hernandez, Trent Jeffreys, Anirudha Marwaha, Mathew Thomas
225 Experimental Investigation and Process Design in a Senior Lab Experiment
Kenneth R. Muske

> OUTREACH
211 A Simple Viscosity Experiment for High School Science Classes
TM. Floyd-Smith, K.C. Kwon, J.A. Burmester F.F Dale, N. Vahdat, P Jones

> CLASS AND HOME PROBLEMS
239 Fuel Processor System for Generating Hydrogen for Automotive Applications
Panini K. Kolavennu, John C. Telotte, Srinivas Palanki

CHEMICAL ENGINEERING EDUCATION (ISSN 0009-2479) is published quarterly by the Chemical Engineering
Division,American SocietyforEngineering Education, and is edited at the University ofFlorida. Correspondence regarding
editorial matter, circulation, and changes of address should be sent to CEE, Chemical Engineering Department, University
of Florida, Gainesville, FL 32611-6005. Copyright 0 2005 by the Chemical Engineering Division, American Society for
EngineeringEducation. The statements and opinions expressed in this periodical are those of the writers and not necessarily
those of the ChEDivision,ASEE, which body assumes no responsibilityfor them. Defective copies replaced ifnotified within
120 days ofpublication. Writefor information on subscription costs andforback copy costs and availability. POSTMASTER:
Send address changes to Chemical Engineering Education, Chemical Engineering Department., University of Florida,
Gainesville, FL 32611-6005. Periodicals Postage Paid at Gainesville, Florida and additional post offices.


Summer 2006










r !j =department
---- U s_____________________________________


Chemical Engineering at


the University of Sherbrooke


J. PETER JONES, BERNARD MARCOS, AND GERVAIS SoucY


he University of Sherbrooke is a French-language
university in the city of Sherbrooke, Quebec. It is 100
miles east of Montreal, about 30 miles from the Ver-
mont border, and approximately due north of Boston, Mass.
The Main Campus, which houses the university administration
and eight faculties, and the separate Health Campus -with the
Faculty of Medicine and Health Sciences, and forming part
of the Sherbrooke University Hospital Centre Complex-are
all located in Sherbrooke, at the heart of the beautiful Eastern
Townships region of southern Quebec. The areas well known
for its many rivers, lakes, and mountains. Part of the Northern
Appalachian chain of mountains, it is a favored cottage, ski-
ing, and recreation area for Montrealers, among others.

THE UNIVERSITY
The university includes nine faculties and offers more than
260 study programs, at both the undergraduate and graduate
levels.
The university has been experiencing unprecedented
development: Since 2001, 650 people have been hired. In
fact, more than 40% of its currently employed professionals


joined the university within the last four years. In keeping
with this growth, the university is presently investing some
$310 million to renovate and make additions to existing
buildings. Close to 35,000 students attend courses at the
University of Sherbrooke, some 85% coming from outside
the region; more than 1,300 international students are enrolled
on a yearly basis.
A study, published annually for the past three years in The
Globe and Mail national newspaper, reveals that the Univer-
sity of Sherbrooke has consistently been the most appreciated
university in Quebec and is among the three top-ranked uni-
versities in Canada. According to MacLean magazine's annual
study, the University of Sherbrooke enjoys the best overall
reputation in Quebec. The university is first in Canada for the
excellence of its cooperative system, which allows students to
alternate between study terms and paid-work terms, and for
the high quality of services offered to its students. The co-op
education system provides a large number of students with
paid-work terms, giving them an opportunity to combine the
theoretical notions acquired in the formal classroom with their
practical experiences as received in the workplace. Students


Copyright ChE Division ofASEE 2006


Chemical Engineering Education



















Known as
The Agora, the
campus fountain
is a frequent
fair-weather
gathering spot of
Sherbrooke students
and visitors alike.

















involved in Sherbrooke's cooperative education system earn
more than $30 million in salaries annually, for more than
4,000 paid-work terms.

A LEADER IN RESEARCH AND CREATION
The University of Sherbrooke has identified the fields of
expertise it intends to develop, in both teaching and research,
to meet tomorrow's requirements both nationally and inter-
nationally. As well, the university is addressing increasing
numbers of requests for "partnerships" from institutions in
Europe, Latin America, North Africa, and Asia concerning
its master's and doctoral degree programs, especially in the
fields of education, administration, cooperative management,
and applied ethics. It also receives the most royalties of any
Canadian university from past inventions by professors and
researchers. Thus, it has received more than $79 million
royalties to date, including $14.3 million in the 2002-2003
academic year alone. In addition, some 22 spinoff companies
have been created by the University of Sherbrooke over the
past 20 years.
In fact, the university holds title to 300 patents (both
established and pending)-51% having been transferred to
businesses. Notable among these innovations is the ACELP
technology, developed at the University of Sherbrooke, which
Summer 2006


has become the standard in mobile telephony (with more
than a billion users) and on the Internet (with more than 500
million users).

RECENT EARLY HISTORY
Sherbrooke's Department of Chemical Engineering started
its existence as a process engineering section in the Depart-
ment of Mechanical Engineering. Originally, there were three
professors: Bernard Coupal, Andr6 Marsan, and a French
military co-operant, Bernard Koehret. They were later
joined by Maurice Ruel, Esteban Chornet, and Normand
Th6rien. In December 1971, under the determined leadership
of Coupal, the department was established as a full depart-
ment in its own right, with a distinct program of studies in
chemical engineering.
The battle to become a separate department was difficult
as other departments were wary they would lose scarce re-
sources, but Coupal advised them that the establishment of a
Department of Chemical Engineering could be done at "zero
cost." Some of the older professors from other departments long
remembered his statement, and used it liberally when later the
department was fighting for an increased share of resources.
The department continues to be the only one in a Quebec
university to have a fully cooperative program.













Sherbrooke's Chemical Engineering Faculty


SNicolas Abatzoglou's activities are in the areas of
thermo-catalytic chemical reactors and the behavior of
particulate systems in reactive and nonreactive indus-
trial processes. His work is being used by companies
for the gasification process commercialization, for
conditioning of industrial gases, and for pharmaceuti-
cal product formulation. He has extensive scientific
and industrial R&D experience in fields at the "juncture" of energy and
environment as well as in dry formulation of pharmaceutical products,
and has initiated a substantial program on Process Analytical Technology
for the pharmaceutical industry. He teaches various courses including the
capstone design course, reaction engineering, pharmaceutical engineer-
ing processes, and separation and purification in biotechnology.
Maher Boulos is a leading figure in the thermal
plasma field, along with colleagues in Minnesota,
France, Switzerland, and Japan. His work through
the Sherbrooke Plasma Research Center has been
broadly based and includes many novel experimen-
tal studies, but it has also included industrial scale
development studies, modeling, and-with collabo-
rators- more "in depth" theoretical work.
Esteban Chornet is a leading international figure
in research whose work has led to the production
of chemicals and energy values from biomaterials
and organic wastes. The first chemical engineer in
Canada to obtain the prestigious Steacie Fellowship
for research, he is the founder of the research center
on the transformation of biological materials. Much
of his recent work has been applied to environmental concerns (e.g.,
"green"chemical engineering, biomaterials recycling, and environmental
concerns). He has initiated a number of high-tech spinoffs in Quebec and
has also developed spinoffs in his native Catalonia.
Nathalie Faucheux, a biochemist with a Ph.D. in
biomedical engineering, plans to determine how
biomaterials and cells can share information with
one another. She is one of the few researchers in the
world working to gain a deeper understanding of how
biochemical signals, triggered by contact with a bio-
material, activate a cell's capacity to survive, multiply,
and function. The cutting-edge materials she is using are based on grafts
of small molecules called peptides, which, among other things, promote
cell adhesion. She holds a Tier 2 Canada Research Chair.
Francois Gitzhofer was recruited for his expertise in
materials engineering from the University of Limoges,
France. He has become director of the department's
Plasma Research Group, which is continuously
evolving toward a broader role in fuel cell develop-
ment and other energy-intensive applications. His
focus is on creating new ways of making coatings on
various substrates, using both established and newly emerging plasma
technologies.
Denis Gravelle, one of the earliest professors to
join the department, is an expert in the application
of thermodynamics to thermal plasma systems and
in the use of spectroscopic methods/techniques for a
fuller understanding of plasma torch dynamics. A key
teacher of thermodynamics at both the graduate and
undergraduate levels, he is also involved in the lab
courses, aiming at the integration of basic engineering
concepts of thermodynamics, transport phenomena,
and reaction kinetics.
Michele Heitz, the associate dean for students (Engi-
neering Faculty) and a full professor, teaches introduc-
tion to chemical engineering, thermodynamics, and


chemical thermodynamics. She is also in charge of integrating projects
for both the chemical and the biotechnological first-year students. Since
2002, she has also taught air pollution control and design, introduction to
biochemical engineering, and chemical kinetics and reactor engineering.
Her current research projects include air treatments by biofiltration as
well as biodiesel production and biomass and whey valorization-both
topics involving various chemical and biotechnological approaches.
Peter Jones arrived at the Sherbrooke shortly after
the department was created. He has developed a
research area in industrial water treatment and the
application of statistical methods to environmental
problems, and more generally, to experimental re-
search. A past chairman of the department, he was
S director of the environmental engineering and sci-
ence master's program and vice dean for research.
Jerzy Jurewicz is a specialist in the area of plasma generation, using
either direct or high-frequency AC currents. He ap-
plies this expertise to the development of new reac-
tors for the synthesis of new products, especially for
nanometric powders. Jerzy is responsible for courses
in safety in the first year and also in process safety
courses in the fourth year, and is a key member of the
design course team.
Bernard Marcos has done much
work in the field of expert systems and neural networks
as well as pursuing educational research such as the
use of a system of intelligent tutorials. He was the
first director of the new program in biotechnological
engineering.
Pierre Proulx has been involved in mathematical
modeling of thermal plasmas since his graduate
studies under the supervision of Maher Boulos in the
Plasma Research Center. His current projects entail
mathematical modeling of complex reactors. He
teaches transport phenomena and process control.
JoRl Sirois, the most recently hired
professor, comes with a very strong background in
biotechnology and was recently the chief technology
officer at a start-up company in this field. His focus
is in the areas of characterization, modeling, and
optimization of cell metabolism and the design and
scale-up of bioreactors.
Gervais Soucy is current chair of the department.
His teaching is concentrated in unit operations. His
research field is in the application of new technolo-
gies for various processes in the aluminium industry.
He has also developed expertise in thermal plasma
technology to produce carbon nanostructures.
Normand Therien, recently retired from the depart-
S ment, is an expert in the application of mathematical
techniques to environmental problems. His most
important work has been in the area of modeling
of hydroelectric reservoirs, where he has been an
S important figure in determining how mercury can
accumulate in reservoirs and render fish unfit for hu-
man consumption.
Patrick Vermette is a researcher at the Research Centre on Aging and
is an engineer and professor in the department, with a joint appoint-
ment in the Service of Orthopaedics. He has built a
state-of-the-art laboratory for surface science and tissue
engineering studies. He and co-workers are involved in
fields including biomaterials, angiogenesis, colloids and
interface science, drug delivery systems, bioreactors, tis-
sue engineering, and haematopoietic stem cells.


148 Chemical Engineering Education











THE COOPERATIVE PROGRAM
The Faculty of Engineering at the University of Sherbrooke
was a very early adopter of the cooperative system of engi-
neering in North America. The University of Waterloo had
initially adopted the cooperative system in 1956, and the
University of Sherbrooke followed in 1966, before the De-
partment of Chemical Engineering was formed. The co-op
system seeks to prepare students for their future careers)
by providing the practical experience that meets employer
requirements in the workplace. Thus, the work term offers
students the opportunity to acquire practical experience and
to develop competencies (knowledge, skills, attitudes, values,
etc.) relevant to their future careers. Cooperative education,
a pedagogical approach whereby students spend alternate
trimesters studying in the classroom and earning wages in
the workplace, also offers many valuable features to potential
employers. As a pioneer of cooperative education in Quebec,
the University of Sherbrooke is proud to be a leader in this
expanding field. A result is that the University of Sherbrooke
now ranks second in Canada-and is among the top five
advanced learning institutions in North America-for the


importance given to its cooperative education system. In
the Department of Chemical Engineering, students are able
to achieve this gradual integration by switching alternate
trimesters between their paid terms in the workplace and
their study terms at the university. When students graduate,
they will have served five work terms (15 weeks / term) in an
industrial company (90%) or in a research laboratory (10%).
Since there are eight academic sessions, the student obtains
his B.Ing. degree after a period of 52 months.


THE UNDERGRADUATE PROGRAMS

Chemical Engineering

The Department of Chemical Engineering has always kept
up to date with the needs of employers in Quebec and the
rest of Canada.
The studies program in chemical engineering was com-
pletely overhauled for students set to begin in September 2001
and graduate in December 2005. Table 1 presents the curricu-
lum of the reformed chemical engineering program. We have
since initiated five cohorts, or graduating classes, to this new


Summer 2006 14S


TABLE 1
Chemical Engineering

Session Description

S-1 Introduction to chemical engineering
The role of the engineer, safety and risks, chemistry, communications
S-2 Measurement techniques for use in the laboratory and the factory/plant
Instrumentation, chemistry, chemical analysis techniques, reports, controls
T- Work term #1
At the end of the first year, the student should be capable of describing the chemical engineer's role and of undertaking control actions and performing
analyses, both in the laboratory and at the plant; thereby displaying, at an early stage, a satisfactory competence level in performing the necessary tasks.
S-3 Transport and exchange in fluids
Fundamentals of chemical transport/transfers in processes
T-2 Work term #2
S-4 Design of the basic units employed for a chemical process
Advanced chemical transfer/transport, chemical reactor and associated units calculations
At the end of the second year, the student should be capable of modeling the operations of several parts of a functioning chemical process plant.
T-3 Work term #3
S-5 Industrial scale plant operations
Control methods, techno-economics, process control laboratories
S-6 Design basics of industrial-scale chemical processes
Types of processes, process simulation, environmental and safety aspects
At the end of the third year, the student should be capable of designing the unit parts and creating the basic overall process concept for an industrial-
scale process.
T-4 Work term #4
S-7 Combining process design skills and experiences I
Integration of all aspects required to establish, modify, and operate a chemical industry installation at an important scale
T-5 Work term #5
S-8 Combining process design skills and experiences II
Integration of all aspects required to establish, modify, and operate a chemical industry installation at an important scale
At the end of the fourth year, the student should be capable of designing the unit parts and creating the basic overall process, taking into account rel-
evant aspects of process economics as well as social and environmental issues.


I










I







I








I





























Sherbrooke students pitch in on a project in one of the
university's well-equipped labs.


regime. This overhaul was dictated by our knowledge of the
companies that hire our graduates. We were also influenced
by the tradition of innovation in pedagogical methods used
in the university's Faculty of Engineering. We have aligned
our developments with the goal that engineering graduates
of our programs will be responsible for the development of
new products and processes. Most of these new products
and processes are not even mentioned in traditional course
materials for aspiring chemical engineers, yet our students
must now develop ability to work creatively in these areas.
We have therefore developed three distinct avenues for go-
ing forward:
Students are responsiblefor their own development,
which is central to the educational process. They must
combine their technical development with the simultane-
ous improvement of their leadership skills, their entrepre-
neurship, their teamwork skills, and their respectfor
their profession.
Students must now take full and early advantage of the
new -* *i q .ri ,, and information technologies at their dis-
posal, especially the software products specific to their
profession and their technical competence as chemical
engineers.
Courses cannot be approached separately. We and the
students must find the commonalities .-: .. -, the session-
al projects program, activated during each of the first,
third, and fourth years.

The result has been an immersion in the practice of chemi-
cal engineering from the very first session. Students were
often ill-prepared for some of the challenges they faced, but
through their initiative and determination they were able,
nevertheless, to produce very interesting results. The projects
performed in those early sessions required a lot of "digging"
to find pertinent information. This information was then ap-


plied to experimental setups, which they also had to design.
They did, however, have considerable help from professors
leading the courses) for each session, along with assistance
from departmental technicians. The project is a capstone
design project, with which we have now had considerable
success for a number of years. These projects are presented
at the CSChE competitions. We have won the SNC-Lavalin
prize a number of times.
The 40 courses offered in the program are distributed in
the following manner:
11 courses as general engineering courses in mathemat-
ics, thermodynamics, and materials
17 courses in chemical engineering, transport phenom-
ena, unit operations, and reactor design
six courses in humanities and social science, law, ethics,
and engineering economics
six courses in the students'chosen major

Students are mentored by their more-senior peers, who help
them become accustomed to the range of department opera-
tions and also provide them with professional contacts at the
very beginning of their professional careers. The unifying
projects chosen for each session are an excellent initiation
to the later work terms, following two sessions spent in the
department. Through the projects students also learn about the
human and societal aspects of their chosen profession.
Biotechnological Engineering
The numerous recent developments in biotechnology and
in the medical sciences have led the Engineering Faculty to
readapt its curricula under the belief that the most precious
asset of a profession is its intellectual core. In an era of rapid
evolution in biotechnology-based industry, it is imperative
that the biotechnological engineering discipline define its own
core. It must strengthen its core through scholarly activities
and diverse applications. Considering that biotechnology
constitutes a broad field, biotechnological engineers need to
integrate skills in engineering principles, process engineering,
and biological sciences, without being restricted to particular
applications. Biotechnological engineering programs must
take into account the complexity of living systems with
their discrete and nonlinear relationships. The integration
of complex engineering principles is not a simple task, and
the biology-engineering barrier is an obstacle that has to be
overcome. It is not sufficient to incorporate biological science
courses into a chemical engineering curriculum hoping that
students will be capable of integrating both concepts. Bio-
technological engineers must eliminate the present gap and
correct misunderstandings between traditional engineers and
biologists. They must accept the fact that living organisms
are not entirely predictable. Consequently, they must master
basic knowledge of living organisms and bioproducts, and
of the fundamental unit operations and simulation tools used


Chemical Engineering Education











by engineers. They must understand the physiology of pro-
karyotes and eukaryotes as well as the engineering concepts
used in bioprocesses. Biotechnological engineers must also
be able to operate and control small- and large-scale culture
systems of cells and micro-organisms for the production of
products of commercial potential (e.g., proteins, antibiotics)
as well as the downstream processing, including separation
and purification of biomacromolecules. Finally, they must be
skilled in project management and quality control. Broadly
speaking, biotechnological engineers will be called upon to
solve problems through the development of bioproducts and
bioprocesses that use living organisms or the products they
synthesize.
The biotechnological engineering program at the University
of Sherbrooke was originated by the departments of biology
and chemical engineering. It took four years to build the
program, which now offers an integrated training in biotech-
nological engineering. Table 2 presents the curriculum of the
new Biotechnological Engineering program at Sherbrooke.
The program is divided into eight terms that include labora-
tory studies, applied projects, and lectures.
Creating a new discipline may present some drawbacks for


the employment of new graduates. Industry will need to learn
what a biotechnological engineer is -just as it understands
what chemical engineers and biologists are. That is why the
biotechnological engineering program was developed with
industrial partners. These industrial partners are regularly
updated on the curriculum's development. Another area of
concern is that biotechnological engineers may be too nar-
rowly trained and too application-oriented. As explained
previously, the biotechnological engineering program is a
science-based program that should ultimately alleviate a
too-narrow perspective.
The goal of the training is to prepare a generalist engineer
who is able to manage the evolution of the biotechnology
industry. The wide breadth of knowledge of tomorrow's
biotechnological engineers will be an important advantage
to them.
Several segments of the bio-industry are relevant to the
employment of future biotechnological engineers: bio-
pharmaceutical and drug companies; agribusiness and food
companies; environmental biotechnology companies; bio-
medical instrumentation companies; biomaterials; and the
tissue-engineering sector.


TABLE 2
Biotechnological Engineering

Session Description
S-1 Introduction to Biotechnological Engineering
Role of the engineer, safety and risks, biochemistry, information technology
S-2 Introduction to Biology
Biology, cell biology, functional biology, microbiology, laboratory techniques
At the end of the first year, the student should be able to describe the role of the engineer and to make measurements and perform laboratory analysis.
S-3 Genetics and transport phenomena
The fundamentals of genetics and chemical transport
T-1 Work term #1
S-4 Design of basic units
Advanced transport, unit operations, experimental protocols
At the end of the second year, the student should be able to model the behavior of a number of the units that make up a biotechnological process.
T-2 Work term #2
S-5 Operation of industrial unit processes
Bioreactors, process control, engineering economics, biological polymers
T-3 Work term #3
At the end of the third year, the student should be able to design a process and the functional units to make it operate.
S-6 Downstream operations
Separation and purification, materials and biomaterials, biomolecular engineering
T-4 Work term #4
S-7 Design of biotechnological processes
GMP-GLP standards, process simulation, design
S-8 Integration of the abilities required to design a chemical process
Integrate all the aspects related to the building, the modification, and the operation of a large biotechnological industrial installation
At the end of the fourth year, the student should be able to design a process to include all of the economic, environmental, safety, and societal
aspects for functioning in today's marketplace.

Summer 2006 151











Double Degree Program with Bishop's University
Sherbrooke students may choose to earn a double degree in
engineering and liberal arts in ajoint program with Bishop's
University. The engineering program includes four, four-
month, paid-work internships. The remainder of the liberal
arts program is undertaken by taking selected courses at
Bishop's University, located a few kilometers away from the
University of Sherbrooke.
The liberal arts degree provides for a broad education in
the social and human sciences, allowing students to develop
fully as individuals in a chosen humanities specialty: history,
literature, philosophy, fine arts, or theater.
This exclusive program is the result of close cooperation
between the University of Sherbrooke and Bishop's Univer-
sity. It allows students to study in both of Canada's official
languages while also experiencing the two unique and distinct
university cultures.

Project-Based Learning Approach
The project-based learning approach is used to integrate
coursework within an academic session in first-year projects
and the third-year and final-year design project. Project-based
learning is a new approach in the educational field. It was cho-
sen by the department to fill new professional and industrial
requirements. The main features of our approach are:
The project involves solution of a problem taken from
a "real world" case. For instance, the last-year design
project was an oil sands exploitation situation; the first-
year project was the design of a process for the valoriza-
tion of lactoserum.
The student is responsible for this "on the job" learning
and the initiative is to be taken by the students. The class
is divided into small teams and collaborative learning
principles are used to find and share new knowledge. At
the end of these projects, teachers have often noted that
their students have improved appreciably in the applica-
tion of self-learning skills.
The project results in a "deliverable" (i.e., a process,
product, flow sheet, report). In the last year, the deliv-
erable (for the last-year project) was the design and
specifications for the oil sands plant; the end product (for
the first-year project) was the design and monitoring of a
pilot plant for lactoserum valorization.
The work lasts for a realistic amount of time. Each proj-
ect is spread over two sessions.
Professors involved are considered an advisory commit-
tee and the approach is student-centered. The teacher
team performs the "follow-up" each week and provides
advice, if necessary. The team realizes the assessment
of the project. The project is also evaluated by other
students.

Students subsequently present the project at public con-


ferences and the annual Canadian Chemical Engineering
Congress. During the last 10 years, Sherbrooke students have
won many awards at the Congress of the Canadian Society
for Chemical Engineering for the quality of their first-year
and final-year project presentations. These projects enable
students to substantially improve their mastery of both oral
and written communications.

RESEARCH
Research has been a very significant part of the department's
activities from its beginning. The department received fund-
ing of more than 4,000,000 CAD last year. This funding is
primarily in the form of grants, so there are fewer overhead
charges than American schools are likely to bear. This fund-
ing has been principally used to support graduate students
and researchers, as well as to build and maintain very well-
equipped laboratories.
Research in the department is conducted in a number of
areas, including aluminium production technology, biomass
conversion, biotechnology, environmental engineering, fuel
cell technology, flow modeling, and plasma technology.
The research has evolved over the years as a result of
recruitment of professors and substantial research funding.
In 1973, Bernard Coupal obtained a very substantial grant to
develop applications for peat moss during the first year of the
department's existence as an independent entity. This grant
provided a very important impetus to work in the department
and led to a very substantial and lasting work program on
biomass transformation.
Environmental engineering developed because we started
a master's degree program in environmental engineering
and science. Biotechnology research was developed as we
recognized the need for specialists in our new biotechno-
logical engineering program. The interest in this research has
progressed hand in hand with the hiring of new professors
working in the field.
Thermal plasma technology was not common in chemical
engineering departments before the 1970s but the arrival of
Maher Boulos led to very substantial growth in this area at
Sherbrooke. The need to model gas/plasma flows in plasma
torches led to general interest in the modeling of flow, heat trans-
fer and kinetics, and particle behavior in a variety of systems.
Fuel cell research evolved because of our specialized
knowledge, which already existed in the plasma research
center for materials, especially deposition on surfaces using
thermal plasmas.
Other research "specialities" were taken up because of
their importance to certain industrial sectors existing in the
Quebec economy, notably in aluminium production. The
pharmaceutical industry is very developed in Quebec, and we
are consequently developing very active research in this area,
including PAT (Process Analytical Technology).


Chemical Engineering Education












Testimonial of a doctoral student's experience
in a joint program with a university in Europe
"The Department of Chemical Engineering at the Uni-
versity of Sherbrooke provided me with an opportunity to
do my doctoral research in collaboration with a Belgian
laboratory at the Catholic University of Louvain (UCL).
This collaboration lies within the scope of an agreement of
joint direction established between the two universities.
"This European collaboration enabled me to work for
some two-and-a-half years within the Bioengineering Unit
at UCL. I discovered innovative enzymatic methods for
the elimination of recalcitrant phenolic compounds. This
experiment also allowed me to establish a network with
other European laboratories working in fields related to
my research task."


GRADUATE EDUCATION
The department has provided a strong graduate program
from its earliest days. Professors who were in the process
section of the mechanical engineering department had large
contingents of graduate students. Upon creation of the chemi-
cal engineering department, we were accorded, in addition
to our existing undergraduate programs, the master's and
doctoral programs.
There are presently approximately 50 post-graduate stu-
dents working in the department. We graduate about eight
master's degrees and five Ph.D. degrees per year.


INFLUENCE OF OUR PROFESSORS IN
CREATING HIGH-TECH SPINOFFS
The University of Sherbrooke generally and the Depart-
ment of Chemical Engineering in particular have been very
successful in obtaining licensing fees and creating spinoff
companies. Esteban Chornet, with the contribution of Nicolas
Abatzoglou, has created a number of companies, many very
successful in the tasks of coproducing useful products and
energy from biomass and organic wastes. Maher Boulos has
created a company, based in Sherbrooke (Tekna), with more
than 40 full-time employees, specializing in the area of thermal
plasma technologies. This company exports its products all over
the world, achieving particularly strong results in Japan.

THE FUTURE
The future is bright. We are recruiting large numbers of
students into the biotechnological engineering program. The
incoming class recruitment in chemical engineering is stable,
with approximately 35 new admissions per year. We are in the
process of recruiting additional professors for both chemical
and biotechnological engineering.
The University of Sherbrooke provides a distinctive edu-
cational experience because of its "French" character. We
encourage those predominantly English-speaking students
who also have knowledge of "basic French" to come to
Sherbrooke to simultaneously "perfect" both their chemical
engineering and their French language skills for use in their
future careers-in Canada, and beyond! 7


Winter visitors to
Sherbrooke will
see some of the
prettiest scenery
snowy climes
have to offer,
such as this view
of nearby St.
Benoit-du-Lac
Monastery.


Summer 2006











S usan educator




Susan Montgomery


of the


University of Michigan


ScoTT FOGLER AND LARA ZIELIN
The University of Michigan Ann Arbor, MI 48109


Undergraduate program advisor Su-
san Montgomery was three years
into a tenure track position at the
University of Michigan (UM) in education
research when she realized that teaching and
advising were her true passions. Driven by
those passions, Susan did the unconventional
thing and became a lecturer in 1999. Since
then, she has been a "mom" for over 1,000
ChE students, who appreciate the warm
and supportive community she helps create
within a big university atmosphere.
"Susan Montgomery literally holds
together the undergraduate curriculum
at UM," says Ron Larson, chair of UM's
chemical engineering department. "She is
the most appreciated faculty member among
the undergrads. The key element that makes
her successful is her singular focus on the
students and their needs. While other faculty
members also care deeply about students,
their research and administrative portfolios
limit the extent to which they can involve
themselves in the concerns of the students.
There is simply no substitute for having a
member of the faculty who is devoted ex-
clusively to the students."
This is no small feat in a ChE program
as large as UM's. Total enrollment hovers
around 350 students. I \1 can be daunting,"
explains Larson. "Even within the 'com-
munity' of a department, students can get
lost." Susan works hard to combat this by
maintaining a connection with each chemi-
cal engineering class. "Not only does she
know every undergraduate that comes up
through the program, she keeps track of them
as they move on to their future careers,"
says Larson.
154


While she loves the classroom and has taught a wide range of undergraduate
classes-from introduction to engineering to process design- Susan has a hard
time choosing a favorite topic. "The real fun is seeing students transition from
one phase of their careers to the next," she says. "The thrill is watching students
blossom."
Susan works tirelessly to make students' growth and advancement a reality.
In addition to her advising and lecturing, Susan is the principal author of the
Visual Encyclopedia of Chemical Engineering Equipment, a CD-ROM designed
to help beginning ChE students understand how chemical engineering equip-
ment works.
The CD-ROM stemmed from her research in the Multimedia Educational Labo-
ratory (MEL) at UM, which focused on studying the diverse learning styles of
chemical engineering students, and developing multimedia educational software
to address those learning styles. Susan then analyzed student use of this software
to discern what types of interactions were preferred by what students. The goal
was to help future educational software developers better understand the role that
different interactions could play in addressing the needs of a variety of learners.
Copyright ChE Division of ASEE 2006
Chemical Engineering Education












c I cCYCLONES / HYDROCYCLONES: Equipment Design


vCn

Co


Underflw


S iot


The picture shows the parts of a cyclone.
For more information on a given part, click
on.its name.
For a given particle size, a portion of the
eaCi v.ii eL' o, ^ bUsW e.i- 1. "e
iridFr .'&d ,. *-he ,lia.I- .?rm rhe
ovewflow Theheavier the particle, the
greater the chance that it emts out the
bottom
The V.umc o.'r'u m rNc.a 6-e aor CrloCre
oilocnr'anl"-. i i e a. r'li pait 1le i ,:t
,a- ciwrnio Tel a rj-, e pamc la, .xj' op
and half at the bottom. A coarser Oarger)
outpoint means that only larger particles can
be separated A finer (smaller) utpoint
means that smaller particles can be removed


-BACK c F o uoo FIIMPREVIO, F P oo, CA
BACK TO PREVIOUSMrNU J fe l^^! PAGE 0 OF H B


Figure 1. A screenshot from Susan's encyclopedia CD-ROM showing
an oversize view of equipment, left, and the main navigation page, right.


The CD-ROM includes animations and pictures of real
equipment as well as examples of applications of the equip-
ment. From the main menu, the user can branch into a variety
of different topics and operations, such as the corresponding
pieces of equipment for processes including heat transfer,
reactors, materials handling, and more.
The CD-ROM encyclopedia has been extremely well re-
ceived both in academia and industry. Figure 1 shows a screen
capture from the CD illustrating a sample introductory over-
view of an item of equipment-cyclones and hydrocyclones
in this instance. The possibilities for illustrating chemical
engineering practice are almost endless, and many of the
screens in the CD show dynamic operation of the equipment,
as in the case of bubble-cap distillation columns, screw ex-
truders, filters, and cyclones, to mention just a few. When a
visit to a chemical plant or refinery cannot be made, the Visual
Encyclopedia is an excellent substitute for illustrating various
operations that might otherwise have to be described more
passively. It has been used for numerous industrial training
courses. The CD is also included in two of the most popular
textbooks in the field: Rich Felder's Elementary Principles of
Chemical Processes; and Scott Fogler's Elements of Chemical
Reaction Engineering.
In addition to the Encyclopedia CD-ROM, Susan and her
colleagues have developed two other CDs. The first, titled
Engineering Fundamentals in Biological Systems, provides
real-world applications in fundamental processes, such as
material balances on an artificial kidney. The second CD,
Material and Energy Balances, provides interactive problem
solving in real-world environments including the car pre-
painting system in Ford Motor Company's Wixom Assembly
Plant, and Ann Arbor's wastewater treatment plant. The CD
also includes tutorials on Pxy-Txy diagrams, psychrometric
charts, and enthalpy-concentration diagrams. The CDs are
distributed through the CACHE Corporation.

Summer 2006


FAMILY TIES
Susan's background and upbringing have much to do with
her success in the field. She was born in Peru, of a Peruvian
father and an American mother. Her father is a civil engineer
who shared his passion for engineering with his daughters. He
used to take Susan and her sister to job sites, showing them
how things worked. Susan recalls going for a walk with her
dad when they chanced upon a street replacement project.
Her dad showed her all the layers that made up a street, and
they met the construction workers.
Susan and her family lived for a year and a half in Ni-
caragua, where her father was an Agency for International
Development consultant with their ministry of public works.
Forced to leave suddenly in 1978 when a civil war broke out,
Susan, her sister, and her mother moved to Ann Arbor, Mich.,
where her American grandmother lived.
The summer after her junior year in high school, Susan
attended a Women in Science and Engineering program at
Carnegie Mellon University and came home announcing
that she intended to be a chemist. Her father corrected her,
"reminding" her that she would be an engineer.
Susan graduated from high school at age 15, a feat she
attributes to an excellent kindergarten that allowed her to
complete first, second, and third grades in one year. She
completed her undergraduate work at UM, then went on to
Princeton University for her graduate work. Susan's advisor
at Princeton was Professor Ludwig Rebenfeld, of the Tex-
tile Research Institute, where her research focused on flow
through porous media. But her passion was in teaching, and
her goal was a faculty position.
Susan had doubts, however, about her ability to complete
the Ph.D. Thrice during her time at Princeton, she announced
to Prof. Rebenfeld that she was quitting the program. Prof.
Rebenfeld offered unwavering support, but two other factors


CHEMICAL ENGINEERING EQUIPMENT. Main Menu


Cl.eK i ec I o re car. one. beil to i cer i r ir. lT 'po i a. re P oI



PROCESS
PARAMETERS HEATTRANSFER REACTORS
PARAMETERS

,,FWMEERS SEPARATIONS: POLYMER
FLOWMET CHEMICAL PROCESSING

TRANSPORT SEPARATIONS: MATERIALS
AND STORAGE MECHANICAL HANDLING


QUIT HELP INDEX


---"I'''''















Right, Susan celebrating
her fourth birthday. Far
right, a beaming Susan on
her graduation day at
Princeton. Below, Susan
and her older sister Betsy
proudly display a snow-
man they "engineered."


made her tough it out and complete the program. The first factor was
knowing that without a Ph.D., she would not be able to fulfill her
goal of becoming a faculty member. The second was the scolding
of her grandmother, Margaret Hampshire, an independent woman
with whom she lived while an undergraduate at UM. Where oth-
ers, upon hearing that she was considering dropping out, offered
condolences, her grandmother replied with comments such as, "We
didn't work this hard for you to drop it all now. You get back there
and get that Ph.D.!"
Perhaps this is why Susan can now speak so well to the number
of women students who are struggling to find their place in the ChE
program. "After a month, they may come to me feeling like maybe
they're not good enough," she says. "I encourage them not to focus
on whether or not they're 'good enough' but on whether or not this
path is taking them where they want to go." The approach seems
to work. Susan says a number of alumni have thanked her for the
advice she offered early on in their academic careers. "They're out
there doing what they love now, so they can look back and be glad
they stuck with the program," she says.
Susan has the strength to help many, but when she needs someone
at her side she calls on her sister, Betsy Vera. Susan says Betsy has
stood by her continuously. They remain best friends and stay close,
even though Betsy now lives in Chicago.
"I have learned so much from my family," says Susan. For ex-
ample, she recalls watching her mother complete her undergradu-
ate studies in her 40s, and then go on to earn a master's degree in
Latin American Studies from Georgetown University. "Now that I
156


am in my 40s, too, I appreciate having my mother as a
role model for living the life you are meant to live," she
says. Susan also says her mother's death at the young
age of 52 solidified the importance of living your life
now, versus waiting until retirement to do things you've
always dreamed of.

ENTERING THE FIELD
Susan got her first job on a fluke. Like many under-
graduate students, she was struggling to find funding for
spring semester of her third year. She was just leaving
the dean's office, disappointed, when she ran into Scott
Fogler, a UM professor of engineering, whose class she
had just finished taking, and acing-she had been the
top student. He inquired about how things were going
and, upon learning of her situation, he immediately of-
fered her a summer job working in his laboratory and
helping him with his textbook. This marked the begin-
ning of more than 20 years of collaboration in various
educational projects.
Susan's first teaching position was as a TA for the
junior-level laboratory course at Princeton, where she
caught the bug for education. "I could see the light bulb
go off in students' minds," she says. An internship at a
local community college teaching pre-algebra at night
made her intrigued about learning styles, and the de-
Chemical Engineering Education












































velopment of learners through their college careers. She
was quite intimidated when asked to be responsible for
the whole laboratory course at Princeton the following
year, but that experience only cemented her decision to
become a faculty member.
During this time, Susan also recalls going to the engi-
neering library at Princeton to search for a research article
in Chemical Engineering Science. Instead, she discovered
Chemical Engineering Education, stacked right next to
the journals she was supposed to be reading. All plans
for the afternoon were scrapped as she spent long hours
browsing through CEE instead.
Also during her time at Princeton, Prof. Rebenfeld
supported her attendance at the 1990 ASEE National Con-
ference in Toronto, back in the days when few graduate
students attended the conference. Participants at that time
stayed in college dorms, which really helped colleagues
get to know one another better. "I always speak of this
conference as the time in which I 'found my people',"
she says. "Their dedication and passion for teaching
matched my own."
After graduate school at Princeton, Susan returned to
UM to complete a two-year postdoctoral appointment
developing educational software for chemical reaction
engineering and problem solving. Once again, she col-
Summer 2006


Left, Susan poses with her sons and her father at the Plaza de
Armas in Lima, Peru. Above, Susan (middle) and her sister, Bet-
sy Vera, flank Prof. Dale Briggs at a 1984 undergraduate gradu-
ation reception; to her left are her grandmother and mother.


laborated with her former professor, Scott Fogler, along with two
dozen outstanding chemical engineering students. It was these stu-
dents who taught her how energizing it can be to supervise teams
of undergrads.
Again, Susan experienced a sea change. "At the time, my plans
were to teach at a small undergraduate institution, or community
college," she says. "But my love of UM and Ann Arbor meant I
eventually accepted a tenure-track faculty position focusing on
academic research, the first in a chemical engineering department."
Susan's research focused on the use of multimedia to address diverse
learning styles, and once again she supervised teams of undergradu-
ate students in developing educational software.
Susan has a long list of accomplishments to her name. In addition
to those already mentioned, in 1994 she started a student chapter
of ASEE at UM, the third such chapter and one of the few long-
standing active chapters. The chapter has remained strong through
the years, creating real change in the culture and appreciation of
teaching through activities such as workshops and panels on engi-
neering education issues.
"I strongly feel that the ASEE student chapter changed the culture
of teaching and graduate student training at the College of Engineer-
ing," she says. "Many of the activities we organized, which centered
around preparing students for faculty positions, have been adopted
by the college." These activities include academic job-search work-
shops and panel discussions on working toward tenure.
To further prepare students, Susan regularly teaches a graduate
course, "Teaching Engineering," that draws 50 graduate students and
trains them for academic positions. The students learn to develop
syllabi and course materials, practice presentation and teaching
skills, and are introduced to different learning styles. They also
learn to deal with student issues that may arise. "This has become
157










an invaluable class for would-be future faculty members,"
says Sharon Glotzer, UM professor of chemical engineering.
"It attracts students from around the College of Engineering,
including postdoctoral students."
Susan's concern for the well-being of graduate students and
staff extends beyond the academic arena. Susan works to assist
students with both academic disabilities and psychological
issues. "Many times these issues manifest themselves during
college years," she explains, "and bad grades are some of the
early warning signs." To aid students, Susan says she tries
to remove any stigmas around the topic of mental illness by
sending out e-mail messages to students about depression,
educating faculty
about the issue, and
encouraging those
who need assistance
to seek professional
help. Her important
work has not gone
unnoticed: Recently
she was asked to take
part in a video titled
"Depression on Col-
lege Campuses."
In this sense, many
faculty think of Su-
sanas pioneer. "She
forges new ideas and
utilizes new resourc-
es to make the cur-
riculum more effec- Susan, right, with sons lan, 12,
tive," says chemical Canadia
engineering lecturer
Barry Barkel. And it's not just students who benefit from her
tireless work-so, too, do alumni. "Susan occupies the unique
position of being the primary focal point of the department for
both undergraduate students and alumni," says Barkel. "She
is the face of the department for many people."

BEYOND RESEARCH AND TEACHING
Susan's family members likely think of her as a pioneer, too.
She and ex-husband Sean Montgomery, whom she met when
they were both undergraduates at UM, have two boys-Ian,
12, and Nicky, 7,-whom she has taken on summer excursions
to places such as the Grand Canyon and the Canadian Rockies,
in keeping with her philosophy to take adventures now and
not wait until retirement. This summer, they will embark on
a trip to Peru with Susan's sister, Betsy, where they will visit
family members as well as journey to Machu Picchu.
Even when not traveling, life with two boys- and two cats,
Smokey andAten-is understandably very active. Both boys
are involved with karate and Ian plays on various team sports.
"We also enjoy going for walks and bike rides, and going


and
in Ro


canoeing down the Huron River," Susan says.
For Ian, watching his mom forge her own path in engineer-
ing may have inspired him. He has aspirations to one day be a
robotics engineer. Nicky thinks that might suit him as well-if
he doesn't make it as an NBA player first.
Susan has not forgotten her Hispanic heritage, and has in-
stilled this pride in her sons. Ian and Nicky's friends have come
to look forward to her alfajores--Peruvian treats she prepares
for any and all occasions. She also serves as faculty advisor to
the Society of Hispanic Professional Engineers, participates
in numerous sessions organized by the Minority Engineering
Program Office (MEPO), and recently started "Ingenieros,"
an informal Spanish
conversation group.
"Dr. Montgomery
has been an integral
part of the diversity ef-
fort here in the College
of Engineering," says
MEPO program direc-
tor Derrick Scott. "She
rarely turns down a re-
quest to participate in
our initiatives to attract
and retain underrepre-
sented minority engi-
neering students."
Susan is careful,
however, not to get
Nicky, 7, on a vacation to the toobusy. \lyboys are
Bckies. the loves of my life and
I want to share in every
aspect of their growing
up," she says. After missing Ian's first soccer goal while out
of town to be an ABET observer, Susan was determined that
she would forgo traveling for business for a few years. "Ron
Larson, the department chair, supported my efforts to create
a balance between my academic position and my family life,"
she says. Despite this, Ian loves to point out that Susan missed
his second soccer goal, which he scored while she was video-
taping Nicky and his friends playing on the sidelines.
She did bend her own rules and take one business trip in
2002 to the ChE division Summer School for ChE faculty in
Boulder, Colo. As Susan told the participants at the welcom-
ing session: "It doesn't matter what city you are in, if you are
surrounded by your friends and colleagues of the ASEE ChE
division, you'll always feel at home."
It's hard not to feel at home around Susan, whether in Colo-
rado, Michigan, or Peru. Her excellence, determination, and
love for people translates into every activity she performs.
"In short, every department needs a Susan Montgomery on
its faculty," says Ron Larson. "But they can't have ours." 7
Chemical Engineering Education











Mj =1 laboratory


AN AGITATION EXPERIMENT


WITH MULTIPLE ASPECTS








JORDAN L. SPENCER
Department of Chemical Engineering Columbia University, New York 10027


A gitation and mixing are important in a wide variety
of areas in both the traditional and modem process
industries,1 and it is appropriate that chemical en-
gineering students see related experiments in the unit opera-
tions laboratory. This paper describes a teaching experiment
involving both agitation and mixing that illustrates -using a
quite simple and relatively inexpensive apparatus a number
of aspects of this field. In particular the experiment involves
not only simple and direct measurements, for example of
torque and power as functions of stirring speed, but also more
sophisticated data-acquisition and processing methods, for
example indirect measurement involving the use of a model
and a parameter estimation method.

APPARATUS
The apparatus is shown schematically in Figure 1 (next
page). The major components of the apparatus are:
o A torque table, .. i1, of a 12-inch-diameter alumi-
num circle mounted on a tapered roller bearing set in a
12-inch-square aluminum base plate. Even with a load
of about 20 kg, the torque needed to set the upper plate
in motion is less than 0.00706 N m (1 oz inch). An arm
attached to the upper plate bears a load cell (Omega
Engineering, Model LCGC) connected to a panel
meter (Omega Engineering). The data are acquired by
a LabView program at 0.2-second intervals.


o A variable speed DC motor (Cole-Parmer) with speed
controller and torque indication. The speed can be
varied from 60 to 2400 RPM, and the maximum torque
is 45 oz-in (0.318 Nm).
o Two six-bladed turbines (of diameter 14.4 and 7.5 cm
and blade width 3.32 and 1.83 cm), and two three-
bladed propellers of diameter 14.4 and 7.5 cm. All are
mounted on 3/8-inch glass-epoxy shafts inserted in a
chuck on the motor shaft.
o A polycarbonate tank of ID 29.2 cm and volume about
18 L. This tank is equipped with four removable stain-
less baffles of width 2.43 cm, mounted on an acrylic
top plate. A stainless funnel is mounted near the top of
the tank, to be usedfor adding a conductive tracer, for
example NaCl or KCl at 30 g/L. A platinum electrode
conductivity cell is mounted near the bottom of the tank
180 degrees from the funnel. The cell is connected to a
conductivity meter (Amber Sciences) that sends a 0- to

Jordan L. Spencer is emeritus professor of
chemical engineering at Columbia Univer-
sity. He received his B.S. in 1953 and his
Ph.D. in 1961, both from the University of
Pennsylvania and both in chemical engi-
neering. His research and teaching interests
involve control and optimal control, and the
development of chemical engineering teach-
ing experiments, including Web-operable
experiments.


Copyright ChE Division of ASEE 2006


Summer 2006
































Figure 1. Schematic diagram of the apparatus,
showing baffled tank on torque table with load
cell, motor and impeller, conductivity probe,
funnel for tracer addition, and optical sensor
and lamp for bead detection.


0.30
0.25
E 0.20
"' 0.15
0.10
0.05
0.00
0 200 400 600
RPM


Figure 2. Torque as a function of turbine speed for w
18-liter tank. Also shown is a curve of the form Torm



60-
50
40-

0
20-
10 V
1
0 *
0 5 10 15 20 25
Frequency dimensionlesss)


Figure 3. Power spectrum of torque signal for 7.5-c
18-liter baffled tank at 800 RPM.


1-volt signal to the computer. A phototransistor (OPF-703) mounted
on the side of the tank 3 cm from the bottom is used to detect the
approach of almost neutrally buoyant 0.9 cm toroidal plastic beads.
The fluid near the sensor is illuminated by a 20 Watt desk lamp.

o A second polycarbonate tank of inner diameter 17.5 cm, also
equipped with four stainless baffles, and with volume about 4 L. All
dimensions of the smaller tank/impeller combination are propor-
tional to the dimensions of the larger tank/impeller combination,
with factor of 0.551.

o A heat transfer probe constructed by ., 1,, a 90 W resistance
heater in the center of a 5-cm-diameter by 7.5-cm-long aluminum
cylinder. A transistor-based temperature sensor (LM35-CAZ) is also
mounted, off-center, in the cylinder. The probe could be suspended
near the middle of the 20 L tank, about 5 cm from the wall.


RESULTS
Presented below are some typical results, with comments related to
what the results illustrate for the student.
Torque and Power
The 18-liter tank was filled with water to a depth of 25 cm. With the
baffles in place and the 7.5-cm-diameter turbine mounted in the motor, the
stirring speed was varied over a range of RPMs. The
torque was measured by the load cell and averaged
by a LabView program. Typical results are shown
in Figure 2, where the torque (N m) is plotted as
a function of RPM. Also shown is a curve of the
form Torque = k RPM2, which fits the data almost
perfectly, as expected.J1
Since the torque signal was found to vary with
time (with an amplitude about 10% of the average
torque), especially at high RPM where the flow in
800
the tank is quite turbulent, it was of interest to look
at the power spectra of the data. A typical spectrum
(Figure 3) shows that most of the power is at low
water in baffled frequencies, less than 0.15 cycles/sec. It is probable
ue: k RPM2. that the fluctuations reflect the presence of eddies or
vortices generated by the impeller.

Non-Newtonian and High Viscosity Fluids

In order to examine the behavior of a non-Newto-
nian liquid, the 4-liter tank (with baffles) was filled
to a depth of 15 cm with commercial ketchup. The
7.5-cm-diameter flat-bladed turbine was used to stir
the ketchup. At 200 RPM the surface of the ketchup
was stationary, clearly a non-Newtonian behavior.
At 400 and 800 RPM the ketchup flowed smoothly
at the surface. The data were well, but not perfectly,
30 fitted by a curve of the form Torque = k RPM2.
Data were also collected under the same conditions
using corn syrup (Karo), a Newtonian fluid with a
m turbine in viscosity about 2,500 times that of water. For these
runs the torque was a linear function of the RPM,
as expected.
Chemical Engineering Education










Tracer Studies of Mixing Times-Modeling
and Parameter Estimation

Mixing times have been estimated1, 2] by visual ob-
servation following the addition of a dye or conductive
tracer to the agitated liquid. This method, while satis-
factory in some cases, has the disadvantage that it is
inherently subjective, and cannot be used with cloudy
or opaque fluids. A more objective method is based on
acquiring and processing conductivity data following
the addition of a conductive tracer (e.g., NaC1 solution)
to the agitated fluid in the baffled or unbaffled 18-liter
tank. The data are acquired by a LabView program,
typically 100 baseline points in 20 seconds, followed
by 400 conductivity points in 40 seconds. The tracer
data are saved from Excel as a space-delimited (.prn)
file readable by a QuickBASIC parameter estima-
tion program.
The model used to fit the data, shown in Figure
4, consists of six well-mixed tanks. Three, of equal
volume, correspond to downward-moving fluid in the
core of the tank. The other three tanks (all of the same
volume) correspond to the fluid moving upward along
the walls of the tank. Symmetry around the propeller
shaft is assumed, so that only three shell tanks are
shown. Salt solution injection is assumed to take place
at the upper right, and a conductivity probe is located
at the lower left. The model contains two undetermined
parameters, denoted b1 and b,, with b1 the fraction of
the known tank volume in the three core tanks, and b2
the volumetric flow rate (L/s) downward through
the core tanks and upward through the shell tanks.
The volume of a core tank is b1V/3, and the volume
of a shell tank is (1 b1)V/3. An objective mixing
time is three times the tank volume, denoted V,
divided by b,.
Figure 5 shows the normalized conductivity vs.
time data for salt solution injection into the 18 L
tank, with agitation provided by a 7.5-cm-diameter
downward-driving propeller rotating at 60 RPM.
Also shown is the best-fit curve corresponding to
the model discussed above. The best-fit parameter
values were b = 0.484 and b2 = 1.368 L/s. The
conductivity does not rise for about five seconds,
corresponding to the time needed for the salt
solution to move from the wall of the tank to the
center, down to the bottom of the tank, and then
over to the conductivity probe located opposite the
injection point. Then the conductivity rises and
drops rapidly as the bolus of salt solution passes
the probe. After some further oscillations the
conductivity reaches a constant value. The curve
based on the six-pool model fits the data fairly
well, but certainly not perfectly. This is because
Summer 2006


Figure 4. Six-pool model of the flow pattern in the tank,
used for analysis of conductive tracer-injection data. The core
pools (tanks) represent the downward-flowing liquid
in the center of the tank, the shell pools represent the upward-
flowing fluid near the tank wall. Parameter b, is the fraction of
the tank volume V in the core, and b2 is the volumentric flow
rate (L/s) through the array of well-mixed tanks.
The state variables are the tracer concentrations
in the six tanks.





2.5

2.0
DATA
0 1.5-

S1.0-- v -"
0 BEST FIT



0.0 ,f ,
0 10 20 30 40
Time (seconds)



Figure 5. Conductivity data vs. time, and best-fit response from
model, for tracer injection into 18-liter baffled tank at 60 RPM using
7.5 cm propeller Parameters: b1=0.484, b2=1.368 L/s.


CORE
TANKS


SIX-POOL MODEL OF
FLOW PATTERN


SHELL
TANKS













At high Reynolds numbers the flow

in the tank is, at least in some sense,

strongly turbulent. This implies that

particles suspended in the tank

move in a chaotic and uncorrelated

way, and thus move independently.




the model is much too simple to correspond exactly to
the highly turbulent, three-dimensional, and complex
flow pattern actually existing in the baffled tank, even
at low RPM. The parameter b1 (the fraction of tank
volume) was 0.484, and b2 (the circulation rate) was
1.368 L/s. When the stirring speed was doubled to 120
RPM the conductivity rose much earlier, overshot less,
and settled out about three times more rapidly-cor-
responding to more vigorous mixing. The value for
the circulation rate parameter b2 was 2.765, more than
double that for the 60 RPM run.

Similitude
In theory, for tanks that are geometrically similar, at
the same Reynolds number the power numbers should
also be equal. As a test of this theory, the torque is
measured at 600 RPM using water and a flat-bladed
turbine in the 4-liter baffled tank, and the Reynolds
number is calculated. Then the torque is measured
in the geometrically similar 18-liter tank, at an RPM
corresponding to the same Reynolds number. When
the experiments were performed, the calculated and
measured torques typically agreed within 20%.

Heat Transfer Coefficient
It is well documentedO1 2] that heat transfer coeffi-
cients in agitated tanks depend strongly on the intensity
of agitation. In order to demonstrate this, the heat
transfer probe was immersed in water in the baffled
18 L tank. The power to the probe was turned on, the
stirring speed was set, and the probe temperature was
allowed to come to steady state, which occurred in a
few minutes. The difference between the probe and
water temperature was plotted as a function of RPM.
As expected, the temperature difference was highest
at 0 RPM, dropped monotonically as RPM increased,
and approached a nonzero constant as the RPM ap-
162


preached high values. This reflects the fact that the resistance to heat
transfer is the sum of a constant resistance due to the aluminum wall
of the probe, and a film resistance at the probe surface that varies
with the 2/3 power of RPM. From the high RPM asymptote the
first resistance can be calculated, and from a second point the heat
transfer coefficient can be found as a function of RPM. In general
the results were consistent with Eq. (1):


k = a N67 = aN2/3


where k is the heat transfer coefficient (Wm2 /C), a is a constant,
and N is the stirring speed (s 1).


0.50

m 0.40

0 0.30

) 0.20

zO 0.10
0
0.00
0 5 10 15 20
Time (seconds)


Figure 6. Optical sensor voltage for beads in 18-liter tank at
400 RPM. Each drop in voltage corresponds to the entry of a
bead into the field of view of the phototransistor. Only drops
below a selected voltage are counted as events.


6.0

'a 5.0


S4.0

S3.0

2.0
0.0 1.0 2.0 3.0 4.0
Interval Length (seconds)


Figure 7. Semilog plot of distribution of bead arrival intervals.
The straight line corresponds to a Poisson distribution.
Chemical Engineering Education










Note that all real-world measurements involve both signal and noise.

In most cases the information is contained in the signal and we

attempt to minimize the noise. But in some cases the noise

also contains useful information.

The results ... illustrate this.


Particle Dynamics
At high Reynolds numbers the flow in the tank is, at least
in some sense, strongly turbulent. This implies that particles
suspended in the tank move in a chaotic and uncorrelated way,
and thus move independently. In order to test this prediction
about 200 plastic toroidal beads of approximate diameter 0.9
cm were added to the tank. The 7.5-cm turbine was run at 400
and 800 RPM. A typical sensor signal is shown in Figure 6,
corresponding to a sample rate of 100 per second. Each drop
in voltage represents the entries of one or more beads into
the illuminated region. The program that acquired the data
also used simple logic to identify close-bead approaches, to
calculate the time interval between entry, and to construct
an interval distribution function. If the beads in fact move
independently, this is expected to be a Poisson distribution.
Typical results, shown in Figure 7 in a semilog plot, corre-
spond reasonably well to a straight line and thus to a Poisson
distribution. The slope of the line is related to the average
frequency of bead events, but also depends on the efficiency
of detection of bead approaches, which is not easily known.
The lower values at low time intervals probably correspond
to the fact that the data acquisition program is not able to
differentiate between two or more bead entries that occur at
almost the same time.
Note that all real-world measurements involve both signal


and noise. In most cases the information is contained in the
signal and we attempt to minimize the noise. But in some
cases the noise also contains useful information. The results
above illustrate this.
Salt Crystal Dissolution
When a crystal of a soluble salt, for example NaC1, sits in
water that is not moving, it dissolves relatively slowly. But the
opposite holds for a crystal suspended in strongly turbulent
water. Approximately one gram of a coarse (about 150 crystals
per gram, the crystals being of variable size) kitchen salt was
added to water in the baffled 18 L tank. At 0 RPM full dis-
solution required more than 60 minutes. The signal from the
conductivity probe recorded by a LabView program is shown
in Figure 8. Using the 7.5-cm-diameter turbine, dissolution
was almost twice as fast at 800 RPM as at 400 RPM. These
results demonstrate that the apparatus described above can
be used in a wide variety of studies of the effect of stirring
speed, impeller design, and other parameters on the rate of
dissolution of (or extraction from) solids, as long as the solids
release a conductive tracer.

CONCLUSIONS
The agitation and mixing experiment described above is
based on a relatively simple and inexpensive apparatus. But
it illustrates a number of aspects of the subject, including the





Figure 8. Conductivity
as a function of time
following addition of
1 gram of NaCl crystals
to the agitated
18-liter baffled tank
at 400 RPM.



60 80


Summer 2006


0.70


r 0.66


S0.62
0

0.58
0 20 40
Time (seconds)











dependence of torque on stirrer speed, impeller design, baffle
design, and nature of the fluid involved. The principle of
similitude can be tested. The experiment also illustrates very
general and more sophisticated concepts, including the use
of modem data acquisition software and nonlinear regression
methods to estimate the parameters of a model of the flow
pattern in an agitated vessel. The apparatus is well adapted to
studying the rate of dissolution of salt crystals, and the effect
of agitation intensity on heat transfer from a solid surface.


Finally, students are able to acquire and process essentially
stochastic data to obtain some information on the turbulent
flow in the vessel.

REFERENCES
1. McCabe, WL., J.C. Smith, andP. Harriot, Unit Operations of Chemical
Engineering, 4th Ed., McGraw-Hill, New York (1985)
2. Uhl, V.W, and J.B. Gray, Mixing, Theory, and Practice, Vol. 1, Aca-
demic Press, New York (1966) 7


Chemical Engineering Education










classroom


USING A COMMERCIAL SIMULATOR

TO TEACH SORPTION SEPARATIONS






PHILLIP C. WANKAT
Purdue University West Lafayette, IN 47907-2100


Since modern practice of chemical engineering uses
specialized process simulators extensively (e.g., Aspen
Plus, CHEMCAD, HYSIM, and PROSIM), chemical
engineering departments need to prepare students to use these
tools. For example, distillation columns are designed almost
exclusively using process simulators, and if the equilibrium
data is deemed reliable, the column will be constructed with-
out any laboratory or pilot data. Most chemical engineering
departments now use one of the steady-state process simula-
tors in separations and/or design courses.[1, 2]
The steady-state simulators do not include adsorption,
chromatography, and ion exchange (collectively, sorption),
which are normally operated as unsteady-state processes.
Formerly, sorption systems were designed by a combination
of data and rules of thumb. Recently, it has become more
common to use a more fundamental design procedure based
on solution of the partial differential equations governing
the heat and mass transfer in the column and the algebraic
equations for equilibrium and pressure drop. In industry, the
detailed simulations are always accompanied by laboratory
and often pilot plant data.
Chemical engineering graduates who understand the
fundamentals of sorption processes and are familiar with
sorption simulators will have a competitive advantage. This
paper discusses the use of the commercially available Aspen
Chromatography simulator to teach sorption separations. The
Summer 2006


course outline, grading procedure, assignments, computer
laboratory operation, and testing procedure are delineated.
Student survey results and the author's opinion of the effec-
tiveness of teaching with this simulator are presented.

THE COURSE
ChE 558, "Rate-Controlled Separation Processes," is a
three-credit, dual-level elective course that has been taught
off and on for almost 30 years.[3] The topics covered always
include sorption separations, and depending upon the profes-
sor, might also include crystallization, electrophoresis, or mem-
brane separations. I have used Rate Controlled Separations [4]
although this book is currently difficult to obtain. This course
has always been taught in a lecture style with homework and
often a course project.


Copyright ChE Division of ASEE 2006


Phil Wankat is a Distinguished Professor of
chemical engineering at Purdue University who
earned his degrees from Purdue and Princeton.
His technical research is in separation processes,
and he recently finished his textbook, Separation
Process Engineering, 2nd Edition of Equilibrium
Staged Separations, Prentice-Hall, 2006. He is
also co-author of the bookTeaching Engineering,
available free at edu/ChE/Newsand Events/publications/teach-
ing engineering>.











Three considerations led me to change the teaching method.
First, since I believe that the sorption separation processes
will eventually be designed almost entirely using simula-
tors, proper preparation of graduates will require teaching
with simulators. Second, the understanding of an average
ChE 558 student was too low. Since I had observed student
improvement in a distillation course when a simulation lab
was incorporated, 21 I expected an increase in understand-
ing if a similar change was made in ChE 558. Third, I had
proposed in the educational part of two NSF proposals to
teach ChE 558 with a simulator, and now I had to deliver
on these promises.
In spring 20051 changed ChE 558 to focus entirely on sorp-
tion separations. The nominal schedule had a one-and-a-half-
hour lecture on Tuesdays and a one-and-a-half-hour computer
laboratory using the Aspen Chromatography simulator on
Thursday (see Table 1). This schedule had fewer lectures on


sorption separations than in previous years, but tests covered
the same amount of material on these topics. The total amount
of material in the course was reduced by removing the mem-
brane separation material, which is now often included in the
required undergraduate course on separations. The course was
taken by four undergraduates and three graduate students.
Only one of the students had previous experience with an
unsteady-state simulator, but all had previous experience
with Aspen Plus, which has a somewhat similar graphical
user interface to Aspen Chromatography.
The grading scheme used a straight scale (85-100 = A,
75-85 = B, 60-75 = C, 50-60 = D) as guaranteed grades, but
I reserved the right to use lower cut-offs if that was appropri-
ate. The two regular tests were each 25% of the grade, the
lab exam was 20%, lab attendance 9%, lab assignments 6%,
homework 5%, and the group course project was 10%. Stu-
dents were encouraged to work together on lab assignments


TABLE 1
Schedule ChE 558, Spring 2005. Readings are from Reference 4.

Date Class Room Subject Reading
T, Jan 11 1 110 Intro. Adsorption & Chromatography 207-228
Th, Jan 13 2 111 Lecture -Adsorption: thermo/phys. prop. flow; start solute movement 228-251
T, Jan 18 3 110 Lecture Solute movement 239-251, 296-305
Th, Jan 20 4 111 Lab 1 Intro to Aspen Chromatography Skim 268-274
T, Jan 25 5 110 Solute movement/thermal effects-focusing 251-268
Th, Jan 27 6 111 Lab 2 Chromatography/adsorption basics 288-296
T, Feb 1 7 110 Heat & Mass Transfer, local equilibrium solution 268-277, 296-305
Th, Feb 3 8 111 Lab 3 Convergence
T, Feb 8 9 110 Chromatography Linear solutions 305-316
Th, Feb 10 10 111 Lab 4 Chromatography 316-321, 336-347
T, Feb 15 11 110 Chromatography Linear solutions 316-331, 334
Th, Feb 17 12 111 Lecture Constant pattern and scaling 365-393
T, Feb 22 13 110 Plateaus & Nonlinear behavior, start MB and SMB 393-400, 521-533
Th, Feb 24 14 111 Lab 5 Thermal effects 405-412
T, Mar 1 15 110 Test 1
Th, Mar 3 16 111 Lab 6 Flow reversal systems 405-418
T, Mar 8 17 110 Moving Beds and SMB; review test 499-537
Th, Mar10 18 111 Lab 7 TMB and SMB 521-533
SPRING BREAK
T, Mar 22 19 110 Ion Exchange 452-484
Th, Mar 24 20 111 Lab 8 Ion exchange 475-481
T, Mar 29 21 110 Ion exchange 475-491
Th, Mar 31 22 111 Lab 9 LAB EXAM
T, Apr 5 23 110 PSA/Gas separation 400-418, 421-438
Th, Apr 7 24 111 Lab 10 Lab demo -ADSIM PSA Aspen Chromato. obtaining data from article Read article
T, Apr 12 25 110 PSA/Gas separation 421-431
Th, Apr 14 26 111 Lab 11 -Project
T, Apr 19 27 110 Lab 12 -Projects
Th, Apr 21 28 111 Test 2
T, Apr 26 29 110 Work on projects
Th, Apr 28 30 111 Lab 13 Project reports and demos

166 Chemical Engineering Education












FIle Edt View Tods Flsheet Run Window Help
D 0WIaI, tt 11 14n IM Diynamic j| R mII1 |6 9
7TrroxWid


5 1mulation
I Flowsheet
I Chromatography
I ystemLibrary
Component Lists


jChromatography
SComponent Lists
RFlowsheet
SGlobals
iOpen Library
oSolver Options
g5ystemLibrary


I1


r n 7 ldbin 100o5 J k v I g


A/ O A Ao 0- \ %J1 IEJ


-limes.N.ewRoman -zJ10 -J1B I U


I
Equation S2.AM_If6 AMTrue AMIf7 moved from ELSE branch to THEN branch
Equation 51.AM_If6 AM True AMIf7 moved from ELSE branch to THEN branch
Equation B3.AM_If3 AMTrue AM_If4 moved from ELSE branch to THEN branch
Equation B2.AM_If7.AM_True.AMIf8 moved from ELSE branch to THEN branch
Simulation has 1223 variables, 898 equations and 4181 non-zeros
_JlI


Figure 1. Screen-
shot of Aspen
Chromatography
Interface with
flow sheet for
a simple
chromatography
system.


Ws artl LP W [E U Lneacujlty-listJ Ed Re... I J College of Engineering... M crosoft Word I L Microsoft PowerPoint -...


spend Chromatogr... I 10:11l


and homework. The complete course syllabus is available
from the author at .
Homework assignments were problems from the textbook
plus one straightforward simulation. The textbook problems
were similar to the test problems; of course, new problems
were written for the tests. Since the students were all able to
come to class early, they were given two hours for each test.
Unfortunately, due to a mistake in solving an ion exchange
problem on the second test, this problem, although solvable,
was about an order of magnitude too difficult. I adjusted
scores based on the performance of the second-best student
in the class (the best student appeared to be an outlier whose
performance was not representative of the class). The students
appeared to be satisfied with the fairness (or generosity) of
this procedure.

ASPEN CHROMATOGRAPHY COMPUTER
LABORATORY
Aspen Chromatography is an algebraic-differential equa-
tion-solving program with a user interface for the solution of
liquid adsorption and chromatography problems (see Figure
1). This simulator is very powerful and a trained user can
often solve in a few hours a problem that used to take months.
Aspen Chromatography uses the method of lines to solve the
partial differential equations. The user can select both the
differencing method to be used and the integration method
Summer 2006


to solve the resulting ordinary differential equations. Aspen
Chromatography licenses are expensive for companies, but
are reasonably priced for universities and can be bundled with
other Aspen Technology programs. It cost $400 to add an As-
pen Chromatography license for 60 users to Purdue's Aspen
Technology order for University Lifecycle Package Bundle
#1 (60 users) that cost $2,000. The current Version 12 is quite
stable and reasonably user friendly, but not as user friendly as
Aspen Plus. My experience with Aspen Plus is that 98-99% of
the difficulties students have are due to operator error. With
Aspen Chromatography about 80% of the students' difficul-
ties are caused by operator error. As expected, the numerical
integration routines, which use the method of lines to solve
the partial differential equations, have difficulty converging
when the profiles are steep and the isotherms are nonlinear. In
general, the resources and expertise that have been developed
for teaching with steady-state simulators1, 2 5] are not avail-
able for sorption separations. More troubleshooting and more
computer assistance will be needed.

Since much of my current research involves simulation
of chromatography and simulated moving-bed systems with
Aspen Chromatography, I am familiar with this simulator and
my graduate students are very familiar with it. The graduate
students and post-doc supported by the NSF grants were en-
listed to help with the computer laboratory. With their aid, I
developed 10 laboratory assignments including a laboratory
167












test. Each of the first eight laboratories showed how to build
a flow sheet for a new aspect of Aspen Chromatography in
a cookbook fashion, and then had the students solve simula-
tion and design problems. Excerpts from the first laboratory
assignment are presented in Table 2. All lab assignments are
available from the author at .

As the semester progressed the amount of detail in the in-
structions was decreased. Most of the students stayed in the
lab after the nominal closing time to finish the take-home as-
signments that accompanied the labs. The material covered in


TABLE 2
Excerpts from First Lab Assignment
A complete set of instructions for all labs is available from
.

The goal of this lab is to get you started in Aspen Chromatography. It
consists of a cookbook on running Aspen Chromatography and some
helpful hints. We will also simulate a real separation. Keep this lab
assignment. You will want to refer back to it.
1. Log in to the computer. Go to Start, Programs, ChE Software,
AspenTech, Aspen Engineering Suite, Aspen Chromatography 12.1,
Aspen Chromatography This opens a window if you are at a station
that allows you to access the hard drive. Otherwise, you will get a
message that essentially says, 'The working folder is unavailable."
In this case, change working file to your N drive. Click on OK, and
window should open. If not, run in circles, scream and shout, and ask
for help.
2. We will first develop a simple chromatography (or adsorption)
column system. To do this, go to the menu bar and on the left side,
File. Click on File and go to Templates, and in that window click on
"Blank trace liquid batch flowsheet," and click on Copy. It will ask
for a file name. Use something like columnnl" This will be saved in
your working file. NOTE: In all file names and names for compo-
nents, columns, steams, and so forth there must be NO spaces.
3. In the "Exploring simulation" box (LHS), click on "component
list." Then in box below (Contents) double click on Default. This lists
A and B. Change these names to the names of the components to be
separated (fructose and dextran T6). First, click "Remove all" button.
Then in window below type in first component name (e.g., fructose)
and click on "add" button. Do the same for all other components.
Then click OK.
4. Now draw the column. Click on the + to the left of "Chromatog-
raphy" in the 'Exploring Simulation" box to open other possibilities.
Click on the word "chromatography "This should give "Contents of
Chromatography" in box below. Double click on the model you want
to use (Reversible since it is most up-to-date). Click and drag the
specific model you want: in this case "chrom_r_column," and move
to the center of the Process Flowsheet Window. This gives a column
labeled B1. Left click on B1, then right click to open a menu. Click
on Rename. Call the block something like "column. "Click on OK.

19. If you have time, do this next step. If not, save your file (remem-
ber the file name), exit Aspen and do this step outside of class. The
two peaks are not completely separated. There are a number of ways
they can be separated more completely. Double the value of
L, to L = 50 cm. Click on Rewind, change L in the column dimen-
sions table, and then rerun the one-minute pulse input. When you run
pure solvent, a pause time greater than 10 minutes is needed since
doubling column length will double time for material to exit. Do this
run and look at the result. Separation is better, but still not complete.
Print your plot and label it. This plot will be handed in with the lab
assignment. Save your file (remember the file name) and exit Aspen.


TABLE 3
Handout on What to Expect in Lab Exam
The exam will be open book and open notes. You may not open or
use any of your old Aspen Chromatography files.
Part A. (50 points) Generic Problem. This is a demonstration that
you can do a basic Aspen Chromatography simulation. Open up
Aspen chromatography and use a "Blank trace liquid flowsheet tem-
plate. Set up a chromatographic column with one feed, a column,
and a product. Use specified models for the column, feed, product,
and connecting streams. Set up the system to process compounds that
will be specified in the test. Have Aspen do discretization with xyz
procedure with NN nodes (these will be specified in the test). Use a
model with convection plus a specified form of dispersion, constant
pressure, and velocity. If needed, the dispersion coefficients will
be supplied. Use a linear, lumped parameter model with a specified
driving force and constant mass transfer coefficients (they will be
given). The isotherms will be given and the units for q and c will be
specified. Operation is isothermal. The column length and diameter
will be given. The adsorbent has the following properties: ee =0.aa,
Ep = 0.bb, KD = 1.0, Qs = cc kg/m3. The following feed values will be
specified: flow rate, pressure, and all component concentrations. Use a
specified integrator with a specified fixed or variable time. Use default values
of the tolerances. Develop a graph of the product concentrations (on the same
scale) versus time.
1. Run a breakthrough curve for zz minutes. Print, label, and turn in
your plot. Use the history to accurately determine the center of the
breakthrough curve and the t.mz for one of the components where tMrz
is measured from 0.05 times the feed concentration to 0.95 times the
feed concentration. These calculations should be shown on the plot.
2. Input a dd-minute feed pulse and develop with pure solvent for a
total time of zz minutes. Print your plot, label, and turn in.
There should not be any convergence problems in Part A.
Part B. (50 points) The second part of the lab test will be a design
problem for one of the other processes that we have studied (e.g.,
flow reversal, adiabatic operation, SMB, TMB, ion exchange).


TABLE 4
Homework Assignment 4
1. Use the Lapidus and Amundson solution with Eefftve to predict the
behavior of fructose in a column packed with silica gel. The feed is
50 g/liter, the feed pulse lasts for eight minutes, and then it is eluted
with water. The flow rate is 20 ml/min. The other values are:
Value Units Description
L = 200.0 cm Length of adsorbent layer in
column
D = 2.0 cm Internal diameter of column
ee = 0.4 m3 void/m3 bed Inter-particle voidage
Ep = 0.0 m3 void/m3 bed Intra-particle voidage
dp = 0.01 cm particle diameter (needed to
find E.,,w)
ED = 0.15 cm2/min Constant Dispersion
Coefficient
Lumped parameter with concentration driving force.
k a = 5.52 1/min Constant mass transfer
m, P p
coefficient
Isotherm is linear
K' = 0.69 dimensionless Isotherm parameter (q and c
both in g fructose/liter)
2. Solve problem 1 using Aspen Chromatography
3. Compare your solutions for problems 1 and 2 at the peak center
time predicted by the local equilibrium solution, peak center time
minus four minutes and peak center time plus four minutes.

Chemical Engineering Education










the laboratory was cumulative, and by the end of the semester
the students were able to simulate rather difficult problems
without detailed instructions.
Part A of the lab test was a demonstration by the students
that they had learned how to use Aspen Chromatography for
simple simulations. Two weeks before the test the students
were given the generic form of part A (Table 3). They were
encouraged to supply data and parameter values to generate
their own form of the test and then practice solving it. Part
B, a design problem, proved to be more difficult. The lab test
was given during a normal lab period that was extended to
two hours. Since there were only seven students in the class
and I knew them all well, no special precautions beyond
proctoring the exam were taken to ensure honesty. (When
I gave an Aspen Plus lab test in a core |


junior class with 95 students, I wrote
a different test for each of the five lab
sections and disabled both e-mail and
access to student files.)
During the 10th lab, students first
watched a computer demonstration of
the use of ADSIM for pressure swing
adsorption. Gas separations can involve
large changes in flow rates which are not
modeled by Aspen Chromatography.
Then the students did a simulation with
Aspen Chromatography that required
them to determine the parameters need-
ed for the simulation from a literature
paper. In the earlier labs the students had
been given all the necessary parameters
since that makes troubleshooting of stu-
dent difficulties much easier. Students
were told that the purpose of learning
how to extract parameters from the
literature was to prepare them for the
course project.


The course project was to develop a
new Aspen Chromatography problem and solution suitable
for one lab period. This is a form of Felder's generic quiz.J61
Students were required to use equilibrium and mass transfer
data from the literature and/or the Internet, not from the text-
book or from Aspen Chromatography demonstrations. They
were told that projects that considered operational methods
not taught in the lab or that combined different operational
methods would be most impressive. Student groups presented
an oral report, including a computer demonstration, and turned
in a written report. As a treat for the students, I ordered pizza
to be delivered after the oral reports were presented. The
student projects -nonisothermal ion exchange, ion-exchange
with flow reversal, and SMB separation-were quite well
thought out.
Since seven students do not divide evenly into groups, I
Summer 2006


divided the class into groups of 3,3, and 1. 1 used this unusual
procedure because one of the graduate students is doing his
thesis research with me and during the course of the semester
he had much more practice with Aspen Chromatography than
the rest of the class. He agreed to be a group of one, and the
class accepted my rationale when the groupings were pre-
sented. The other two groups were made as equal as possible
based on grades in the course.

EXAMPLE PROBLEM
Students solved a number of chromatography and ad-
sorption problems during the semester. The real strength of
numerical analysis is it can solve problems with complicated
nonlinear isotherms that cannot be solved analytically. To
| avoid the "black box" effect, benchmark-


ing of numerical solutions with analytical
solutions was done for linear problems
where analytical solutions exist. One
convenient analytical solution is the
Lapidus and Amundson solution41] with
an effective dispersion coefficient that
includes the effects of dispersion and
mass transfer.'71 Homework assignment
4 (see Table 4) illustrates benchmarking
of analytical solutions. This assignment
requires students to solve a simple, sin-
gle-component chromatography problem
with a large pulse of feed by the Lapidus
and Amundson method and numerically
with Aspen Chromatography.
The effective dispersion coefficient
that lumps all dispersive effects into
axial dispersion and assumes negligible
mass transfer resistance was estimated to
be 8.062 cm2/min. This is much greater
than the axial dispersion coefficient value
0.15 cm2/min because mass-transfer re-


distance controls dispersion. The Lapidus
and Amundson solution requires the use
of superposition as a step up followed by a step down eight
minutes later.
The same problem was solved numerically with Aspen
Chromatography using two of the higher-order differencing
schemes, Buds (Biased Upwind Differencing Scheme, a 4th-
order method) and QDS (Quadratic Differencing Scheme),
and the default UDS1 (Upwind Differencing Scheme 1)
with 50 nodes. The solutions all used the Gear method with
a fixed time step for integration. The QDS solution was done
first with the actual value of the mass transfer coefficient
and axial dispersion coefficient, and then with a very high
mass-transfer coefficient (essentially no resistance) and the
effective dispersion coefficient.
A screenshot of the Aspen Chromatography solution using
169


Aspen Chromatography
is an algebraic-differential
equation-solving program
with a user interface
for the solution
of liquid

adsorption and
chromatography
problems. This simulator is
very powerful and
a trained user can often
solve in a few hours
a problem that
used to take
months.











Buds with 200 nodes is shown in Figure 2 and a screenshot
of the solution using UDS 1 with 50 nodes is in Figure 3. The
Lapidus and Amundson solution and the higher-order nu-
merical solutions were bell-shaped curves and looked almost
identical. UDS1 with 50 nodes also produced a bell-shaped
curve, but it is much more spread out and has a lower peak
concentration than the other curves because of significant
numerical dispersion. The curves are different enough that
students can easily see the differences by comparing Figures
2 and 3. Thus, the use of UDS 1 with 50 nodes is numerically
inappropriate for this problem.


Since the Lapidus and Amundson and the
higher-order numerical analysis curves are so
similar, differences can only be ascertained
by looking at exact values of concentrations
and times (Table 5). The concentrations
predicted by the Lapidus and Amundson
solution are: t = 25.575, c = 25.0 g/liter;
t = 29.575, c = 48.54 g/liter (peak maxi-
mum); and t = 33.575, c = 24.975 g/liter. The
Lapidus and Amundson solution has its peak
center at exactly the time predicted by the
local equilibrium solution (29.575 minutes).
The peak concentration, peak time, and the
predicted times for concentrations of 25.0
and 24.975 g/liter are given in Table 5 for
the five different solutions. Since the two
QDS solutions are quite close to each other,
the use of an effective dispersion coefficient
is valid for this linear system. All of the rea-
sonable solutions (excluding UDS 1) are quite
close, with a small shift in times. Although
the Buds solution with 200 nodes is the best
fit to the analytical solution, in practical
terms it doesn't matter which is used. One
of the lessons students learn from this and
other benchmarking exercises is that they
must pay close attention to numerical
convergence.

RESULTS
A survey on the computer laboratory was
developed, and a research exemption was
obtained from the Purdue Institutional Re-
view Board for Human Subjects Research.
The students all responded to the survey
(Table 6) on the last day of class. To avoid
biasing any of the responses, the survey was
administered by the undergraduate secre-
tary; I was not in the room while the students
filled out the survey, and the process was
completed before the students knew there
would be a pizza delivery.


The students' responses to the survey (Table 6) show that
previous knowledge of different computer applications varied
from no knowledge to comfortable. General comfort levels
with computers were high. With the exception of the speed
of the Distributed Academic Computing System (DACS),
which allowed remote access to Aspen Chromatography,
laboratory operation was rated as about right. The students
thought that both the computer labs and the lectures helped
them learn sorption separations and that combining lecture
and lab was an appropriate way to teach this material. Most
of the comments are positive and reinforce the advantage of


Fibe Vi To Run Wind Help
an Ial ai m a coI l Dynamic J.I ..I II NI I4YW l .IiaNfaFaB |

Fructose Concentration


15 20
Time Minutes


Ready Paused | l ocl IDynac at40.0o f4.0 Mnutes |
Figure 2. Screenshot of Aspen Chromatography solution for problem 2 in
Table 4 using Buds with 200 nodes.

File -ew Too- Run windo HI*
g 6a1r I l an| cn I W Dynamic Dj > .ii M I...ll. I M MiEM
1" ... \
Fructose Concentration















10 15 20 26 30 35 40 45
IPaud | cal ynm cat 45.0 of 5.0 Mnutes
Figure 3. Screenshot of Aspen Chromatography solution for problem 2 in
Table 4 using UDSI with 50 nodes (an inappropriate choice).

Chemical Engineering Education


25 30 35 40












having a computer lab. At the same time
they filled out the survey, the students
Comparison of Solutions for Problems in Table 4 e ed t the sta ard cure evalu
responded to the standard course evalu-
L & A Aspen Chromatography Solutions action questionnaire required in all ChE
Soln. Buds QDS QDS (100 nodes) UDS1
200 nodes 100 nodes [Ez=E, 100 nodes courses. Course evaluation questions that
MTC=100,000] ask for global ratings correlate positively
with student learning.181 These core ques-
Peak time 29.575 29.6 29.5 29.5 28.7 tions were, 1. "Overall, I would rate this

Peak conc. 48.54 48.63 47.84 47.87 34.30 course as:" and 2. "Overall, I would rate
this instructor as:" The choices were:
Time, min 25.575 25.47 25.43 25.38 24.90 Excellent=5, Good=4, Fair=3, Poor=2,
@ upward curve, c=25.0 and Very Poor=1. The scores obtained

Time, min 33.575 33.44 33.43 33.38 32.46 for these questions-4.1 and 4.6, respec-
@ downward curve, c= 24.975 tively -collaborate the impression that


TABLE 6
ChE 558 Computer Laboratory Survey
(The average values and comments in italics are based on student responses.)
I. Computer experience before taking ChE 558. Rate your experience with the following applications (name package used where asked) using the following scale:
1 = Never used it before 558. 2 = Knew a little about it before 558. 3 = Used it some before 558. 4 = Was comfortable with it before 558.
Avg
Spreadsheets .................... ........... ................. .. .......... 2 ............... 3 .............. 4 .............4 .0
Intern et ....... ... ..... ... .... ..................... ............. 2 ............... 3 ..............4 .............4 .0
D A C S ........................... ....... ..................1 ............. .2 ...............3 ..............4 .............2 .3
A spen C hrom atography ................................................ ..............2 ...............3 ..............4 .............1.1
A spen Plus ....................................... .. .... .. ... ..... ...... ..... .. ... 2 ............... 3 .............. 4 .............3.1
O their steady-state sim ulator ......................................... ..............2 ...............3 .............. 4 .............1.6 ...........Pkg? P ro III
{M athlab, M athcad, M aple,M athematica} ..................1 ..............2 ...............3 ..............4 .............3.1 ........... Pkg? M athematica4.M atlab5
D E Q -algebraic eqn solver ............................................ ..............2 ...............3 .............. 4 .............1.0 ...........Pkg?
D ata B ase ........................................ ................ ...1 ...........2 ...............3 .............. 4 .............2.0 ...........Pkg? A access 3
Statistical package ........................................................ ..............2 ...............3 .............. 4 .............2.9 ...........Pkg? JM P 3. C crystal B all I
Program m ing languages) ............................................ ..............2 ...............3 .............. 4 .............2.1 ...........Pkg? FOR TRAN 1. C ++ 1.C 2
O th e r ............................................. ........................ ..............2 ...............3 .............. 4 .............1 .0 ...........P k g ?
II. Computer comfort level. Rate your comfort level with the computer:
1 = Uncomfortable 2 = Neither comfortable nor uncomfortable 3 = Reasonably comfortable 4 = Very comfortable
Avg-
General comfort level using computer before class ......1 .............2 ..............3 ..............4 .............3.7
General com fort level using com puter now .................1 .............2 ..............3 ..............4 .............3.7
Comfort level using Aspen Chromatography now ........1 .............2 ..............3 ..............4 .............3.4
Comments:
III. Computer Laboratory Operation. Please circle the appropriate response.
Avg.
The computer speed with direct installation (without DACS) was: ........... 1. slow ...........2. about right ........3. fast .................. 2.1
Com puter speed using DA CS was: .............................................................. 1. slow ...........2. about right ........3. fast ....................... 1.4
The laboratory assignm ents were: ................................................ .................. 1. too long......2. about right ........3. too short ................ 2.0
Computer lab should be scheduled for: .................................... ................. 1. less time.....2. same time .........3. longer time ............ 2.0
The assistance available during lab from the graduate student and the professor was:
1. inadequate...2. adequate.............3. very good................2.4
Comments: On one survey the term "graduate student" was underlined.
IV. Learning. Please answer these questions with the following scale:
1 = Strongly disagree 2 = (Between 1 & 3) 3 = Neither agree nor disagree 4 = (Between 3 & 5) 5 = Strongly agree.
Avg.
The computer labs helped me learn adsorption and chromatography ............................................. ... 1 2 ...3 ...4 .....5 ............... 4.7
The lectures and homework on the theory helped me learn adsorption and chromatography ................ ..... 2 ...3 ...4 .....5 ...............4.9
The format of ChE 558 (combining lecture and computer laboratory) is appropriate for this subject ... 1 .....2 ...3 ...4 .....5 ...............4.6
Comments: "Because of the complexity of solving chromatographic problems, being able to see what actually happens in a column was quite nice." *
... ., Tues./Thurs. schedule worked great!" "Without lab a lot of material would be lost" "More classroom time would be helpful to reinforce some
material"
V. Suggestions for improving 558 computer lab:
"Run the simulations before the students run them." "More labs with more lab time, cover a little more material."



Summer 2006 171











the students thought this was a good course.
I believe the students learned sorption separations in more
depth in spring 2005 than in previous years. This seemed to
be true across the spectrum of student abilities (good students
learned more than good students previously, average students
learned more than average students previously, and struggling
students learned more than struggling students previously).
Since in previous years the course also covered membrane
separations, the breadth of coverage was less in 2005; how-
ever, the students learned sorption operations better despite
less lecture time spent on this topic. Obviously, the 2005
students are also prepared to use the simulator.

DISCUSSION AND CONCLUSIONS
The students generally liked the format of lab and lecture
and thought it helped them learn; however, these students
all volunteered to take this elective knowing there would be
a computer lab. Students who feel uncomfortable using the
computer probably took other electives.
During the semester a faulty installation of Windows caused
difficulties running Aspen Chromatography in the computer
laboratory. For several weeks the students needed to log into
DACS, which was slower than the direct installation. Once the
problem was identified and Windows was reinstalled, we had
no difficulties with the direct installation of the software. The
comment in Table 6, "Run the simulations before the students
run them," probably referred to this difficulty running some
of the labs on DACS. Lab 6, with flow reversal, ran without
problems when I tested it using the direct installation of Aspen
Chromatography in my office, but would not run on DACS.
The student group that later did its course project with flow
reversal had no difficulty following the original lab instruc-
tions and obtaining solutions with a direct connection. It is
important to have reliable computer support before scheduling
use of any simulator.


If there are transferable skills in learning how to use simula-
tors, students who become skilled with, for example Aspen
Plus, will learn to use another simulator faster. This appeared
to be true for Aspen Chromatography. Thus, even if they never
use simulators taught in the curriculum, the experience of
learning to use these simulators will probably help graduates
efficiently learn to use simulators on the job.

ACKNOWLEDGMENT
The assistance of Nadia Abunasser, Jin-Seok Hur, Weihua
Jin, and Dr. Jeung-Kun Kim is gratefully acknowledged.
The computer personnel in ChE-George Bailey and Eric
Pratt-were extremely helpful in making the computer lab
run smoothly. This project was partially supported by NSF
grants CTS-02112008 and CTS-0327089. This paper was
presented orally at the AIChE Annual Meeting, Cincinnati,
November 2005.

REFERENCES
1. Rockstraw, D.K., "Aspen Plus in the ChE Curriculum," Chem. Eng.
Ed., 39(1), 68 (2005)
2. Wankat, PC., "Integrating the Use of Commercial Simulators into
Lecture Courses," J. Eng. Ed., 91(1), 19 (2002)
3. Wankat, PC., "An Elective Course in Separation Processes," Chem.
Eng. Ed., 15(4), 208 (1981)
4. Wankat, PC., Rate-Controlled Separations, Kluwer, Amsterdam (1990)
(There were earlier printings by Elsevier and Blackie. Kluwer is now
part of Springer.)
5. Seider, WD., J.D. Seader, and D.R. Lewin, Process Design Principles,
Wiley, New York (1999)
6. Felder, R.M., "The Generic Quiz: A Device to Stimulate Creativity
and Higher-LevelThinking Skills,"( ..... I i.' 19(4), 176 & 213
(1985)
7. Dunnebier, G., I. Weirich, and K.U. Klatt, "Computationally Efficient
Dynamic Modeling and Simulation of Simulated Moving Bed Chro-
matographic Processes with Linear Isotherms," Chem. Eng. Sci., 53,
2537 (1998)
8. Centra, J.A., Reflective Faculty Evaluation, Jossey-Bass, San Francisco,
(1993) 1


Chemical Engineering Education












Random Thoughts...






HOW TO TEACH


(ALMOST) ANYBODY (ALMOST) ANYTHING



RICHARD M. FIELDER AND REBECCA BRENT
North Carolina State University Raleigh, NC 27695


It seems it's no longer enough for you to teach about the
Navier-Stokes equations and potential flow past sub-
merged objects. The ABET coordinator says that students
in the fluids course have to be taught oral communications too,
and the department head got inspired at some workshop and
now wants to teach critical thinking in every course, includ-
ing fluids. You argued at the faculty meeting that it's all you
can do to get through fluids in the fluids course but got little
sympathy, and it looks like there's no way out of it.
You probably have some questions at this point, like, (a)
Exactly what are those skills I'm supposed to teach? (b) Can
they be taught (as opposed to you either have them or you
don't)? and (c) How? For an answer to (a), we invite you to
check out an article we wrote on learning objectives, teach-
ing strategies, and assessment methods that address ABET
Outcomes 3a-3k.1 The answer to (b) is, yes. This column
suggests some answers to (c)-how do you enable students
to develop and improve a targeted skill, whether ABET-re-
lated or not? While we don't guarantee that the techniques
we'll recommend will always work for all students, we're
confident the results will be good enough to satisfy ABET
and your department head, and-as long as your expectations
are realistic-you.

1. R.M. Felder and R. Brent, "Designing and Teaching Courses to Sat-
isfy the ABETEngineering Criteria," J. Engr. Education, 92(1), 7-25
(2003), pdf>. If you're not an engineering educator or you are one and just
got backfrom the latest Mars expedition, let us explain that Outcomes
3a-3k are specified attributes engineering students in accredited
programs should have by the time they graduate. They include the
usual technical abilities but also such things as communication
skills, the ability to work effectively in multidisciplinary teams, and
an understanding of the professional and ethical responsibilities of
an engineer.
2. N.E. Gronlund, How to Write and Use Instructional Objectives (6th
Ed.), Upper Saddle River, NJ: Prentice-Hall, 2000. See also R.M.
Felder andR. Brent, "Objectively Speaking," Chem. Eng. Ed., 31(3),
178-179, 1997, html>.


> Write detailed learning objectives and let the
students in on them
Learning objectives (or instructional cli,.. il',. i are
explicit statements of what students should be able to do to
demonstrate that they have learned what you want them to
learn.2 The objectives must specify directly observable actions
(list, explain, calculate, derive, model, critique, design .. .).
Verbs such as "learn," i-. i,,," "understand," and "appreci-
ate" are unacceptable. You can't see students understanding
something; to know whether or not they understand, you have
to ask them to do something you can see that demonstrates
their understanding. For examples, see p.; il-l. /,. ,- '5"i, in nJ, 2 if>, a study guide containing a
subset of the learning objectives for the introductory chemi-
cal engineering course. Even if you don't know the course
content, you should be able to convince yourself that if the
students can do everything on those two pages, they have


Richard M. Felder is Hoechst Celanese
Professor Emeritus of Chemical Engineering
at North Carolina State University. He is co-
author of Elementary Principles of Chemical
Processes (Wiley, 2005) and numerous
articles on chemical process engineering
and engineering and science education,
and regularly presents workshops on ef-
fective college teaching at campuses and
conferences around the world. Many of his
publications can be seen at edu/felder-public>.
Rebecca Brent is an education consultant
specializing in faculty development for ef-
fective university teaching, classroom and
computer-based simulations in teacher
education, and K-12 staff development in
language arts and classroom management.
She codirects the ASEE National Effective
Teaching Institute and has published articles
on a variety of topics including writing in un-
dergraduate courses, cooperative learning,
public school reform, and effective university
teaching.


Copyright ChE Division of ASEE 2006


Summer 2006










probably learned what the instructor wants them to learn in
that part of the course.
Our first recommendation is to write detailed learning
objectives and give them to the students as study guides
for exams and as guidelines for other assignments such as
project reports and oralpresentations. Make sure your objec-
tives cover all the skills you would like students to master,
especially high-level skills (such as critical and creative
thinking) and the ABET-mandated outcomes that have not
been traditionally addressed in engineering courses, such as
communication and lifelong-learning skills. When you are
explicit about your expectations, the likelihood that your
students will meet them goes up dramatically, especially if the
expectations involve difficult or unfamiliar material.2

> Teach skills before you assess them
Take problem formulation as an example-another one of
those ABET outcomes few of us ever thought about before
they showed up in the Engineering Criteria. Suppose an
objective in your fluids course states that the students should
be able to make up (and solve) fluid dynamics problems
whose solutions call for creative thinking. If you simply
assign students to do that, most won't get what you're after
and you'll see mainly problems that look like, "Given X and
Y, calculate Z." That shouldn't surprise or disappoint you,
since nobody ever taught them how to do what you're now
asking them to do.
If, however, you first explain and illustrate what you're
looking for, then show several good and bad examples and
have the students work in small gi',cit\ in class to critique
them, then give and grade two assignments that include the
same task and perhaps show some submissions from students
who .. -i il,.. idea, you'll start seeing creative problems.3 Doing
all that would allow you to check off both problem formula-
tion and critical thinking on the list of outcomes addressed in
your course. You can do the same thing for, say, writing techni-
cal memos (communication and critical thinking), analyzing
workplace case studies (professional and ethical awareness),
or critiquing an engineering-related article in the popular press
(professional awareness, communication skills, understanding
the societal impact of engineering solutions, knowledge of
contemporary issues, and lifelong-learning skills).

> Use rubrics for grading
Problem-solving and multiple-choice tests are relatively
easy to grade, and if the grading system is rational, students
should have no trouble understanding what they did wrong
and why they got the grade they did. The same is not true of
written project reports, essays, case study analyses, and oral
presentations. When students just get grades and a few writ-
ten comments as feedback, they may understand why they


were marked down but may have little idea about how to do
it better next time.
There's a better way. Use a rubric to grade ,w ilIt,,. that
involves li.,. i, ,.ji.1,. n .. ., ill,, ..i, t of the grader. Decide
on criteria you will use to evaluate the memo, report, or pre-
sentation (e.g., technically accurate, complete, appropriately
documented, well organized, well written, good visuals,
sound theoretical analysis .. .); assign weights to each crite-
rion; and-for a four-point rating scale-briefly summarize
the attributes of a 1, 2, 3, and 4 for each criterion.4 Then,
when you give students illustrative written products or oral
presentations to critique in class, have them use the rubric
individually and then compare their ratings, and then share
your i.i, ,i. with them. When you hand back their assign-
ments, give them your completed rubric as well-and watch
how they improve on the next assignment. As with learning
objectives, when a grading system is clear, the students are
more likely to meet expectations.

> If a skill is important to you, assess it
Once you've communicated your learning objectives and
assessment method for a particular skill, and you've pro-
vided examples and practice in applying the skill, then (and
only then) is it legitimate to test the students' mastery of the
skill-and at that point, you definitely should. Engineering
students barely have enough time to keep up with their assign-
ments; they don't have time to dig deeply into everything in
all of their courses. If they are sure that something requiring
effort on their part to learn won't count toward their grade,
most won't bother to learn it. That's not laziness-that's ra-
tional behavior. The assessment drives the learning: if a skill
is important to you, assess it.
Here's our challenge to you: (a) Pick a problem-solving or
professional ("soft") skill that your students have always had
trouble mastering; (b) write one or more learning objectives
that list the things you might ask students to do to demon-
strate mastery of that skill; (c) share the objectives with the
students -if possible, as part of an exam study guide; (d) give
your class examples of what you're looking for and several
opportunities to practice the skill, in and out of class; and
(e) assess the students' mastery of the skill (using a rubric if
subjective judgment is involved) by asking them to do some
of the things specified in the objectives. If you see better
performance than you've ever seen (which we always have
when we've done all that), consider making this strategy a
permanent addition to your teaching toolkit. 7

3. R.M. Felder, "On Creating Creative Engineers," Eng. Ed., 77(4),
222-227(1987), gineers.pdf>.
4. You can see an illustrative rubricfor evaluating studentpresentations
at .


All of the Random Thoughts columns are now available on the World Wide Web at
http://www.ncsu.edu/effective_teaching and at http://che.ufl.edu/~cee/
Chemical Engineering Education











curriculum
-0


HYPER-TVT: DEVELOPMENT AND IMPLEMENTATION OF

AN INTERACTIVE LEARNING


ENVIRONMENT





MARINA SANTORO AND MARCO MAZZOTTI
ETH Swiss Federal Institute of Technology Zurich CH-8092 Zurich, Switzerland


Rapid advances in information technology and easy
access to the Web have motivated students and educa-
tors to use more and more Internet and multimedia
technology for educational purposes. Educators and students
are challenged by the great potential of the Internet for de-
livering and sharing a large amount of information among a
greater public, and by the possibility of creating alternative
and breakthrough ways of teaching and learning by using
advanced software. Today, in almost every field of educa-
tion, a broad range of e-learning material is available online.
"Computer-based learning system" has been the catch phrase
for the last few years to indicate a wide area of systems using
the Web and multimedia technology for education.
These systems can be classified depending on the main
functionality and research focus as: computer-aided educa-
tion systems; multimedia/virtual laboratory; distance-learning
systems; or intelligent tutoring systems.1l This classification
identifies different levels of e-education: from the simple
integration of computer technologies into the traditional
teaching system-where face-to-face meetings and personal
relationships are still of primary importance-to the stage
where the information transfer rate is no longer managed by
a human tutor but it is adaptively controlled by an intelligent
computer system.J1 In all cases the objective is always to use
the technology not just for technology's sake, but to enhance
the quality of higher education. This contribution aims to
present Hyper-TVT, a public interactive learning environment
on separation process technology ().
The name Hyper-TVT derives from the word "Hypertext" and


the German acronym of the course name, i.e., "Thermische
VerfahrensTechnik," meaning thermal separation processes.

MOTIVATION
In recent years, the educational concept has been extended
toward the integration of new generations of Internet- and
computer-based courses intended to lower the gap between
theoretical knowledge and practical experience in the mod-
em curriculum for process and chemical engineering. This
concept addresses the nowadays frequent request of students
for continuous, adequate preparation (i.e., solving practical


Marina Santoro is a chemical engineer at the
ETH Swiss Federal Institute of Technology
Zurich, in Switzerland, specializing in sepa-
ration process technology, with particular
emphasis on the development of the new
learning technologies and of e-learning tools.
She has developed the e-learning system for
chemical engineers, Hyper-TVT.


Marco Mazzotti is professor of process
engineering at ETH Swiss Federal Institute
of Technology Zurich, in Switzerland. He
received his B.ChE. and his Ph.D. from
Politecnico di Milano, in Italy. He teaches
unit operations and mathematical methods
and has published more than 100 scientific
papers on separation processes.


Copyright ChE Division of ASEE 2006


Summer 2006










problems and facing real issues), thus allowing them to be-
come better aware of the physical reality of the processes, the
industrial world, and their future profession. Likewise, Lewin,
et al., claimed that the "instruction of chemical engineers
should reflect the challenges they face in industry."[2] Indeed,
teaching and learning in the field of chemical engineering can
be enormously enhanced by the use of information technology
and Web tools. In comparison to traditional textbooks, these
new avenues can offer videos, animations, and interactivity
that help students visualize the reality behind equations and
diagrams, thus compensating for the lack of practical experi-
ence in common engineering curricula.
In that respect, a very good reference for an e-learning en-
vironment is, in our opinion, the one developed by Fogler and
Gulmer in the field of chemical
reaction engineering.[3 The Web
site is offered by the University
of Michigan and is freely ac- u
cessible online. Besides texts,
equations, and diagrams, this
Web site provides interactive A *
tools for self-assessment, vid-
eos, and audio descriptions to
demonstrate the concepts of
chemical engineering, helping ing
users visualize the applications
within the industrial world.


In the field of separation
process technology as well, all
top-ranking technical universi-
ties provide e-learing systems
and online courses in an effort
to extend traditional chemical
engineering classes. Access


Figure 1. Illustration of the
respect to traditional ec


to these modules, however, is
very often restricted to enrolled
students of the course or to faculty,[4 5] and a lot of material
is available only in closed environment via CD-ROM or
password-protected LAN.[61 In other cases, the courses are
freely accessible but a lot of the online material provides
only course information and assignments, 71 or provides the
course syllabus as pdf files or PowerPoint presentations.
Although this static material is of high quality and has the
benefit of being always available online, a very low degree
of interactivity, or none at all, is offered and therefore this
approach exhibits only minor advantages as compared to
traditional textbooks.
A lot of interesting material covers only some specific
topic of the separation processes, such as distillation. Com-
prehensive theory, pictures, schemes, diagrams, and videos
are published online by research groups at universities and
companies working in the field.8 111 Finally, other universities
and organizations are active in the e-learning field, provid-
176


ing databases and links to facilitate searches for material,
organizations, communities, journals, and publications about
e-learning technology, such as ChemEngInfo,'21 World Lecture
Hall,13l EuPaCE.net,[141 and MERLOT.["15
The class on separation process technology is compulsory
in most curricula in process and chemical engineering. It
demonstrates the application of chemical engineering prin-
ciples within an industrial context for effective design of
processes, particularly of multistage separation processes.
Topics covered are: fundamentals of separation processes;
absorption and stripping; flash evaporation; distillation; and
liquid-liquid extraction. These subjects are particularly suited
for the development of e-learning tools due to the synergy be-
tween theoretical issues and practical experience. Hyper-TVT
is a freely accessible e-learning
system for students of chemical
and process engineering at the
ETH Swiss Federal Institute of
Technology Zurich, and also for
all individuals and institutions
involved or interested in process
and chemical engineering educa-
tion and practice. Its purpose is to
complement the already existing
tools, and, at the same time, to
compensate for their limitations
in access and scope, and to take
full advantage of interactive
and multimedia technology.
Therefore Hyper-TVT is not
comparable to commercial pro-
cess simulators such as HYSYS.
plant, CHEMCAD, or Aspen
role of Hyper- TVT with
lte of H er th Plus, which have different scopes
education methods.
and purposes.
Hyper-TVT is a didactic sup-
port, which is complementary to, but not a replacement for,
traditional lectures and textbooks. In fact, combining Hyper
TVT with conventional education elements-e.g., lecture,
textbook study, and discussion with the teacher-contributes
to enhancing teaching effectiveness and flexibility. The Web
site offers both educators and students a number of tools such as
videos, animations, simulations, and a self-assessment environ-
ment, thus creating a better balance between synchronous and
asynchronous approaches in teaching and learning, as illustrated
in Figure 1. The design concepts, the didactic content, and the
technical features of Hyper-TVT will be discussed in detail in
the following chapters.
DESIGN CONCEPT
"Textbook" structure
The structure of the Web site is organized in chapters and
paragraphs, as it would be for a traditional textbook. Figure
2 is a screenshot of the page content of the "Contactors"
Chemical Engineering Education











lesson as an example of page structure. The choice
of minimizing any structural complexity has two
purposes: to let students and educators focus bet-
ter on the Web site content, and to make using the
tool as easy as possible. Within each paragraph the
topic is presented using explanatory text, images,
and interactive diagrams and schemes. Every para-
graph develops a certain concept independently and
completely by integrating the multimedia material
within the text and by providing logical links to other
pages of the learning environment.

Navigability

Due to its simple structure and the implementa-
tion of a navigation system, the Hyper-TVT Web
site is easy to browse. In every section of the Web
site, menus help answer three fundamental ques-
tions[16]: Where am I?; Where have I been?; Where
can I go? Within every lesson, an interactive table of
contents highlights the user's current location within
the lesson, helps to find other related content, and
suggests the logical learning sequence, as shown
in Figure 3.

Interactivity

The added value of a Web-based educational system i
interactivity offered to students and educators. The H


Hyper- hermische Verfahrensfechnlk


sm &J


I Lessons | Unks Downloads Calculator Contact us Your feedback Search
I-Contactors: industrial equipment

S1. Gasb-Lquud contractors ...rh
S2 Gas-Liquid contractors: Tray columns MIA eld,
* 3 Gas-LUquid contractors: Packed columns
3.1 Description
3.2 Packing typ
3.3 Companson packed columns vs.tr ay columns
S4 LUquid Llquid contractors


Prof


ure 2. Structure of the "Contactors" lesson with chapters
paragraphs. A screenshot ofa video on column internals
is shown on the bottom right.

TVT system stimulates interactivity through pictures, anima-
tions, a modeling and simulation environment, videos, and
self-assessment tools, all of which are very consistent with
the learning objectives in the domain of process and chemi-


Hyper- hermische Verfahrenstechnik


flat2


_r_


ILQaLLUI -CIrrCU LUIUIU III


a uctued picking





Sb.Packingsupport




Sd-Randwwn packing
dlE


Figure 3.
Packing types
in the
"Contactors"
lesson. On
screen the
path is
indicated by
orange-
highlighted
items in the
menus.


The type of packing materials in industrial use can be distinguished into random
packing (also called dumped) and structured packing (also called ordered,
stacked or arranged).


Summer 2006


ml-Gas-Liquid C

3.2 Packing types


c.lquid dit tbulor










cal engineering. Two
examples of animations
about distillation topics
are shown in Figure 4. c
Reusability Recycle D
Recycle DxL
Hyper-TVT teaches
some of the key separa-
tion processes technolo- F
gies, i.e., a fundamental Fr T
subject in the process
and chemical engineer-
ing curriculum that is
well established. For this M
reason, the content of or Z
Hyper-TVT is not ex- MimoBsX.
pected to be subjected
to revision in the near
future, if ever.
Therefore the Hyper-
TVT Web site is avail-
able, now and in the fu-
ture, to students, educa- Figure 4. Interactive Flash
tors, and practitioners of
different institutions, countries, and educational sectors,
higher education, vocational training, professional organ
tions, industry, and schools.
Outreach to other categories of users
The target users of Hyper-TVT consist of all individ
and institutions involved in process and chemical engineer
education. Its simple structure, links, and interactive modi
however, make Hyper-TVT easy to be followed by oth
who, though not chemical or process engineers, still nee
want a quick and complete overview of the separation proi
technologies.

DIDACTIC CONTENT AND METHOD
Hyper-TVT consists of seven lessons presenting class
trial using text, images, animations, interactive tools,
simulation environments. These are:
1. Introduction of separation processes
2. Fundamentals of separation processes
3. Contactors
4. Absorption and stripping
5. Flash evaporation
6. Distillation
7. Liquid-liquid extraction

The first lesson is a presentation of the separation proce;
and their role and importance in the industrial context. In
second lesson, basics of thermodynamics and mass tran
are revisited; these are fundamental for further understand
of multistage separation technologies. Lesson number th
178


J Pressure
r Temperature
U Equilibrium stage


NumIe agses. 6


animations within the "Distillation" lesson.


on contractors, pres-
ents an overview of
the industrial equip-
ment for gas-liquid
and liquid-liquid
separation. Lessons
four, six, and seven
cover the design of
three of the most im-
portant multistage
separation processes.
The fifth lesson, on
Flash evaporation,
introduces many
concepts useful to
understanding distil-
lation. The didactic
approach first pro-
vides students with
all basic concepts
and tools needed.
Then, students are
challenged to use
the new material to


solve problems given as home assignments. Besides the les-
sons, three additional sections of the Web site contain videos
that can be streamed, PowerPoint and pdf files that can be
browsed, and tests that can be used for self-assessment. The
videos focus on industrial equipment for separation processes
(i.e., lesson three) and have been partly created for this use
and partly provided by companies, e.g., Sulzer ChemTech,
Kithni, and FRI (Fractionation Research, Inc.). In the videos,
an off-screen narrator guides the visitor into a virtual tour of
the real equipment to observe directly and in detail phenomena
that neither words nor photos alone would be able to clarify.
A screenshot of a video about column internals is shown in
Figure 2. Videos can have a great didactic value, and not just
in the field of chemical and process engineering, because
they overcome physical barriers and bring the world into the
classroom. The PowerPoint and pdf files are a collection of
mathematical derivations of the equations used in the lessons
for process design. Their format allows students to use them
interactively online or to download them for further reading.
The test section provides multiple choice and descriptive
questions. Students also have the option to submit completed
questionnaires, receive support online, and access and print
homework assignments as pdf files. Another important ele-
ment of the Hyper-TVT Web site is its database. The database
contains five specific search categories, e.g., text, notation,
images, glossary, and bibliography. The "text" category
provides links to pages where keywords are mentioned.
"Notation" contains detailed explanation of all the symbols
used in the Hyper-TVT Web site. "Images" is a collection of
diagrams and graphics already present in the lessons, but it
Chemical Engineering Education











Feedback from the ETH Zurich students indicates 95% appreciated the

learning environment and found it useful, both during lectures and outside

the classroom ....


also contains additional pictures, schemes, and photos of real
equipment provided by various companies. The "glossary"
category is a compact dictionary of terms specific to the les-
sons presented in Hyper-TVT. Finally, the "bibliography"
provides a list of suggested textbooks, handbooks, and further
readings. All five categories can be accessed by a search tool.
This allows visitors and students to make quick searches in a
specific area of interest and find direct links to the Web site's
relevant section.
A further aspect of the Web site is its toolbar. This provides
three additional links: to some of the most important chemical
engineering companies in the field of the separation process
technology; to the required-plug-ins Web page; and to an
evaluation section for online feedback and comments.

TECHNICAL FEATURES
Hyper-TVT is an independent learning environment in
which technical tools and software have been chosen not only
to suit the pedagogic needs of the class on separation processes
but also to facilitate its future management, maintenance,
and upgrade. Access is also possible from limited-band con-
nections, and compatibility has been guaranteed for all main
browsers on both PC and Mac machines. The architecture of
the Web site is based on the PHP scripting language, which
allows for creation of dynamic pages and the use of databases,
particularly of MySQL. Javascript has been used to build
the navigation menus and the interactive tools, i.e., the table
of contents and the self-assessment environment. Animated
diagrams, schemes, and interactive presentations have been
shown with Flash, Real Player, and QuickTime. Each has also
been used for video streaming. Finally, MathML (version 1.0),
has been used to edit and display the mathematical equations.
At ETH Zurich, Hyper-TVT has been one of the first educa-
tional projects to use MathML, thus providing a useful and
successful experience for other projects as well.

DISCUSSION AND CONCLUSION
The development of Hyper-TVT started in July 2001 and
its realization has taken three-and-a-half years with one per-
son working on it full time. A prototype of the Web site was
released to students in June 2002, and their feedback has been
used during revision of the site and completion of the project.
The whole environment is continuously revised based on feed-
Summer 2006


back collected in different ways. The ETH students attending
the class are asked to fill out an evaluation form at the end of
the semester, and an online evaluation form allows students
and educators from other institutions to provide helpful com-
ments. On one hand, these have been used for modifications,
refinements, and improvements of some parts of the Web
site. On the other hand, and more in general, feedback from
ETH Zurich students indicates 95% appreciated the learning
environment and found it useful, both during lectures and
outside the classroom, as support material and in completing
homework assignments. The Web site is used very intensively
in preparation for the final exam. The major advantage, as in-
dicated by students, is the interactive and audio-visual content,
i.e., the flash animations, videos, diagrams, and pictures. As a
further positive comment, students and practitioners underline
the easy navigability offered by the Web site. Very positive
reactions have been received from educators of other institu-
tions and technical universities as well. The result, although
possibly not yet statistically relevant, is very encouraging
about the usefulness of the Web site.
Use of the Web site is not a requirement of the separation
processes class for students at ETH Zurich. Rather, the Web
site is presented at the beginning of class as additional support
and its interactive material and videos are used during some
of the lectures. In this way, students get a first impression of
the real world of separation process technology while gaining
familiarity with the Web site, thus making its use easier. The
purpose of this approach is to increase student interest about
the proposed subject with more motivating material and tools,
thus stimulating their self-study skills and responsibility.
The most difficult issue we have experienced as instructors
has been integration of the Web site during traditional lectures.
In fact, this implies an allotment of time in the lecture plan
for use of the computer-based didactic material. This is not a
simple task since the traditional course has to be reorganized
and restructured to implement a new hybrid methodology (tra-
ditional and computer-based lecture). It requires an additional
effort and a lot of motivation by the instructor. For students
as well the integration of the Web site into their traditional
way of learning is challenging. Many studies have been con-
ducted to determine what factors impact the perception and
acceptance of new e-learing tools on students. 17 Such tools
generally require students to become more independent and











more responsible of their personal learning processes, and
to apply new technologies-an additional workload that is
not always well received at first. With the time and guided
assistance of the tutor, however, students usually recognize
and appreciate the benefits of the computer-integrated edu-
cational system. Typically, their learning efficiency and
performance improve, resulting in increased self-confi-
dence and greater motivation.
The Hyper-TVT system is available online ( per-tvt.ethz.ch>) and can be accessed without restrictions
through the Web.
Hyper-TVT shows how dissimilar pedagogic methods- tra-
ditional and Web-based-can be implemented in parallel to
offer a more stimulating and productive learning environ-
ment. Students, in turn, gain tools for self-paced learning
and become more autonomous. Use of Hyper-TVT can foster
their skills in analysis, synthesis, and evaluation. Finally, the
lecture time with the instructor can be invested more effec-
tively to discuss advanced issues. Our experience in using
Hyper-TVT has been very positive so far, and we encourage
students, educators, and chemical and process engineers to
explore it as well.


ACKNOWLEDGMENTS

Hyper-TVT has been developed within the Filep program


of ETH Zurich, which the authors thank for funding and
support.

REFERENCES
1. Shin, D., E.S. Yoon, K.Y. Lee, E.S. Lee, "A Web-Based Interactive
Virtual Laboratory System for Unit Operations and Process System
Engineering Education: Issues, Design, and Implementation," Comp.
& Chem. Eng., 26, 319-330 (2002)
2. Lewin, D.R., WD. Seider, andJ.D. Seader, "Integrated Process Design
Instruction," Comp. & Chem. Eng., 26, 295-306 (2002)
3.
4.
5.


6.
7. 2005/CourseHome/index.htm>
8.
9.
10.
11.
12. htm>
13.
14.
15. rowsecat=101>
16. Nielsen, J., Designing Web Usability, New Riders, USA (2000)
17. Platteaux, H., "How Students Perceive e-Learning Situations: The Case
of the SVC Embryology Course, "Proceeding of the 5th International
Conference on New Educational Environments, Lucerne (2003) 7


Chemical Engineering Education











curriculum
--- ^K__________________________-0


INTEGRATING BIOLOGICAL SYSTEMS

in the Process Dynamics and Control Curriculum




ROBERT S. PARKER
University of Pittsburgh Pittsburgh, PA 15261
FRANCIS J. DOYLE III
University of California at Santa Barbara Santa Barbara, CA 93106
AND MICHAEL A. HENSON
University of Massachusetts at Amherst Amherst, MA 01003


The discipline of chemical engineering is evolving, as
evidenced by the recent wave of departmental name
changes that reflect both the increasing number of
chemical engineering faculty involved in research on biology-
oriented topics, and the fact that the percentage of chemical
engineering undergraduates obtaining initial employment
with companies in the biotechnology and biomedical sectors
increased from 4.6% in 1998 to 10.3% in 2001-02.11] A series
of MIT-organized and NSF-sponsored workshops examined
the current state of undergraduate chemical engineering
education and recommended a sweeping set of changes.[2]
Foremost among the proposed changes were the introduction
of biology as a core science, the importance of addressing
complexity, and the expanded use of the systems approach.
The present discussion focuses on these three elements within
the context of the traditional process dynamics and control
curriculum.
The dynamics and control course, typically taught late in the
junior or senior year, is a natural point for including biological
systems content along with chemical process material. Due to
the focus on general principles rather than specific processes,
biological systems can be integrated without detracting from
the coverage of more traditional applications. This expanded
vision of the system dynamics and control curriculum requires
the following difficult issues to be addressed: (1) how can
these complex systems be introduced in a meaningful way
to undergraduate chemical engineers with little background
in biology?; and (2) what changes are required to include
biological content without sacrificing the traditional core of
process dynamics and control? The objective of this paper is
to provide some practical answers to these questions using the
experiences of three courses taught at our respective institu-

Copyright ChE Division ofASEE 2006
Summer 2006


tions. The first two examples illustrate the introduction of
biological content into the traditional process control course,
while the third example focuses on the development of a new
course in which the systems approach is applied to a diverse
set of biological problems.

Robert S. Parker is an associate profes-
sor in the Department of Chemical and
Petroleum Engineering at the University
of Pittsburgh. His educational interests
focus on the area of dynamical systems
analysis and control. He is currently
involved with the implementation of an
integrated curriculum and the develop-
ment of cross-cutting biological prob-
lems to assist students with integrating
material across courses.
Francis J. Doyle III holds the Duncan
and Suzanne Mellichamp Chair in
process control in the Department of
Chemical Engineering at the University
of California at Santa Barbara. He holds
joint appointments in Electrical Engi-
neering as well as in the Biomolecular
Science and Engineering Program. His
educational interests are in process
dynamics and control, systems biol-
ogy, and the Introduction to Chemical
Engineering course.
Michael A. Henson is a professor of
chemical engineering at the University
of Massachusetts, Amherst. His edu-
cational interests are in the areas of
process modeling and control. He is
involved in a variety of educational
initiatives including development of a
cross-disciplinary biological-systems-
engineering curriculum and participation
in a CACHE task force on systems-biol-
ogy education.











INTEGRATION OF BIOLOGICAL SYSTEMS
CONTENT
A typical process dynamics and control course covers a
broad range of new material at a rather brisk pace. To produce
students who can apply traditional dynamic analysis and
controller design techniques is a formidable challenge even
when the focus is purely on chemical process systems. The
addition of biological content along with the requisite model-
ing and analysis techniques requires a carefully crafted course
to avoid leaving students overwhelmed. A possible structure
for a semester-long course is illustrated by the syllabus in
Table 1, where NL is the number of lectures allotted to the
specific topics listed in all caps. Bold entries represent new
topics specific to biological systems. Italicized entries are
theoretical topics often considered optional in a traditional
course but which are viewed as important for a biologically
oriented course.


The introduction of state-space
models and associated analysis
tools is essential for the treat-
ment of biological systems due to
their complexity (e.g., high order,
multivariable, highly nonlinear),
which often precludes simple
Laplace domain treatment. A few
lectures on matrix algebra and
linear state-space systems are
necessary to review core material
and ensure that students with defi-
cient backgrounds understand the
basic concepts. When combined
with the linear systems analysis
lecture, this material allows the
calculation of eigenvalues to de-
termine stability and matrix rank
for the analysis of controllability
and observability. The nonlinear
systems theory lecture includes
the traditional topic of Jacobian
linearization as well as
introductory coverage
of phase plane analy-
sis, multiplicity, and
bifurcations. Biological Chen
Chen
systems are inherently Conti
nonlinear, given the ex- simple
istence of saturation proce
phenomena, stable os- Biote
cillations, etc. As such, Conti
a student must have a cell c
working knowledge of Biom
Baror
nonlinear systems to metal
be able to identify such drugmeta
be able to identify such drug


behavior and analyze system response in the presence of
nonlinear phenomena. Without question, this topic could
comprise a course unto itself. Some basic tools (e.g., phase
planes, limit cycles, bifurcation) are easy enough to teach
in a class or two, however. These provide students with an
ability to identify nonlinear system characteristics, even if
they cannot design a linearizing-state feedback controller to
address the underlying nonlinearity. Feedback is a concept
that is introduced naturally in the context of biological system
examples. The representation of biological control systems
using various elements of the traditional block diagram is
particularly effective. This approach, however, should be used
carefully to avoid concealing the complexity of the underlying
biological processes.
Throughout the topic sequence in Table 1, a number of
examples serve to highlight the breadth of opportunities for
application of the theoretical concepts presented in the course.
Table 2 provides a list of potential case studies. For each


TABLE 1
Proposed Syllabus for a Biologically Oriented Dynamics and Control Course
NL-number of lectures all caps=topic area bold=new topics italicized=optional topics
NL Topics
4 DYNAMIC MODELING
Principles of fundamental modeling; chemical and biological process examples; introduction
to empirical modeling
7 LINEARAND NONLINEAR SYSTEMS ANALYSIS
Matrix algebra and linear state-space systems; linear systems theory; introduction to nonlinear
systems theory; dynamic simulation; chemical and biological process examples; introduction
to the Laplace transform
7 FEEDBACK SYSTEMS
Basic principles of feedback; physiological control systems; homeostasis as a setpoint-free
feedback system; feedback in biochemical reaction networks; closed-loop response analy-
sis; servo vs. load behavior; feedback control of chemical process systems; closed-loop drug
delivery
8 FEEDBACK CONTROL SYNTHESIS
Basic principles of model-based controller design; PID controller design and tuning; advanced
single-variable control techniques; multivariable control techniques; modelpredictive control;
chemical and biological process examples
4 ADVANCED TOPICS
Large-scale systems andplantwide control; parameter estimation and experimental design;
state estimation; introduction to systems biology

TABLE 2
Possible Case Studies for the Process Dynamics and Control Course
ical Processes
nuous and/or fed-batch polymerization reactor; distillation column; continuous pulp digester; paper machine;
e plantwide example (e.g., reactor and separator); semiconductor process (e.g., lithography); photovoltaic film
ssing; fuel cell
chnological Systems
nuous and/or fed-batch fermentor; yeast energy metabolism; cell stress response (e.g., heat shock); eukaryotic
ycle; bacterial chemotaxis
medical Systems
*eceptor vagal reflex (blood pressure control system); insulin-dependent diabetic patient (glucose-insulin
bolism/control); circadian rhythm gene regulatory network; anesthesia control; drug delivery for HIV treatment;
delivery for cancer treatment
Chemical Engineering Education











topic where examples are listed in the syllabus, two chemical
process and two biological system examples could be used to
develop lecture materials, in-class exercises, and recitation
problems. Ideally the biological problems are divided equally
between the biotechnology and biomedical lists.
A major conclusion of the MIT-organized education work-
shops was that multiscale phenomena should be incorporated
throughout the undergraduate chemical engineering curricu-
lum.[2] A useful connection between the traditional chemical
and biological examples listed in Table 2 is the wide range of
time and length scales at which these systems can be analyzed.
Polymerization reactor models can be developed using input-
output representations,[3] detailed descriptions of the individual
polymer particles and their interactions,[4 6] or a variety of
scales in between.7 9] Analogous models can be developed
for microbial fermentors where lumped descriptions of cel-
lular processes are provided by unsegregated modelst1 13] and
detailed descriptions of the individual cells are provided by
cell population models. 14] While the introduction of biological
systems content is not necessarily required to illustrate these
concepts, we feel that an integrated program of chemical and
biological examples will reinforce key concepts and demon-
strate that these diverse examples are conceptually similar.

UMASS CHE 446: INCORPORATING
BIOTECHNOLOGY
The process dynamics and control course at the University
of Massachusetts ()
has traditionally focused on Laplace transform methods and
chemical process applications. This course usually repre-
sents the only extensive exposure to dynamic modeling and
feedback control in the undergraduate curriculum. Biological
systems were chosen as an appropriate vehicle for introducing


the key elements of biological transformations, multiscale
phenomena, and systems-level analysis identified in the MIT-
sponsored education workshops.[21 Rather than completely
change the existing course content, a more conservative
approach based on the integration of biological systems and
the requisite analysis techniques was pursued.
The current syllabus for the UMass course (ChE 446) is
shown in Table 3, where new topics introduced in the past two
years are italicized. The first few weeks are focused on fun-
damental modeling because undergraduate students typically
have little experience formulating dynamic balance equations.
Two biological examples-a continuous yeast fermentor
model and a structured yeast cell model-are introduced
and revisited throughout the semester. Both time domain and
Laplace domain analysis techniques receive extensive cover-
age. A major focus is the formulation and stability analysis of
linear state-space models. Engineered and natural-feedback
systems are introduced in parallel to highlight their common
features and unique properties. While most of the material on
single-loop controller synthesis is traditional, an introduction
to time domain controller design and analysis techniques is
provided to parallel the Laplace domain methods. The final
few weeks are focused on multivariable control systems with
an emphasis on linear model predictive control.
To accommodate the new material on biological systems
and time domain techniques, material previously covered in
the course had to be de-emphasized or virtually eliminated.
Topics that received reduced coverage included transfer func-
tion models, Laplace domain analysis and design techniques,
advanced single-loop control strategies, and traditional chemi-
cal process examples. Frequency domain techniques received
very limited coverage. While these topics are admittedly
important, a broader view of dynamic systems and feedback


TABLE 3
Syllabus for UMass Course ChE 446: Process Control
NL=number of lectures italicized=new topics all caps=topic area
NL Topics
5 FUNDAMENTAL MODELING
Basic principles; chemical process examples (nonisothermal chemical reactor; binary flash unit; binary distillation column);
biochemical system examples (continuous fermentor model; metabolically structured yeast cell model)
7 DYNAMIC SYSTEM ANALYSIS
Linear algebra (solution of matrix equations, state-space models; eigenvalues and eigenvectors); time domain analysis (basic
stability concepts, linearization of nonlinear models, linear stability analysis, continuous fermentor example); Laplace trans-
forms; transfer function models; empirical models; parameter estimation
6 FEEDBACK SYSTEMS
Process control systems; biological feedback systems (engineered vs. natural feedback systems, yeast sulfate assimilation
pathway, baroreceptor vagal reflex); closed-loop transfer functions; closed-loop stability
7 FEEDBACK CONTROL SYNTHESIS
PID-controller tuning; internal model control; time domain controller design (state feedback, pole placement, model matching,
continuousfermentor example); feedforward control; cascade control
5 MULTIVARIABLE CONTROL
Control loop interactions; decentralized control; discrete-time models (discretization of continuous-time models, convolution
models, prediction models); model predictive control (controller design and tuning, constraint handling, real-time optimiza-
tion, continuous fermentor example)
Summer 2006










control was deemed to be more important given current trends
in the chemical engineering profession. In fall 2003, each
student was asked to evaluate the biological systems content
using a score ranging from "5" if they strongly agreed the
objective was achieved to "1" if they strongly disagreed the
course objective was achieved. Results obtained from the 21
respondents are summarized in Table 4. The average scores
are similar to those obtained for the other course objectives,
thereby indicating that the biological content was successfully
integrated into the course.

PITT CHE 0500: INTRODUCING BIOMEDICINE
The biology component in the Systems Engineering I:
Dynamics Modeling course (ChE 0500, pitt.edu/~che0500>) at Pittsburgh focuses on the analysis
of, and controller synthesis for, biomedical systems at the
whole-organism level. By integrating the research activities
in modeling and control of diabetic and cancer case studies
within the undergraduate class, students are exposed to a
novel application area. This format has resulted in a steady
flow of undergraduates interested in undergraduate research,
and an increased interest in graduate study. Students at Pitt
were posed the same questions as those at UMass; responses
can be found in Table 4. While confidence in dynamic balance
construction is not as high as that shown in the UMass course,
the other questions return similar quantitative responses indi-
cating that the biomedical topics were well received.
ChE 0500 is approached from a model-based perspective;
approximately half of the course is focused on modeling
systems using both fundamental and empirical approaches,
in both continuous and sampled-data (i.e., discrete) domains.
From the fundamental modeling perspective, the students are
taught to distinguish pharmacokinetics (the time profile of a
drug) from pharmacodynamics (the disease dynamics, effect
of the drug on the disease, and toxicity) in much the same
way valve dynamics and process output response are captured
by separate blocks in a block diagram. The remainder of the
course focuses on the model-based synthesis and analysis of
classical and advanced control systems, as in Table 1.
As a case study, consider the insulin-dependent diabetic
patient depicted in Figure 1. Fundamental model construction


introduces students to the key variables of the diabetic-patient
problem and demonstrates the utility of skills developed
elsewhere in the curriculum (e.g., dynamic mass balance with
reaction, transport resistance) in the modeling of biomedical
problems. Students then work with this model, or suitable
lower-order approximations, 5] throughout the semester on
in-class problems, homework, etc.
The case study methodE16 is commonly employed in teach-
ing to facilitate in-depth treatment of problems in limited



SInsulin Meal Exercise
SInfusion Disturbance t Disturbance
..... ...................................


Brain
I i

Heart/ _
Lungs


I Gut i- \




SKidne -




Periphery
L ....- ,-- ,-- -- -,- ,--,-. -,-, -..-- -

t Glucose Measurement


Figure 1. Open-loop schematic of the diabetic patient.
Small solid blocks represent the fundamental model, with
manipulated input insulin delivery rate, meal distur-
bance, exercise disturbance, and glucose concentration
measurement.


TABLE 4
Student Responses to Biological Systems Content in the UMass (21 respondents) and
Pitt (17 respondents) Process Control Courses
Score
Question UMass Pitt
I can construct a dynamic model of a biological system. 3.83 3.23
I can perform dynamic system analysis and controller design in the time domain. 3.78 3.71
I can apply dynamic system analysis techniques to biological systems to evaluate 3.89 3.76
properties such as stability.
I can describe the relevance of feedback control theory to biological systems. 3.83 3.77
Chemical Engineering Education











classroom time. An added benefit would be to use a unifying
application, thereby allowing students to focus their attention
on a single problem. The diabetic patient is one such problem,
and case studies from the literature have been mapped onto
the course outline (Table 1). The map in Table 5 provides a
guide to focused literature reading that allows biomedically
motivated problems to be quickly brought into the classroom.
Case study-specific tables of this form are most useful to
faculty who are not dynamics and control experts, but who
are responsible for teaching the course, because the dynamics
and control class is a challenging course for nonexperts to
teach. A collection of these paper-topic maps, for traditional
and biological case studies, would provide those teaching
the dynamics and control course with a variety of examples
tailored to each section of the course.


UCSB CHE 154: A COURSE IN SYSTEMS
BIOLOGY

In addition to the required dynamics and control course, de-
scribed earlier, there is a demand in many chemical engineer-
ing programs for elective courses that facilitate specialization
in either systems engineering or biotechnology/biomedical
engineering. At UCSB, a new course was offered in the
spring 2004 quarter entitled Engineering Approaches to Sys-
tems Biology (ChE 154/BMSE 255). The course is taught at a
dual level (seniors and new graduate students), and fulfills the
track requirement for both systems and biology emphases in
the undergraduate chemical engineering program. The current
syllabus is listed in Table 6, detailing the topics for a single-
quarter course (20 lectures of duration 75 minutes).


TABLE 5
Integration of Sample Case Study (insulin-dependent diabetic patient) with Course Outline Topics
DATA-DRIVEN MODELING
Sorensen FOTD,E23 Bolie two-state linear,"1 Bergman "minimal" model[24]
FIRST PRINCIPLES MODELING
Physiologically based 1.1 .. .. 1. .. 1. 1 n .. .
LINEAR SYSTEMS ANALYSIS
Bolie two-state linear ODEsE15"
LINEAR SYSTEMS ANALYSIS w/ LINEARIZATION
Linearize and analyze Bergman "minimal"[24]
DYNAMIC SIMULATION
All models, including AIDA as a different performance classificationE15 18 23 261
FEEDBACK SYSTEMS
Glucose-insulin interactions[151; nonlinear feedback response[24]; healthy pancreas response23, 27]
CLOSED-LOOPANALY SIS
Sorensen healthy patient[23]
PID CONTROL
Controller design from FOTD,[231 low-order ODEs, E15 and linearized systems and/or effects of nonlinearity[24]
ADVANCED CONTROL
Feedforward for meal disturbancesR28 and exercise,[29] with simple[15 24] or complex[23] case studies
MULTIVARIABLE CONTROL
MISO (glucose and insulin inputs; G, I, and exercise inputs)o301 or MIMO (glucose and insulin control) for a variety of i ....' *
MODEL PREDICTIVE CONTROL
Linear MPC in analytical"151 or data-driven 311 forms; MPC with a linearized model[23 24 32, 33]; nonlinear MPC if desired[23 24 3234]


TABLE 6
Syllabus for UCSB Course: ChE 154 Engineering Approaches to Systems Biology
NL Topics
6 CELLULAR REGULATION
Central dogma; genome sequences; genome expression; genomic circuits; protein, metabolic, signaling networks; high throughput
biological data; biological databases
6 MATH MODELING AND SYSTEMS ANALYSIS TOOLS
Modeling strategies; boolean models; nonlinear ODE models; discrete stochastic models; systems biology modeling packages;
network analysis-robustness, identifiability; design of experiment issues
6 BIOSY STEMS CASE STUDIES
Bacterial chemotaxis; lambda phage virus; circadian rhythm gene network; signal transduction in apoptosis; synthetic biological
circuits
2 COURSE PROJECTS
Midterm progress reports; final presentations
Summer 2006 1,










The course focuses on the emerging problems in systems
biology and computational biology. There is a substantial
level of effort being invested in these areas in both academia
and industry, and the demand for training of students has
increased in proportion. These advances have been facilitated
by developments in both computational modeling and high
throughput biology-enabling a systematic approach to ana-
lyzing complexity in biophysical networks that was previously
untenable. These studies provide increasingly detailed insights
into the underlying networks, circuits, and pathways respon-
sible for the basic functionality and robustness of biological
systems. They also create new and exciting opportunities
for the development of quantitative and predictive modeling
and simulation tools. Model development involves translat-
ing identified biological processes into coupled dynamical
equations that are amenable to numerical simulation and
analysis. These equations describe the interactions between
various constituents and the environment, and involve mul-
tiple feedback loops responsible for system regulation and
noise attenuation and amplification.
The discipline of "systems biology" has emerged in re-
sponse to these challenges, 17 and combines approaches and
methods from systems engineering, computational biology,
statistics, genomics, molecular biology, biophysics, and other
fields. The recurring themes include: (i) integrative viewpoints
toward unraveling complex dynamical systems, and (ii) tight
iterations between experiments, modeling, and hypothesis
generation. In response, there have been a number of courses
introduced in a variety of departments across the country
that address elements of systems biology and computational
biology. These have been targeted at both undergraduate and
graduate audiences, and in some cases involve continuing
education participants from industry. The balance of topics
in the syllabus in Table 6 is approximately one-third on ba-
sic cellular regulation, one-third on applications of systems
engineering tools to biological problems, and one-third on
detailed case studies to illustrate current methodologies and
future challenges. Although the UCSB curriculum is based
on quarters, the same general template could be extended to
a semester-long course without significant modification.
Assignments for this course consist of short homework
problems, primarily at the beginning of the course, and a
major course project. The project entails a midterm progress
report, a final presentation, and a written report. The case study
offers mechanism to tailor the course to a diverse studentpopula-
tion-seniors work in teams with reduced scope, while graduate
students work as individuals on a more detailed project.

OPEN ISSUES
Laplace Domain Methods
Traditional process control courses emphasize Laplace
transform methods for analyzing and designing feedback
systems. While traditional analysis may be facilitated by


Laplace domain representations, the applicability of these
methods to the complex systems commonly encountered in
biological problems is severely limited. Biological systems
are inherently nonlinear with phenomena ranging from pro-
tein interactions in gene regulatory networks to adaptation
in systemic reflexes. Furthermore, modeling of biological
systems at resolutions below the macroscopic scale often
leads to high-state dimension.[14 18] As is evident from Table
1, Laplace domain methods have been de-emphasized and
frequency domain techniques have been effectively removed
from the proposed curriculum. While we do not dispute their
potential value, transform-based methods introduce concep-
tual difficulties that cause many students to lose their physical
insights and view the material as applied mathematics. On
the other hand, the syllabus in Table 1 is sufficiently flexible
that limited coverage of frequency domain methods at the
expense of other topics is possible.

Time Domain Methods
Complex dynamic system models are most effectively for-
mulated and analyzed in the time domain using conservation
equations. Consequently, the syllabus in Table 1 focuses on
linear and nonlinear state-space models. Connections with the
corresponding Laplace domain concepts can be introduced
as appropriate (e.g., stability via eigenvalues vs. poles). On
the other hand, the Laplace transform is a particularly useful
tool for single-input, single-output (SISO) systems with time
delay and/or zero dynamics. We acknowledge that analytical
treatment of zeros in the time domain is more involved than
the corresponding Laplace methods. Time domain analysis of
transportation and measurement delays is most conveniently
performed using a discretized framework based on state
augmentation. Because this approach can lead to potentially
large state dimensions, evaluating student understanding of
this material can be challenging. A possible solution is to
use a combination of relatively simple exam questions and
more detailed homework problems. While control system
design issues can be addressed using continuous state-space
models, we believe that a discrete-time framework is pre-
ferred for introducing data-driven model identification and
sampled-data systems. Recent results have shown that a
properly tuned SISO model predictive controller cannot be
outperformed by a conventional proportional-integral-deriva-
tive (PID) controller.[19 Because we expect this fact to be
reflected in industrial practice, the syllabus in Table 1 offers
increased exposure to controller synthesis techniques based
on discrete-time representations such as step response models.
While a comprehensive treatment is beyond the scope of this
course, model predictive control (MPC) should be foremost
among the topics covered due to its industrial importance.
As outlined in the UMass course syllabus (see Table 3), the
introduction of MPC necessitates limited discussion of real-
time optimization and draws on the discrete-time modeling
tools discussed above.
Chemical Engineering Education










Multivariable Control
While most traditional courses treat multivariable systems
as a straightforward extension of SISO systems, a more com-
prehensive approach that addresses the unique challenges of
multivariable controller design is warranted. A formal intro-
duction to decentralized control would support the systems
viewpoint of multivariable processes -a set of optimal SISO
feedback loops generally does not result in overall system
optimality. Another advantage of introducing MPC is that
multivariable system complexity is handled in a transparent
and systematic manner. Students can gain appreciation for
the effects of constraints and optimization-based methods for
constraint compensation.
Robustness
A critical topic in the analysis of both process control
systems and biological regulation is robustness. While the
remarkable levels of robust performance attained in nature
are enviable from an engineering perspective, this issue is not
widely appreciated in biology. The critical importance of ro-
bustness in understanding disease states, as well as evolution
and development, motivates its incorporation in the system
dynamics and control curriculum. While a detailed theoretical
treatmentE"20 is beyond the scope of a typical undergraduate
course, key concepts of robustness can be emphasized us-
ing simple tools such as sensitivity analysis-effectively
capturing the gains from uncertain system elements to the
controlled output or performance measure. Students would
be well positioned to evaluate parametric sensitivities using
state-space models in the proposed curriculum. Robustness
analysis could also be used to study closed-loop strategies
such as redundancy, feedback, filtering, and modular protocols
commonly used in nature.
Nonlinear Analysis and Control
Most biological systems are not adequately described by
linear dynamic models since nonlinear effects such as satu-
ration phenomena are ubiquitous. Consequently, linear and
linearization-based analysis techniques are rarely sufficient.
Nonlinear analysis techniques, such as phase plane analysis and
bifurcation theory (see Table 3), can be introduced explicitly,
thereby exposing students to theoretical concepts and analysis
tools with wider applicability than Laplace domain methods.
Nonlinear phenomena are also common in industrial plants,
and linear control methodologies often require specialized tools
to handle strong nonlinearities. Linear controllers exhibit poor
performance for some nonlinear processes (e.g., high purity
distillation columns) and completely fail for particularly dif-
ficult processes (e.g., those displaying input multiplicity). Given
increased exposure to linear MPC in the revised curriculum, a
brief introduction to nonlinear MPC is entirely feasible.
Teaching Control for Nonexpert Faculty
Our experience indicates that the process dynamics and
control class is not a popular choice as a teaching assignment
Summer 2006


among nonexperts in the field. This lack of interest is due to
a variety of issues, including the mathematical complexity
of the material and the significant focus on feedback con-
troller synthesis. An additional concern is that the material
is challenging to students, who have had limited exposure
to dynamical systems prior to this course. The syllabus in
Table 1 represents a significant departure from the traditional
controller-synthesis-dominated course to a more balanced
presentation of system dynamics and feedback.
A notable benefit of the proposed syllabus is the degree
of potential customization. While our focus has been on the
introduction of biological systems content, the treatment of
other application areas such as advanced materials can be ac-
complished in a similar manner. This flexibility provides an
excellent opportunity for instructors to integrate their research
interests into the course. In fact, the three courses described
here were heavily influenced by the work performed in our re-
search groups. Possible benefits of such integration include: (i)
increasing the diversity of application examples by encourag-
ing nonexperts to teach the course; and (ii) introducing students
to cutting-edge research that influences their perception of the
field and may affect their future career directions.

SUMMARY
Biological processes have assumed an increasingly im-
portant role in chemical engineering research and practice.
Modifications of the existing chemical engineering curriculum
are necessary to provide undergraduate students the needed
exposure to this emerging field. We believe that the capstone
process dynamics and control course provides an excellent
opportunity to integrate biological systems content and draw
parallels with chemical process applications that have been
the traditional focus of this course. This paper provides a sum-
mary of work on this problem at our respective institutions.
The proposed curriculum allows biological content and
time domain concepts to be introduced in a synergistic man-
ner without adversely affecting the coverage of traditional
material. As outlined in the proposed syllabus, this requires
a decrease in time spent on traditional topics such as PID
controller synthesis, Laplace transform techniques, and fre-
quency response analysis. Advances in feedback controller
tuning (e.g., autotuning and model-based methods) combined
with the availability of simulation/analysis tools (e.g., MAT-
LAB, LabVIEW) bring into question the need for extensive
treatment of pencil-and-paper analytical techniques that are
rarely employed, even at the graduate level. While focused
time on these topics has been reduced in the name of incorpo-
rating biology, it should also be noted that the analysis tools
introduced in the dynamics and control class are applicable to
problems beyond biological systems. Hence, students are no
less prepared for "traditional" industrial positions, and they
are certainly more equipped for positions in pharmaceuticals
and systems biology.












A key hurdle that must be overcome is the lack of instruc-
tional materials to support the new process dynamics and
control curriculum. For the courses outlined above, the authors
are using new textbooks (System Modeling in Cell Biology,
MIT Press) or have developed supplementary materials to
complement existing textbooks. Researchers in process dy-
namics and control can contribute in a variety of ways. The
construction of extended case studies such as Table 5 for
various applications would ease the burden on nonexperts
teaching the course. Software tools such as the Process Con-
trol Modules[21] and Java-based Control Modules[22] are well
suited for introducing traditional concepts and applications.
New software tools are needed to expose chemical engineer-
ing undergraduates to biological complexity and to allow the
application of theoretical concepts to representative biological
systems. Ongoing efforts, such as those organized by MIT and
the CACHE Corporation, are focused on the development of
biologically focused systems courses. A task force headed
by the second author of this paper is currently working on
course revisions as well as software module development as a
means to integrate biological content throughout the chemical
engineering curriculum. More details on this effort will be
made available at .


ACKNOWLEDGMENTS

Support for RSP was provided by the National Science
Foundation CAREER program (CTS #0134129).


REFERENCES
1. AIChE, 2001-2002 initial placement of chemical engineering graduates
(2002)
2. Massachusetts Institute of Technology, Frontiers in Chemical Engineer-
ing Education Initiative (2003)
3. Parker, R.S., D. Heemstra, EJ. Doyle III, R.K. Pearson, and B.A. Ogun-
naike, "The Identification of Nonlinear Models for Process Control
Using Tailored 'Plant-Friendly' Input Sequences," J. Proc. Control,
11, Sp. Issue SI:237-250 (2001)
4. Immanuel, C.D., C.E Cordeiro, S.S. Sundaram, E.S. Meadows, T.J.
Crowley, and EJ. Doyle III, "Modeling of Particle Size Distribution
in Emulsion Co-Polymerization: Comparison with Experimental Data
and Parametric Sensitivity Studies," Comp. ( I .... i 2 1133-1152,
(2003)
5. Semino, D., and WH. Ray, "Control of Systems Described by Popula-
tion Balance Equations I. Controllabity Analysis," Chem. Eng. Sci.,
50(11), 1805-1824 (1995)
6. Semino, D., and WH. Ray, "Control of Systems Described by Popula-
tion Balance Equations II. Emulsion Polymerization with Constrained
Control Action," Chem. Eng. Sci., 50(11) 1825-1839 (1995)
7. Congalidis, J.P, J.R. Richards, and W.H. Ray, "Feedforward and
Feedback Control of a Solution Copolymerization Reactor," AIChE
J., 35(6) 891-907 (1989)
8. Uppal, A., W.H. Ray, and A.B. Poore, "On the Dynamic Behavior of
Continuous Stirred Tank Reactors," Chem. Eng. Sci., 29, 967-985,
(1974)
9. Daoutidis, P, M. Soroush, and C. Kravaris, "Feedforward/Feedback
Control of Multivariable Nonlinear Processes," AIChE J., 36(10)
1471-1484(1990)
10. Henson, M.A., and D.E. Seborg, "Nonlinear Control Strategies for
Continuous Fermentors," Chem. Eng. Sci., 47, 821-835 (1992)


11. Chang, Y.K., and H.C. Lim, "Experimental and Simulation Studies of
MultivariableAdaptive Optimization of Continuous Bioreactors Using
Bilevel Forgetting Factors," Biotech. Bioeng., 34, 577-591 (1989)
12. Semones, G.B., and H.C. Lim, "Experimental Multivariable Adaptive
Optimization of the Steady-State Cellular Productivity of a Continuous
Baker's Yeast Culture," Biotech. Bioeng., 33,16-25 (1989)
13. DiBiasio, D., H.C. Lim, and WA. Weigand, "Experimental Investi-
gation of Stability and Multiplicity of Steady States in a Biological
Reactor, "AIChE J., 27, 284-292 (1981)
14. Henson, M.A., "Dynamic Modeling of Microbial Cell Populations,"
Current Opinion in Biotechnology, 14, 460-467 (2003)
15. Bolie, V.W, "Coefficients of Normal Blood Glucose Regulation," J.
Appl. Physiol., 16, 783-788 (1961)
16. Mustoe, L.R., and A.C. Croft, "Motivating Engineering Students by
Using Modern Case Studies,"Int. J. Eng. Ed., 15, 469-476 (1999)
17. Kitano, H., "Systems Biology: A Brief Overview," Science, 295,1662-
1664(2002)
18. Parker, R.S., J.H. Ward, N.A. Peppas, and FJ. Doyle III, "Robust H
Glucose Control in Diabetes Using a Physiological Model," AIChE J.,
46, 2537-2549 (2000)
19. Pannocchia, G., N. Laachi, and J.B. Rawlings, "A Fast, Easily Tuned,
SISO, Model Predictive Controller, Proc. DYCOPS (2004)
20. Skogestad, S., and I. Postlethwaite, Multivariable Feedback Control,
John Wiley & Sons, New York (1996)
21. Doyle, EJ. III, R.S. Parker, and E.P Gatzke, "Process Control Modules:
A Software Laboratory for Control Design," Prentice Hall International
Series in the Physical and Chemical Engineering Sciences, PH PTR,
Upper Saddle River, NJ (2000)
22. Yang, D.R., and J.H. Lee, "Process Control Education Software Using
Java Applet," AIChE Annual Meeting, formation/research/issicl/che4400/ javamodule.html> (2002)
23. Sorensen, J.T., "APhysiologic Model of Glucose Metabolism in Man and
its Use to Design and Assess Improved Insulin Therapies for Diabetes,"
Ph.D. thesis, Department of Chemical Engineering, MIT, (1985)
24. Bergman, R.N., L.S. Phillips, and C. Cobelli, i I ..... .. Evaluation
of Factors Controlling Glucose Tolerance in Man, "J. Clin. Invest., 68,
1456-1467 (1981)
25. Lehmann, E.D., T. Deutsch, E.R. Carson, and PH. Sonksen, "AIDA: An
Interactive Diabetes Advisor," Comp. Meth. Prog. Biomed., 41,183-203
(1994)
26. Agar, B.U., G. Birol, andA. Cinar, "Virtual Experiments for Control-
ling Blood Glucose Level in Type 1 Diabetes," in Proc. Second Joint
EMBS/BMES Conf., p. 2609 (2002)
27. Nomura, M., M. Shichiri, R. Kawamori, Y. Yamasaki, N. Iwama, and
H. Abe, "A Mathematical Insulin-Secretion Model and its Validation
in Isolated Rat Pancreatic Islets Perfusion," Comput. Biomed. Res., 17,
570-579 (1984)
28. Lehmann, E.D., and T. Deutsch, "A Physiological Model of Glucose-
Insulin Interaction in Type 1 Diabetes Mellitus," J. Biomed. Eng., 14,
235-242 (1992)
29. Lenart, PJ., and R.S. Parker, "Modeling Exercise Effects in Type 1
Diabetic Patients," Proceedings of the 15th IFAC World Congress on
Automatic Control, Barcelona (2002)
30. Parker, R.S., E.P Gatzke, and EJ. Doyle III, "Advanced Model Pre-
dictive Control (MPC) for Type 1 Diabetic Patient Blood Glucose
Control," in Proc. American Control Conf., Volume 5, pp. 3483-3487
(2000)
31. Parker, R.S., EJ. Doyle III, and N.A. Peppas, "A Model-Based Algo-
rithm for Blood Glucose Control in Type 1 Diabetic Patients," IEEE
Trans. Biomed. Eng., 46(2) 148-157 (1999)
32. Parker, R.S., EJ. Doyle III, and N.A. Peppas, "The Intravenous Route
to Blood Glucose Control," IEEE Eng. Med. Biol., 20, 65-73 (2001)
33. Lenart, P.J., and R.S. Parker, "Glucose Control During Exercise in
Type I Diabetic Patients," Proceedings of the Topical Conference on
Bioinformatics and Genomics, AIChE Annual Meeting (2001)
34. Florian, J.A. Jr., and R.S. Parker, "Empirical Modeling for Glucose
Control in Diabetes and Critical Care," Eur. J. Control, 11 (2005) [
Chemical Engineering Education











IMRjt learning in industry


THE ROLE OF INDUSTRIAL TRAINING IN


CHEMICAL ENGINEERING EDUCATION




MAMDOUH T. GHANNAM
United Arab Emirates University Al-Ain, United Arab Emirates


industrial training plays an important role in preparing
engineering students to be future professional chemical
engineers. The training offers a golden opportunity to
acquire numerous technical and nontechnical skills that can
not be obtained in a classroom environment. Some of the
benefits of industrial training are1-'41: observing daily work
activities firsthand in a real setting, gaining the ability to apply
technical and theoretical knowledge to industrial problems,
direct exposure to nontechnical skills such as oral and written
communications, understanding the diversity of the chemical
engineering industries, applying computer software programs
to real industrial situations, teamwork experiences, time man-
agement and deadline objectives, getting familiar with the
industrial environment to set and achieve future career goals,
working effectively in a multidisciplinary environment, and
boosting the student's self-esteem and confidence by gaining
today's industrial skills.

UNDERGRADUATE INDUSTRIAL PROGRAMS
There is no classroom course that could simulate or replace
the industrial experience gained from working with several
operators in a real industrial environment. Numerous universi-
ties allow their students to gain industrial experience through
a variety of programs.


Cooperative Education Program
One such program is the cooperative education program." 3]
Co-op education is based on rotation between schooling and
full-time work periods. It connects undergraduate students
directly with industry to gain strong fundamentals and in-
valuable insight into the chemical engineering profession,
acquire technical knowledge, earn academic credits, and
receive wages. Some universities offer co-op programs on an
optional basis while others are mandatory. Among the colleges
of engineering offering an optional program are the Univer-
sity of South Alabama,[s5 the University of Minnesota,[6] and
the University of Pittsburgh. 71 Mandatory programs can be


Mamdouh Ghannam received his Ph.D.
from the University of Saskatchewan, Can-
ada, in 1991. He was an assistant professor
at Cairo University, Egypt, from 1992-1995,
after which he joined the University of Sas-
katchewan as a research engineer. Since
1998, he has been with the Department of
Chemical and Petroleum Engineering at the
United Arab Emirates University as an as-
sociate professor. His research interests are
coating of Newtonian and non-Newtonian
solutions, and interfacial and theological
properties of crude oil emulsions.


Copyright ChE Division of ASEE 2006


Summer 2006


This column provides examples of cases in which students have gained knowledge, insight,
and experience in the practice of chemical engineering while in an industrial setting. Summer
internships and co-op assignments typify such experiences; however, reports of more unusual
cases are also welcome. Description of the analytical tools used and the skills developed during
the project should be emphasized. These examples should stimulate innovative approaches to
bring real-world tools and experiences back to campus for integration into the curriculum. Please
submit manuscripts to Professor W.J. Koros, Chemical Engineering Department, Georgia Institute
of Technology, Atlanta, GA, 30332-0100.










found in universities such as Drexel,[8] Ryerson,[9] Toledo,[101
and Cincinnati.E11 Cooperative education programs have been
well established in these universities for a long time. For ex-
ample, cooperative education was founded at the University
of Cincinnati111 in 1906 and at Drexel UniversityE8' in 1919.
Most cooperative education programs offer financial benefits
to students. The financial reward for students is usually based
on location and type of task. Students' wages from their co-op
jobs can even help finance their educations. At the University
of Cincinnati,E11 average salaries for co-op jobs are almost
twice the tuition fees. Therefore, students do not have the
burden of having a part-time job, giving them more time to
devote to academics and other activities. At the University
of Pittsburgh's7'] Department of Chemical and Petroleum
Engineering, approximately 40-50% of the students take
advantage of the available co-op program. According to a
study completed by the co-op office in Cincinnati,E11 96% of
graduating students acknowledged that the college program
including co-op experience provided a better education than
the traditional program without co-op.
Most engineering schools that offer cooperative education
have certain requirements for a student to participate in the
program. For example, at the University of Minnesota,[6] the
undergraduate student needs to be in good standing, have
completed all program course requirements including fall
semester of the third year, have completed at least five out
of seven elective courses, and maintain a minimum GPA of
2.8. The co-op student will work full time for one year con-
tinuously in the industry. After successfully completing their
co-op programs, students earn two credit hours per semester,
which count toward technical elective courses.


Industrial Training Program
Around 10 years ago, the College of Engineering at the
United Arab Emirates University (UAEU) recognized the
importance of industrial training and its crucial role in pre-
paring students for professional engineering. In 1995, a very
committed program was established by the Unit of Industrial
Training and Graduation Projects at the college. This pro-
gram is mandatory-i.e., part of the engineering education
curriculum-for all engineering students in the disciplines
of chemical, civil, electrical, mechanical, and petroleum
engineering. Students can be granted 15 credit hours after suc-
cessful completion of their industrial training. The following
discussion focuses only on the chemical engineering industrial
training program (ITP). The industrial training program at
UAEU is selected in this article as an example to address its
benefits and highlight areas that can be improved.

PROGRAM LOGISTIC OF ITP

To ensure that students have enough theoretical background
to comprehend industrial training tasks, the program requires
students have a minimum GPA of 2.5 and to have completed
114 credit hours. Any student with a lower GPA will be
required to complete 126 credit hours (the total credit-hour
requirement to earn a B.S. degree in engineering is 168).
Each eligible student should prepare and submit a file to the
industrial training unit that contains academic records, a one-
page resume, and a completed application form.
The industrial training unit works with each candidate to
find a training position with the available participating in-
dustrial partners. Students will be required to interview with


Municipality 7%-

Central Lab 2%~ armaceutical

Hospital 2%

Glass 1%

Aluminum 7%


Cement 7%

Fertilizer 1%


Oil
48%


17%
Water


1%

Cerarics


Figure 1. Participating industries.
Chemical Engineering Education


___










potential companies to be placed. If a student is not accepted
by one company, he/she will have an opportunity with another
company since the number of industrial partners exceeds
the number of training candidates. The industrial training
unit ensures that each candidate will receive placement in a
relevant industry within the country or abroad.
At the beginning of the training period, each student will re-
ceive a training-program schedule outlining the whole 16-week
period. This schedule includes a weekly job description with
tasks, academic advisor visits, and deadlines for reports and
presentations. The schedule is established by the industrial
supervisor in agreement with the student's academic advisor.
The student is also advised of the evaluation process used to
determine his or her grade. Student evaluation during this
program is based on weekly progress reports submitted to
the academic advisor (15% of total score), academic advi-
sor visits to the industrial site (15%), industrial supervisor's
evaluation report (20%), and final report and oral presentation
(50%). The presentation and final report are assessed by an
examination committee consisting of a college represen-
tative, a departmental faculty member, and two or three
professional engineers.
Upon successful completion of the ITP, student participants
are eligible to register into the final academic year in chemical
engineering. Failed participants must repeat the ITP program
at another industrial site. With this in mind, most students are
committed very seriously to the ITP.
From the above ITP description, it is notable that the objec-
tives are similar to cooperative education in allowing students
to gain industrial experience and learn the basics and funda-


mental of the industry. The ITP, however, strictly considers
students as trainees without financial compensation, not as
employees with wages as in co-op programs.

COMPANIES INVOLVED IN THE INDUSTRIAL
TRAINING PROGRAM
The number of participating companies in the ITP for the
whole college increased tremendously from around 10 compa-
nies during the initial year of 1995-1996 to approximately 140
companies in the academic year of 2004-2005. This increase
in participating companies, both locally and internationally,
reflects industrial appreciation of the important role of the
ITP The total number of chemical engineering students who
participated in the program from 1999-2005 is 198. The
number of academic advisors involved in these activities
changed from one semester to another based on the number
of students. For example, the number of faculty members in-
volved in the first semester of 2004-2005 was four, and in the
second semester of 2005-2006 it was seven. For the chemical
engineering discipline, the specialties of the involved indus-
tries vary widely, as can be seen in Figure 1. The reported
percentage represents the number of students completing the
program for each industry during the period of 1999-2005.
In addition, Figure 2 displays the various regions within the
UnitedArab Emirates, with exception of France and Qatar, in
which students of the UAEU carried out their training. These
percentages also cover the period of 1999-2005.
The following is a brief description of the ITP for a chemi-
cal engineering student who completed his training program at
the oil service company Dowell Schlumberger, inAbu Dhabi,


UmmAlQuw ain 1%AI Ain

11% 6%


Fujairah


Ras Al Khamiah
13%


Qata 1%

Sharja 7%1


DoU


Abu Dhabi
:: :: ::: 43%







2%
France


Figure 2. Industrial locations by region.


Summer 2006










UAE. His training period was February to May 2005. The first
eight weeks of his training included attending orientation,
learning company safety standards and injury protection,
performing laboratory tests, utilizing software, and learning
equipment maintenance.
During the second half of his training program, the student
attended a company course on equipment and machine safety.
He also had the opportunity to learn different aspects of the
business such as the transfer of product. In the laboratory, the
student was able to analyze materials and perform standard
API tests including rheology and stability tests. The student
was also able to use a Fann35 instrument to study the slurry-
flow behavior and measure the slurry viscosity at various
RPM. One sample used consisted of water, antifoam, disper-
sant, retarder, and cement, with composition of 18.65, 1.0,
0.15, 0.2, and 80 wt%, respectively. The
student measured shear stress data versus Bas
shear rate and concluded the yield stress


and viscosity for this sample was 9.83 lbf
100 ft2 and 40.03 cp, respectively.
I visited the student at the end of the
third and 13th week to check on his train-
ing performance and diary, and discuss
the feedback and recommendations of his
industrial supervisor. The student's gradua-
tion project proposal was also discussed.

IMPACTS AND POTENTIAL
IMPROVEMENT OF ITP
Impact
By becoming a part of the ITP, students


gain practical experience, technical knowl-
edge, confidence, time management skills,
teamwork capability, and a better understanding of what
they've learned in class. Based on a study performed by the
Industrial Training Unit at UAEU in the first semester of the
2003-2004 school year, 86% of students who completed the
ITP gained strong technical skills. A total of 69 students hav-
ing completed their industrial training in 42 industrial sites
participated in this study. This training mechanism enables
students to define their career goals and provides an oppor-
tunity to find a permanent employment position.
Additionally, this program is an excellent recruiting tool for
participating companies. Employers will have the chance to
train, evaluate, and select candidates for future job opportuni-
ties. A study was done by the industrial training unit at UAEU
to survey participating industrial employers from the first
semester of 2003-2004 to the first semester of 2005-2006. This
study investigated the responses of the industrial employers
on the trainees' performance with regard to the ABET2000
criteria (a to k criteria).'11 A total of 75 industrial employers


were involved in this study. The study showed that most of
the employers' evaluations exceeded the 70% limit (the ac-
ceptable limit established by the UAEU training office) for
all the criteria of a to k with a few exceptions. These excep-
tions occurred during the second semester of 2003-2004 and
first semester of 2005-2006. The employers' evaluation was
slightly less than the 70% limit for criteria c andj (the ability
to design a system, component, or process, and knowledge of
current engineering trends).
Improvement of ITP
If the faculty members are not significantly involved in
industrial-site supervision, project choice, follow-up, and
academic evaluation of student performance during the train-
ing period, the chances of an undergraduate student achieving
industrial experience are not very good.11i As
an academic advisor participating for the last
on seven years in the ITP, I feel improvements can
be made to the current program. Implementing
y these ideas will enhance the overall perfor-
d by mance and outcomes of the ITP. Suggested
trial improvements are:
nit (a) Site Selection
the To ensure the success of the training
who program for students, an academic advisor
should be involved from the very beginning.
ted The advisor needs to work closely with each
p student in selecting the appropriate training
site since there can be several choices. Doing
wrong so guarantees matching the student with an
al industrial site that meets the program objec-
tives, and avoids having students select an
industry based on convenience rather than
relevancy.
(b) Industrial Field Visits
The department should encourage and enforce a one-day
field visit once a semester for all students, especially freshmen,
to provide early exposure to different daily industrial activi-
ties.[3] Doing so allows students to start forming an opinion
as to what type of industry they want to pursue.
(c) Industrial Short Residence
Due to the short period of the ITP (i.e., four months) in
comparison with that usually spent by a trainee at other
schools such as Pittsburgh, 71 Drexel,I81 Cincinnati,[111 and Min-
nesota,[61 participating departments should develop a yearly
industrial event in which sophomore and junior students spend
two to four weeks during the summer at a local industry in
the residence city. Of course, each student should achieve
certain limited objectives during this period and be required
to provide a written report and oral presentation detailing
his/her activities and experiences gained. To make this idea


Chemical Engineering Education


a stud
perform
the Indus
Training U
86% of
students
complex
the IT
gained st
technic
skills


ed











even more practical, one industrial site can be selected for
each student. This requires the student to spend each summer
break in different departments of an industrial site.

(d) Nontechnical Skills Course

To improve the nontechnical skills of undergraduate
students in general and the ITP in particular, the college of
engineering should develop a special course equivalent to
two credit hours. The course would focus on enhancement of
nontechnical skills such as communication, effective presen-
tation, technical writing, accessing information, judgment,
software applications in industry, job interviews, resume
building, and explaining technical information to nontechni-
cal customers.[4,12]

This course would be offered to all students during the
second academic year, thus allowing enough time before the
ITP stage. In this way, students will be well prepared for the
ITP. They will be able to maximize the benefits and achieve
all the expected objectives of the ITP.

(e) Graduation Project Proposal

One requirement for the successful completion of the ITP
is the submission of a graduation project proposal within the
final report. This proposal should reflect a valuable idea that
attracted the trainee's attention during the training period. The
proposal may then be selected as a graduation project by a
faculty member within the chemical engineering department.
Based on my own experience over the period of 1998-2005,
I noticed problems associated with the graduation project
proposals such as unclear or incomplete proposals and the
lack of technical description and industrial data. It seemed
students completed their proposals just to meet the require-
ment criteria, or did not complete them at all.

The requirement to submit a graduation project proposal
itself is a great idea. For greater benefit, however, it needs
more involvement and commitment from each side. The fol-
lowing are some suggestions to strengthen the process of the
graduation project proposal:

1. All ,,,. i,a, .l. projects should be based on students'
industrial proposals.
2. Clear, limited, and well-defined criteria for ,,iiii.; *
projects should be established by the department at
the beginning of each academic year
3. Each group of three students will be assigned an
academic advisor relevant to their industrial training
site.
4. Forming a solid project proposal should be one of the
main duties of trainees in coordination with the
industrial supervisor and the academic advisor
5. Both the industrial supervisor and academic advisor


should be involved with students to achieve this task
by the end of the training period.
6. The proposal should cover the details of problem
definition, industrial importance, ,., m.- impacts,
alternative solutions, suggested solution and reasoning,
and positive impacts.
7. A significant grade should be assigned to the project
proposal.
8. The academic advisor should involve the industrial
supervisor .1. ., ... i 1 the ,,, ,. i,,i. *. project stages.
9. The industrial supervisor will be entitled to attend the
final project exam and to receive a copy of the final
report with all results and recommendations.

(f) Academic/Industrial Interaction

One of the main benefits of the suggested mechanism of
the ITP and graduation project proposal is the strong link and
interaction made possible between the industrial supervisor
and the academic advisor. This relationship can help both
parties advance their mutual interests, including:

1. Faculty member can make a strong connection with
different industries leading to research and technical
cooperation.
2. Industrial supervisor will have better access to the
university environment for assistance such as techni-
cal recommendations, hiring new graduates, acquiring
samples and data analysis, and p .., i. q ..ri,, in scien-
tific activities sponsored by the university.
3. Research cooperation between the two parties "lit",
in scientific publication reflects well on the image of the
industrial partner

Due to the large number of activities requiring faculty
involvement, these modifications will be better suited for
departments with a small number of students. A larger number
of students would require too much faculty time investment,
negatively affecting research.

SUMMARY

It is clear that industrial experience in any program format
can be beneficial to students. Some programs may have more
benefits than others, but each provides valuable skills and
training for our future engineers. It is important that we con-
tinue to study and evaluate these programs to make improve-
ments and adopt new ideas. What may work in one university
may not work in another. By sharing the fundamentals of the
programs, however, engineering colleges may find useful
information to improve their current programs. After all, the
main objective of any engineering education program is to
produce the best possible engineers and to develop, enhance,
and advance our society.


Summer 2006












REFERENCES
1. Huvard, G., "Make Summer Internship a Learning Experience," Chem.
Eng. Ed., 32(1), 48 (1998)
2. Fricke, A.C., "From the Classroom to the Workplace: Motivating
Students to Learn in Industry," Chem. Eng. Ed., 33(1), 84 (1999)
3. Bradburn, T., "Cooperative Education: A Key Link Between Industry
and Engineers in the Making," Chem. Eng. Ed., 35(1), 58 (2001)
4. Bendrich, G., "Just a Communication Course? Or Training for Life
After the University," Chem. Eng. Ed., 32(1), 84 (1998)
5.
6.


7.
8.
9. <1.111. .:. .. .. l, I.. .1. ,;-2006/>
10. pdf>
11.
12. Newell, J., D. Ludlow, and S. Sternberg. "Development of Oral and
Written Communication Skills," Chem. Eng. Ed., 31(2), 116 (1997)
13. ABET Criteria for Accrediting Programs in Engineering in the United
States, Engineering Accreditation Commission, Accreditation Board
for Engineering and Technology (ABET) Inc., New York (1994) 1


Chemical Engineering Education











MR, laboratoryy


Validating


THE EQUILIBRIUM STAGE MODEL

for an Azeotropic System in a Laboratorial Distillation Column









B.P.M. DUARTE, M.N. COELHO PINHEIRO, D.C.M. DA SILVA, AND M.J. MOURA
Institute Superior de Engenharia de Coimbra 3030-290 Coimbra, Portugal


During their fourth year of undergraduate studies,
chemical engineering students at the Instituto Su-
perior de Engenharia de Coimbra (ISEC), Portugal,
take a full laboratory course in unit operations and process
control. The topics covered include evaporation, distillation,
absorption in packed columns, solid-liquid extraction, dry-
ing, and the control of variables often found in industrial
units (e.g., pressure, flow, level, and temperature) employing
laboratory units or bench-scale kits. The course's basic aim is
the practical demonstration of theoretical concepts taught in
courses on process separation, chemical thermodynamics, and
process dynamics at laboratory scale. It also provides students
with experience in operating and controlling complex units.
Regarding the work on distillation, the students are asked to
validate the steady-state behavior of a laboratory unit used
for separating an azeotropic mixture of aniline and water. The
interest in this binary system arose from an intensive research
program carried out in the chemical engineering department
of ISEC in collaboration with a Portuguese company that
produces aniline-with the aim of optimizing the aniline
production section. In addition, this work also aims to vali-
date knowledge relating to the distillation of heterogeneous
azeotropic mixtures, of which the aniline-water system is a
simple and easily handled example.
The conceptual basis used to describe the phenomena
involved in a distillation column is the equilibrium stage
model, which assumes thermodynamic equilibrium between


the species and perfect mixing in each phase and tray.1, 2]
But it is quite common for equilibrium not to be achieved in
vapor-liquid-liquid dispersions arising from heterogeneous
azeotropic mixtures. This is because of the occurrence of
kinetically controlled phenomena, such as mass transfer,
coalescence, and nucleation in liquid phases. Regardless of

Belmiro Duarte is auxiliary professor in the chemical engineering
department of Institute Superior de Engenharia de Coimbra, Portugal.
He graduated in chemical engineering from the University of Coimbra,
Portugal, in 1991. He received his Ph.D. in chemical engineering from the
same university in 1995, and the M.Sc. degree in business management
from Coimbra in 1998. His research interests include simulation, process
optimization, and modeling.
Maria Nazare'Coelho Pinheiro is coordinator professor in the chemical
engineering department of Institute Superior de Engenharia de Coimbra,
Portugal. She graduated in chemical engineering from the University of
Porto (FEUP), Portugal, in 1986. She received her Ph.D. in chemical
engineering from the same university in 1994. Her research interests
include mass transfer and hydrodynamic characterization in bubbling
columns.
Dulce Cristina Martins da Silva is researcher at CIEPQPF. She gradu-
ated in chemical engineering in 1996, and received her Ph.D. in chemi-
cal engineering in 2006, both from the University of Coimbra (FCTUC),
Portugal. Her research interests include simulation, process optimization,
and control.
Maria Jose' Moura is assistant lecturer in the chemical engineering
department of Institute Superior de Engenharia de Coimbra, Portugal.
She graduated in chemical engineering from the University of Coim-
bra (FCTUC), Portugal, in 1993 and received her M.Sc. in chemical
processes from the same university in 1999. Her research interest is in
biomaterials.


Copyright ChE Division of ASEE 2006


Summer 2006










the mismatch of theoretical assumptions and real behavior,
the equilibrium stage model is still used to represent the
operation of distillation columns phenomenologically, and
several research groups have presented data for three-phase
distillation experiments to validate it.3 61 Thus, this work aims
simultaneously to enable students to gain experience in operat-
ing and analyzing a distillation procedure, to use vapor-liquid
equilibrium (VLE) prediction methods, and to contribute to
the research community's efforts to validate the equilibrium
stage model for the aniline-water system.
Three four-hour sessions are required to complete the work.
The first is devoted to a review of the basic theory behind
mixture thermodynamics and methods for predicting activ-
ity coefficients for VLE, analysis of the column layout, and
understanding its operation and control. Between the first and
second sessions, the students develop an Excel workbook to
obtain the VLE prediction for the aniline-water system at
atmospheric pressure. Each group of three or four students
is asked to use a different prediction method for the activ-
ity coefficients among UNIFAC, UNIQUAC, two-constant
Margules, and van Laar,<71 then to compare the results with
experimental data published in the literature for the aniline-
water system at atmospheric pres-
sure.[8,Appendix G] During the second
session, the students carry out
the experimental work with the
distillation column, running it
until steady-state conditions are
reached. Next, they use Aspen
Plus v. 11.1 to compute the unit feed l
steady state with an equilibrium
stage-model-based module. Fi- head temperature
sensor
nally, in the third session, the
results are presented, compared
with those given by simulation,
and discussed.


EXPERIMENTAL EQUIP-
MENT AND SAMPLING
Aniline, even in low concentra-
tions, is a fairly toxic aromatic
hydrocarbon. Its handling there-
fore requires students take some
safety precautions, that is, to wear
protective clothing, gloves, and
eye/face protection to avoid skin
contact. Furthermore, the work
is carried out in a well-ventilated
area to reduce the toxicity risk.


Figure 1. Schematic
diagram of laboratory unit
used for experiments.


The schematic diagram of the column where experiments
are performed is shown in Figure 1. The unit is formed by
three column sections, each with 10 bubble-cap trays, a total
condenser, and a kettle-type reboiler (reference Labodest
250 from Fischer). The column body is made of glass with
an internal diameter of about 0.05 m.
The feed stream enters in tray 11, and the reboiler, which
is the last tray of the column, operates simultaneously as a
heat exchanger and equilibrium tray, since it contains an
electrical-resistance element with maximum heating capac-
ity of 3 kW.
The schematic representation of trays is shown in Figure
2. The net area for vapor flow (6) is 10% of the total area of
the tray. Two cavities for liquid inlet (1) and outlet (2), with
slightly different heights, separated by a wall (3), form the
bubble caps. The liquid falls from the upper tray into the reflux
zone (4) through the downcomer, circulates around the wall
to the opposite position, and is discharged to the lower tray
through the tube (5). This design increases the contact time
between the vapor and liquid phases.
A solenoid valve controls the reflux flow to the column


Chemical Engineering Education


feed heater -






bottom product
receiver


control unit











(Figure 1). When the valve is open, a glass stem is pulled
magnetically from the seat and the liquid from the condenser
is collected in the distillate receiver. When the valve is closed,
the stem is pushed back, the seat is closed, and, when it is
partially filled, the liquid overflows through a side tube and
returns to the column (Figure 3). The reflux ratio value is set
by employing two time preselectors on the control device, thus
establishing the valve dead band. For the aniline-water binary
system, however, the control of the distillation column based
on the reflux ratio causes severe problems due to the forma-

TABLE 1
Density of Aniline-Water System for Different Temperatures[r Appndx G]
density (kg/m3)
temperature (C) aqueous phase organic phase
20 999 1023
30 997 1014
40 995 1006
50 991 998
60 987 989
70 982 982


Figure 2. Schematic representation of column trays.
1. Cavity for liquid inlet. 2. Cavity for liquid outlet.
3. Wall. 4. Reflux zone. 5. Downcomer to the lower tray.
6. Vapor flow trajectory.


valve open valve
closed


from condenser from condemns
from condenser


glass stcm
al organic
phsLe (hav
retained here


valve t to distillate to column
receiver

Figure 3. Schematic detail of the solenoid valve
operation: open and closed positions.

Summer 2006


tion of two immiscible liquid phases in the condenser-the
organic being denser than the aqueous at the temperature at
which it leaves the condenser. The difference in density means
that, when the valve opens after being closed for a while, a
small volume of liquid retained in the valve seat (below the
side tube level) is poured into the distillate receiver. This
produces significant changes in the distillate composition
since it is rich in the heavier phase (the organic phase). Table
1 indicates the density values for the aqueous and organic
phases formed at different temperatures for the aniline-water
system.[8. Appendix G]
To overcome this drawback, instead of reflux ratio adjust-
ment, an on-off control scheme based on the boiling point limit
in the upper tray of the column is employed, with the head
temperature measured through a resistance thermometer (Fig-
ure 1). Whenever it exceeds the value selected at the control
unit (98.7 oC-the azeotropic temperature), the distillate flow
rate is automatically interrupted, and the condensate starts to
flow back to the column. Consequently, the temperature of the
upper tray starts to fall, and when it reaches the preselected
value the distillate starts to be collected again. This control
scheme leads to small changes in the position of the solenoid
valve, which is open most of the time.
The feed stream is preheated to 99.3 OC before entering
the column in liquid state at atmospheric pressure. The feed
stream is sampled for the quantitative determination of aniline
concentration in the mixture.
When the steady state is reached, the students collect the
distillate and bottom products in graduated receivers for a
period of 90 minutes. At the end, the distillate is trans-
ferred to a separatory funnel, and after the separation of
the two phases, the volume of each layer is measured in
graduated cylinders.
The aniline concentrations in the aqueous phase of dis-
tillate, in the bottom product, and in the feed sample are
quantitatively determined by spectrophotometry, after
appropriate dilution. The absorbance of the solutions is
measured at 279.5 mn in a UV-Vis spectrophotometer unit
after setting the calibration line. The quantitative analysis
of water concentration in the organic phase of the distillate
is determined by titration with Karl Fisher reagent.
For the range of flows used in the experiments, the col-
umn manufacturer indicates that the Murphree efficiency
is 92%. The experimental confirmation of this value is not
performed exactly because of the column configuration,
which does not allow the extraction of liquid samples in
consecutive trays.

MODEL VALIDATION
Aspen Plus v. 11.1 is used to validate the equilibrium
stage model; the steady-state flows, compositions, and
temperatures of the upper tray and bottom stream result-











ing from the model solution are compared with laborato
data. The Radfrac module is chosen to describe the colur
unit since it is based on the rigorous solution of the equili
rium stage model for multistage vapor-liquid fractionatic
steady-state operations.[9 10] The model consists of a set
nonlinear algebraic equations, comprising the material b;
ance (M) and thermodynamic equilibrium relation (E) for ea
component and tray, and the summation of mole fractio
(S) and enthalpy balance (H) for each tray, generally call
MESH equations."111
This system can be augmented with the trays' hydrau
relations and pressure-drop profiles across the column wh
the unit geometry is known. The broader generality of t
Newton-Raphson algorithm led this to be chosen to sol
Radfrac module in rating mode. The Newton-Raphson
gorithm implemented is based on the classic Naphtali a
Sandholm algorithm.[12 First, the number of equations a
variables resulting from unit modeling is reduced through t
condensation of mole fractions, liquid, and vapor flows in
new variables representing component molar flows. Ne
the complete set of variables is ordered, and the result
algebraic equation system solved iteratively by employ
a Newton-type algorithm. The convergence is checked afi
each iteration by comparing the sum of squares of all va
ables, conveniently weighted by scale factors, with a toleran
defined as a function of the number of degrees of freedom t
system involves and inlet flows.
An azeotropic convergence al-
gorithm is chosen to handle the
current binary system that forms
a minimum-boiling azeotrope
in the region of low aniline 200
concentrations. The operating
conditions, including the molar 180 -
flow of distillate stream, the heat
consumed in reboiler, the stages 160
at which streams enter/leave the
unit, the pressure-drop profile
across the column, additional
information regarding the con-
denser operation, and the char- 120
acterization of the second liquid _
phase-formed essentially by an- 100
E
iline-are introduced into Aspen _
Plus. Molar flows are set equal 80
to the experimental steady-state
values, and the pressure drop is -

Figure 4. Comparison of ex-
perimental data from
literature, Appendix G] Vs. VLE
data determined 20 -
through UNIFAC for 0 0.1
aniline-water system.


ry disregarded due to the small flows involved in the operation.
an The characteristics of the feed stream, including its tempera-
b- ture, pressure, and molar composition, are also entered into
n, Aspen Plus. It is considered that there is no sub-cooling in the
of condenser and the Murphree stage efficiency is set to 92%.
al- The Appendix at the end of this paper presents the Aspen Plus
ch Input Summary file for a successfully converged model.
ns
ed RESULTS
The first step in analyzing the results is to compare the
lic VLE data calculated by the Aspen Properties module with
en the prediction obtained by students. Next, both must be
he validated with experimental data. The diagram in Figure 4
ve shows the agreement between the VLE data predicted using
al- the UNIFAC method and the experimental data published in
nd the literature.8 Appendix G] The system presents a heterogeneous
nd azeotrope at 98.7 C and 0.044 of aniline mole fraction, where
he three phases are in equilibrium: a vapor phase and two liquid
[to phases [an organic phase with 30.3% (mole/mole) water, and
xt, an aqueous phase with 98.6% water].
ng This happens because the vapor-liquid envelope overlaps
ng the liquid-liquid envelope, as illustrated in Figure 4. This
ter task allows students to understand that for heterogeneous
ri- azeotropes the vapor formed during boiling has the same
ce composition as the overall liquid, but the three phases in
he equilibrium have distinct compositions, contrary to what
I


'- ~~M 0 O a 0S9 0.505 0.9as o as o. 0 0asc
o a. M frliGior 1O


- T-x (P=1 atm) UNIFAC prediction
- T-y (P=1 atm) UNIFAC prediction
O T-x (P=1 atm) literature data (L-V)
+ T-y (P=1 atm) literature data
+ T-x (P=1 atm) literature data (L-L)


I I I I I I I I I


0.2 0.3 0.4 0.5 0.6 0.7 0.8
Molar fraction of H20


0.9 1


Chemical Engineering Education


} /
7-

I -














happens for homogeneous azeotropes, where the liquid and
vapor formed have the same composition.
The VLE prediction methods using UNIQUAC, two-con-
stant Margules, and van Laar activity-coefficient models
require binary parameters for the aniline-water system, which
students estimate by employing the following procedure and
theoretical basis.
For a binary system containing two liquid phases and one
vapor phase in equilibrium, the fugacities of each compound
in each of the phases are equal. That is:


orgy org
Yas PaXas
fugacity of
aniline (a) in the
organic phase


Org p org

fugacity of
water (w) in the
organic phase


aqp* aq
yas -a Xas
fugacity of
aniline (a) in the
aqueous phase


= aqP*X =

fugacity of
water (w) in the
aqueous phase


P Ya
fugacity of
aniline (a) in the
vapor phase



P Yw
fugacity of
water (w) in the
vapor phase


when the vapor phase behaves like an ideal gas mixture. The
expressions yg and yai represent the activity coefficients of
component i in the organic and aqueous phases, respectively;
xorg and x1q are the molar fractions of component i in the
organic phase and aqueous phases, respectively; Pi is the


TABLE 2
Values of Binary Interaction Parameters for UNIQUAC,
Two Constant Margules, and van Laar VLE
Prediction Models
binary interaction parameters
model A (J mol 1) Aa (J mol 1)
UNIQUAC 1524 575
two-constant Margules 6247 5836
van Laar 13164 4798


vapor pressure of the pure component i at temperature T; P
is the pressure; and y, is the mole fraction of component i in
the vapor phase. The subscript s stands for saturated liquid
phases.
According to the phase rule, a binary system with three
phases in equilibrium has just one degree of freedom, which
means that by fixing the pressure (atmospheric pressure) the
system becomes determined. Setting the activity coefficient
model, the functional forms of ya and Yw can be explicitly
written, and the preceding equations lead to
~ rga A .xrg) Xorg = aq() Xq
org(A Aa A awa( A..x.. xa
las \'a,w' wa' as / as as aw' wa' as as

(3)

and
rgA A oxrg rg iaq A x *aq
Yws (Aa. w.aw )w ws ywsA aw' .a' A sws

(4)

where Aw and A are the binary interaction parameters of the
model chosen for the aniline-water system. When the liquid-
org aq
liquid equilibrium data is available, the fractions Xa Xa,
xs, and xa can be used to evaluate the two parameters
A and A Solubility data of aniline in water and of water
in aniline for a temperature range of 20 C to 100 OC can
be found in the reference,8, Appendx G] thus allowing students to
estimate mutual solubility values at the azeotrope temperature
(98.7 C). The parameters obtained are used to calculate
activity coefficients for subsequent vapor-liquid equilibrium
calculations in the regions of 0 Table 2 lists the parameters obtained by the students for the
UNIQUAC, two-constant Margules, and van Laar activity
coefficient models.
The results of the Radfrac module are compared with
experimental results in Table 3, showing the agreement


TABLE 3
Comparison of Experimental Data with Model Predictions

Operating Conditions
Feed: Distillate:
Flow: 2.623 1/h = 2.619 kg/h Molar flow: 17.72 mol/h
Molar flow: 132.22 mol/h (100% liquid)
Temperature: 99.3 C Pressure: 1 atm Reboiler:
Aniline mole fraction: 4.53x103 Heat duty: 335 W

Laboratory data RadFrac results
Distillate: Distillate:
Temperature: 98.7 C Temperature: 98.98 C
Aniline mole fraction: 0.0353 Aniline mole fraction: 0.0338
Reflux ratio: 0.645
Bottom product: Bottom product:
Molar flow: 114.63 mol/h Molar flow: 114.50 mol/h
Temperature: 101 C Temperature: 100.02 C
Aniline mole fraction: 1.14x106 Aniline mole fraction: 1.16x106

Summer 2006 19S











Block B1 (RadFrac) Profiles Compositions


Stage

Figure 5. Composition profiles of aniline across the column.


between the model's prediction and experimental data, thus
validating the equilibrium stage model for the aniline-water
system. Once the heat duty in the reboiler is introduced into
the model, in rating mode, it is able to determine the reflux
ratio, which just represents an average value since, with the
control strategy implemented, it varies discretely between 0
and oo, depending on the valve position. Indeed, such a value
represents a possible set-point value, if the control scheme
was based on the reflux ratio.
Using the Profiles form, it is possible to view the results
from Radfrac as compositions, temperatures, and flow rates
for each column tray. Figure 5 shows the profiles of aniline
composition in both liquid and vapor phases across the col-
umn. As expected, the aniline concentration in both liquid and
vapor streams increases from the bottom to the top of the col-
umn, with the aniline mole fraction in the vapor always being
greater than that in the liquid since the feed stream is located
to the right of the azeotropic point of the VLE diagram (see
Figure 4). The composition profiles provide strong evidence
that the number of trays is over-projected for the experimental
conditions tested. Indeed, in some of the stages the enrich-
ment of vapor phase in aniline is quite small, thus leading to
the conclusion that the column can successfully separate a
higher flow of feed stream (with a similar composition) with
the same efficiency, provided the heat supplied to the reboiler
increases and flooding does not occur.
Figure 6 shows that the liquid and vapor flows across the
column are approximately constant in the enrichment and


o -0-- Liquid flow
E -- Vapor flow

0
-J
UJ
-o



OOOO Condenser Reboiler

4 6 8 101214161820222426283032
Stage

Figure 6. Liquid and vapor flows across the column.

stripping zones. Since stage 11 (12 in Figure 6) is the feed
stage and the state of the stream is saturated liquid, the liquid
flow rate in the stripping zone is increased by an amount equal
to the feed flow rate. These conditions enable the McCabe-
Thiele graphical construction'13 to be used to estimate the
number of theoretical trays required to perform the separation.
Typical values achieved by students are about 14 trays plus
the condenser and the reboiler.


Chemical Engineering Education


---- Liquid (mole frac) ANILINE
- Vapor (mole frac) ANILINE


Reboiler


o Block COL1 (RadFrac) Profiles TPFQ













CONCLUSIONS

Operating a laboratory distillation column is a good experi-
ment for demonstrating the application of some concepts of
unit operations, vapor-liquid equilibrium prediction, and
process simulation. The experiment described in this paper
embraces a wide range of topics including process control,
chemical analysis, and numerical methods for handling
rigorous distillation models. It also enables students to gain
experience in operating and controlling a distillation unit.
Moreover, the results provide the research community with
sufficient evidence to support the validation of the equilibrium
stage model for the heterogeneous azeotropic system formed
by aniline and water. The results show that the column is too
large for the experimental conditions tested, and additional
knowledge regarding its behavior can be acquired if the inlet
flow is increased and digital temperature meters are installed
in each tray to validate the temperature profile. But the ex-
periment might be successfully applied to other azeotropic
systems, such as a benzene-mononitrobenze-water mixture.


REFERENCES
1. Seader, J.D., and E.J. Henley, Separation Process Principles, John
Wiley & Sons, New York (1998)
2. Perry, R.H., and D.W Green, Perry's Chemical Engineer's Handbook,
7th Ed., McGraw-Hill Companies, New York (1997)
3. Kovach, J.W, and WD. Seider, "Heterogeneous Azeotropic Distilla-
tion: Experimental and Simulation Results," AIChE Journal, 33, 130
(1987)
4. Cairns, B.P, and I.A. Furzer, "Multicomponent Three-Phase Azeotropic
Distillation 1. Extensive Experimental Data and Simulation Results,"
Ind. Eng. Chem. Res., 29, 1349 (1990)
5. Herron, C.C., B.K. Kruelski, and J.R. Fair, "Hydrodynamics and Mass
Transfer on Three-Phase Distillation Trays," AIChE Journal, 34, 1267
(1988)
6. Mudller, D., W Marquardt, T. Hauschild, G. Ronge, and H. Steude,
"Experimental Validation of an Equilibrium Stage Model for Three-
Phase Distillation," in Distillation and Absorption '97, Vol. 1, 149
(1997)
7. Poling, B.E., J.M. Prausnitz, and J.P O'Connell, The Properties of
Gases and Liquids, 5th Ed., McGraw-Hill, New York (2001)
8. Coulson, J.M., R.K Sinnott, and J.E Richardson, Coulson and
Richardson's Chemical Engineering, Vol. 6: Chemical Engineering
Design, 3rd Ed., Butterworth-Heinemann, Oxford (1999)
9. AspenTech, Aspen Plus 11.1. Unit Operation Models, Aspen Technol-
ogy, Inc., Cambridge, MA (2001)
10. Seider, W.D., J.D. Seader, and D.R. Lewin, Process Design Principles.
Synthesis, Analysis, and Evaluation, John Wiley & Sons, New York
(1999)
11. Biegler, L.T., I.E. Grossmann, andA.W Westerberg, Systematic Meth-
ods of Chemical Process Design, Prentice Hall PTR, Upper Saddle
River, NJ (1997)
12. Naphtali, L.M., and D.P Sandholm, "Multicomponent Separation
Calculations by Linearization," AIChE Journal, 17, 148 (1971)
13. McCabe, W, J.C. Smith, and P Harriott, Unit Operations of Chemical
Engineering, 6th Ed., McGraw-Hill, New York (2001)

APPENDIX

AspenPlus Input Summary file

Input Summary created by AspenPlus Rel. 11.1 at 18:15:48


Sat., Jan. 28, 2006.
DYNAPLUS
DPLUS RESULTS=ON
TITLE 'Coluna do Isec'
IN-UNITS SI
DEF-STREAMS CONVEN ALL
SIM-OPTIONS
IN-UNITS ENG
SIM-OPTIONS NPHASE=3 ATM-PRES=1.
PARADIGM=EO
ACCOUNT-INFO USER-NAME= "BELMIRO DUARTE"
DATABANKS PURE11 /AQUEOUS /SOLIDS /INORGANIC / &
NOASPENPCD
PROP-SOURCES PURE11 /AQUEOUS /SOLIDS /INORGANIC
COMPONENTS
WATER H20 /
ANILINE C6H7N-1
FLOWSHEET
BLOCK C1 IN=FEED OUT=DESTIL RESID
PROPERTIES UNIFAC
PROPERTIES NRTL / UNIQUAC
PROP-DATA NRTL-1
IN-UNITS SI
PROP-LIST NRTL
BPVAL WATER ANILINE 2.238300000 362.5433000 .3000000000
0.0 &
0.0 0.0 372.1500000 441.1500000
BPVAL ANILINE WATER -.8969000000 509.3646000 .3000000000
&
0.0 0.0 0.0 372.1500000 441.1500000
PROP-DATA UNIQ-1
IN-UNITS SI
PROP-LIST UNIQ
BPVAL WATER ANILINE .6554000000 -168.0642000 0.0 0.0 &
372.1500000 441.1500000
BPVAL ANILINE WATER -.4676000000 -172.2809000 0.0 0.0 &
372.1500000 441.1500000
STREAM FEED
SUBSTREAM MIXED TEMP=99.3 PRES=I &
MASS-FLOW=2.427 MAXIT=100
MASS-FRAC WATER 0.977 / ANILINE 0.023
BLOCK C1 RADFRAC
PARAM NSTAGE-32ALGORITHM=STANDARD EFF=MURPHREE
&
INIT-OPTION=STANDARD MAXOL=100 TOLOL=0.0001
JMETH=INIT &
LL-METH=GIBBS NPHASE=2 DAMPING=NONE
COL-CONFIG CONDENSER=TOTAL REBOILER=KETTLE
FEEDS FEED 12 ON-STAGE
PRODUCTS RESID 32 L / DESTIL 1 L
P-SPEC 1 1. / 2 1.
COL-SPECS QN=335. MOLE-D=17.72
14
SC-REFLUX OPTION=0
STAGE-EFF 1 0.92 / 2 0.92 / 3 0.92 / 4 0.92 / 5 &
0.92 / 6 0.92 / 7 0.92 / 80.92 / 9 0.92 / 10 &


Summer 2006


201












0.92 11 0.92/ 12 0.92/ 13 0.92/ 14 0.92/ &
15 0.92 16 0.92 17 0.92 18 0.92 19 0.92 &
20 0.92 21 0.92 22 0.92 23 0.92 24 0.92 &
25 0.92 26 0.92 27 0.92 28 0.92 29 0.92 &
30 0.92 31 0.92 32 0.92
T-EST 1371.1 / 2 372.3 /3 372.6 /4 372.6 / 5 &
372.6 / 6 372.6 / 7 372.6 / 8 372.6 / 9 372.6 / &
10 372.6 / 11 372.6 / 12 372.6 / 13 372.8 / 14 &
372.9/ 15373./ 16373./ 17373.1 / 18373.1 / &
19 373.1 20 373.1 21373.1 22 373.2 23 &
373.2 24 373.2 25 373.2 26 373.2 27 &
373.2 28 373.2 29 373.2 30 373.2 31 &
373.2
L-EST 1 3.938E-006 /2 3.98E-006 /3 3.984E-006 / 4 &
3.985E-006 5 3.985E-006 6 3.985E-006 7 &
3.985E-006 8 3.985E-006 9 3.985E-006 10 &
3.985E-006 11 3.985E-006 12 4.364E-005 13 &
4.366E-005 14 4.367E-005 15 4.368E-005 16 &
4.369E-005 17 4.37E-005 18 4.37E-005/ 19 &
4.37E-005 / 20 4.371E-005 21 4.371E-005 / 22 &
4.371E-005 23 4.371E-005 24 4.371E-005 25 &
4.371E-005 26 4.371E-005 27 4.371E-005 28 &
4.371E-005 29 4.371E-005 30 4.371E-005 31 &
4.371E-005
V-EST 1 0. /2 8.86E-006 /3 8.902E-006 / 4 &
8.907E-006 /5 8.907E-006 /6 8.907E-006 / 7 &
8.907E-006 8 8.907E-006 / 9 8.907E-006 / 10 &
8.907E-006 11 8.907E-006 12 8.907E-006 13 &
8.927E-006 14 8.945E-006 15 8.959E-006 16 &
8.971E-006 17 8.98E-006 18 8.986E-006 19 &
8.99E-006 / 20 8.993E-006 21 8.996E-006 22 &


8.997E-006 23 8.998E-006 24 8.999E-006 / 25 &
8.999E-006 26 8.999E-006 27 9E-006 / 28 9E-006 / &
29 9E-006 / 30 9E-006 / 31 9E-006
BLOCK-OPTION FREE-WATER=NO
EO-CONV-OPTI
SENSITIVITY S-1
DEFINE Z1 BLOCK-VAR BLOCK=C1 VARIABLE=RR
SENTENCE=RESULTS
DEFINE Z2 BLOCK-VAR BLOCK=C1 VARIABLE=VRATE &
SENTENCE=PROFILE ID1=2
TABULATE 1 "Z1"
TABULATE 2 "Z2"
VARY BLOCK-VAR BLOCK=C 1 VARIABLE=QN SENTENCE=COL-
SPECS
RANGE LOWER="250" UPPER="350" INCR="10"
SENSITIVITY S-2
DEFINE Z1 BLOCK-VAR BLOCK=C1 VARIABLE=RR
SENTENCE=RESULTS
DEFINE Z2 BLOCK-VAR BLOCK=C1 VARIABLE=VRATE &
SENTENCE=PROFILE ID1=2
TABULATE 1"Z1"
TABULATE 2 "Z2"
VARY MASS-FLOW STREAM=FEED SUBSTREAM=MIXED
COMPONENT=ANILINE
RANGE LOWER="5.0E-6" UPPER="4.0E-5" INCR="2.5E-6"
CONV-OPTIONS
PARAM TEAR-METHOD=NEWTON SPEC-LOOP=INSIDE
STREAM-REPOR MOLEFLOW MASSFLOW
PROPERTY-REP PCES NOPARAM-PLUS 1


Chemical Engineering Education











[MnI classroom
----- s______________________________________


A TIRE GASIFICATION

SENIOR DESIGN PROJECT

That Integrates Laboratory Experiments

and Computer Simulation








BRIAN WEISS AND MARCO J. CASTALDI
Columbia University New York 10027


The Accreditation Board for Engineering and Tech-
nology (ABET) requires that students in accredited
engineering programs complete certain requisites to
graduate. One constituent is engineering design, "the process
of devising a system, component, or process to meet desired
needs,"[1] specifically Criterion 4-Professional Component,
which is particularly focused on a major design project in-
corporating appropriate engineering standards and multiple
realistic constraints. To fulfill the program, the Department
of Earth and Environmental Engineering at Columbia Uni-
versity allows undergraduate seniors the opportunity to work
independently under the supervision of a faculty advisor. The
faculty advisor's purpose is to guide the student's activities
and ensure progress.
The student typically decides on a topic area, such as waste
to energy, in consultation with a faculty member. Once the
overall project area is identified, a rigorous task plan and
schedule are given to the student to begin the design effort.
Clearly, this plan must be consistent with any guidelines
outlined by the department (e.g., midsemester report, final
presentation). In this case, Figure 1 (next page) details the
efforts to be undertaken by the student and that are aligned
with the department's requirements of a fall and spring term
presentation (not shown) and the final report.
One recent student's project involved the conversion of
waste tires by thermal treatment to either energy generation
or chemical synthesis. The field of study was selected based
on the expertise of the mentoring professor and capabilities of
Summer 2006


the laboratory. Weekly meetings between the student and pro-
fessor were arranged and a schedule of activities was drafted
to facilitate progress toward the design. The first semester
(fall) concentrated on researching the topic, creating a design,
and justifying the initial feasibility of the design with eco-

Marco J. Castaldi is an assistant pro-
fessor in the Earth and Environmental
Engineering Department at Columbia
University. He received his B.S. ChEfrom
Manhattan College and M.S. and Ph.D.
ChE from the University of California, Los
Angeles. Prior to joining Columbia Uni-
versity, he worked in industry for seven
years researching and developing novel
catalytic reactors. His teaching interests
lie in thermodynamics, combustion phe-
nomena, and reaction engineering. His research is focused on
beneficial uses of CO2 in catalytic and combustion environments,
waste-to-energyprocesses, and novel extraction techniques for
methane hydrates.
Brian Weiss received his B.S. from
the School of Engineering and Applied
Sciences at Columbia University in
spring 2005. The work with tire gasifi-
cation culminated his education in the
Department of Earth and Environmental
Engineering. Currently, he looks forward
to pursuing similar projects in efficient
chemical conversion as a chemical
engineering graduate student at the
University of California, Berkeley.


Copyright ChE Division ofASEE 2006


203












Plan for Fall Semester
0) Motivation (September 22)
a. Market demand
b. Environmental/Economic/
Global impacts/benfits
c. Type of waste
1) Previous work (October 15)
a. Patent office (Next Week)
(www.USPTO.gov)
b. Textbooks
c. General literature
(journals)
d. Company info
2) Identify most promising processes
(November 15)
a. Product yields
b. Feedstock accessed
c. Prototype built
d. Economics
e. Energy balance
3) Understand (December 10)
a. Flow diagram
b. Chemistry and reactions
c. Thermodynamics
d. Pitfalls
e. Lab work
4) Improve (Ongoing)
a. Imagination
b. Lab work


Plan for Spring semester (Due dates in bold):

0) Review of tire pyrolysis and
combustion characteristics. (Jan 14)
a. Literature search

1) Materials list for Prototype (Jan 21)

2) Lab Work (Feb 25)
a. Combustion TGA, micro GC
(Jan 28)
i. Test Plan (Jan 14)
ii. Kinetics
iii. volatile fraction; CO2/CO;
optimum atmosphere
b. Gasification Same apparatus
(Feb 25)
i. Test Plan (Jan 28)
ii. w/ H20 (g + 1); CO2, CO,
air
iii. GC/MS product analysis
iv. Effect of Temperature;
optimum atmosphere;


3) Computer simulation (Apr 8)
a. Aspen Plus
b. Use TGA data
c. Team with Kimberly?

4) Prototype build (Apr 29)
a. Wire and Cellophane
b. Smoke experiments


5) Final Report (May 2)


Notes to plan:

The prototype build might run overtime
depending on the complexity of the project. The
priorities will be physically observing the design,
i.e. the lab work and the prototype build. The
main benefit of a computer simulation could be a
comparison with the models from last semester
so it need not be the focus.


nomic and thermodynamic calculations. The second semester
(spring) focused on executing laboratory work necessary to
provide data for input to a theoretical modeling of an overall
system. Typically an industry-accepted modeling package is
used to conduct simulation and analysis of an overall process.
The package used during this project was Aspen Plus v. 12.1
by Aspen Technology, Inc.
Other design projects in which students combined ex-
perimental work and computer simulation have demonstrated
positive results. In such projects, students can gain a profound
understanding of industry in a more stimulating setting than


a lecture class. In one project, students were instructed to use
theory and experimentation to execute a boric acid dehydra-
tion process. The result won recognition from education
competitions and Borax Europe, Ltd., an industry leader.[2]
In another case, students were prompted to design, fabricate,
and test single-component, mechanical products.[3] In both
projects, an emphasis was made to complete the curriculum
within one or two semesters. In the latter, computer aided
design was necessary for compressing the work into the
desired time frame. The success of each example project
lay in the ability to organize a broad scope of activities into


Chemical Engineering Education


Figure 1.
The student and
professor drafted
a schedule of
activities at
the beginning
of each semester.
The schedule
followed the
guidelines given
by the academic
department.










manageable blocks.
The first step of the present design project was for the
student and professor to propose a schedule of activities
ensuring the student gained exposure to all aspects of the
design experience. Shown in Figure 1, this type of work plan
allows the student to understand the usefulness of preliminary
calculations in guiding subsequent work. All segments of the
plan were executed. The prototype build was attempted but not
completed, however, because unanticipated opportunities to
present the work arose and demanded more effort be devoted
to refining what was already accomplished. The remaining
sections of this paper are taken from work done by the student
during the two-semester design project.
This paper is an example of how to integrate environmental
issues such as pollution prevention, reuse, and recovery into
the design experience. While this particular project did not
use life-cycle assessment methods and tools, it could be a very
worthwhile effort to employ them to explore the outcomes for
waste tire and other waste-to-energy processes.

BACKGROUND WORK
The project began with a literature survey of academic
papers, government programs, and business efforts. Currently
in the United States, 290 million waste tires are generated an-
nually. Additionally, nearly 300 million tires (~ 6 million tons)
reside in environmentally unsound stock piles. Accordingly,
government impetus has created a market for waste tires that
currently awards tipping fees to users between $50-100 per
ton.[41 Discarded tires have applications as a co-combustion
fuel and ground rubber fills in construction materials. Only
9% of the tires currently generated in the United States go
to landfills. Current applications for tires are both economi-
cally practical and regulated by national standards, but initial
research showed opportunities for improvement. [5 Most tires
are used as supplementary feed in processes for which they
were not explicitly designed. Since many advantages can be
realized from a method specifically adapted to the feed, the
purpose of the project became to design a system specific to
scrap tires.
Since tires possess several distinct qualities as a fuel, ther-
mal processing may offer a broad range of opportunities to
improve existing practices. In the United States, three facili-
ties have been established to convert exclusively scrap tires
to electricity: Exeter Energy in Connecticut, Modesto Energy
in California, and the Ford Heights facility in Illinois. Each
facility was built in the early 1990s with a capacity to handle
8-10 million scrap tires per year on conventional equipment
to produce 25-30 MW of electricity. Only one remains in
business, however, indicating that their profitability is, in the
current environment, only marginal.[61 It is anticipated that,
with rising energy and landfill prices, an innovative approach
may enable a more successful business.


PRELIMINARY DESIGN AND ECONOMICS
Before drafting a reactor design, the economics and thermo-
dynamics of the process were investigated using information
found in the literature. These steps were intended to provide
a "first-cut" analysis and set the boundaries of feasibility and
profitability. A process was proposed with unit operations
including: a standard thermal reactor; an electricity-produc-
ing system with boiler, turbine, and generator; a gas clean-up
system consisting of an electrostatic precipitator, a scrubber
and a stack; and auxiliary capital (pipes, pumps, etc.). Cost-
ing estimates were obtained from a chemical engineering
textbook; size, efficiency, and inflation were incorporated
into the model.[7' Tires consist of a mixture of rubber polymer
(CsH8,) and carbon black with added fillers such as light oils,
fibers, trace metals (zinc), and a steel wire belt. A chemical
and elemental analysis of tires was taken from a published
source and showed a combustion enthalpy of 35 kJ/g.18
The economic model was created for a process scaled to
10 million tires per year (164 thousand tons). The results
showed that 16.7% of the tire's enthalpy of combustion
could be converted to an output of 28.6 MW of electricity.
The revenue generated by the system included tipping fees
from the tires ($100 per ton) and electricity sales ($0.05 per
kWh). At an interest rate of 5%, the revenue after a 40% tax
totaled to $21.5 million per year. The net present value was
$201.5 million, which annualized to $8.4 million per year.
The profits were $13.1 million per year indicating an internal
rate of return of 19%. These results include all relevant parts
such as working capital, debt servicing, permitting, and cit-
ing. The experiences of the other facilities combusting tires
support the estimate.
Although the process proves profitable, the return is lower
than most investors would prefer for new technology, indicat-
ing that a technological breakthrough is necessary to secure
the business. Several reactors for converting waste-to-energy
were researched including fixed beds, moving beds, fluidized
beds, and rotary kilns.9, 10] A fixed-bed type reactor was select-
ed for its cost effectiveness, ease of use, and appropriateness
to the feed. Additionally, it was proposed that the end product
would be syngas (primarily CO and H,), which was thought
to enable more controllable conditions. Syngas can be created
by reforming the tires with CO, and H,0. Because these reac-
tions are endothermic, a heat source is required. Thus it was
proposed that the combustion of tires with air could provide
that heat to drive the reactions. Using sewage sludge as the
water source could minimize costs, extending the scope of
the design to a truly novel integrated waste converter. The
ideas behind the design evolved by employing the principles
of process intensification to existing technologies. Process
intensification basically is the reduction of process volumes
by combining and consolidating multiple unit operations into
one physical unit."11


Summer 2006


205










DESIGNING THE REACTOR

The reactor was sketched in StudioTools (Alias). Antici-
pated flow patterns are shown in Figure 2. Scrap tires and stoi-
chiometric air are fed separately into the combustor through
an annular pipe. Tires fall and combust on a ceramic plate
similar to a grate-type combustion system. The wall of the
combustor has two layers: an outer impermeable steel barrier
and an inner screen within which primary air emerges. This
concept was adapted from the gas turbine industry to maintain
moderate metal temperatures of the combustor. The purpose
of the double wall is to maintain a unidirectional flow pattern
that minimizes the amount of hot gases impinging on the wall.
Secondary air actively cools the struts supporting the ceramic
plate and assists the combustion. The combustion product
leaves the combustor from below and enters the gasifier. The
material to be gasified falls from the top in a counter-current
flow to the hot gases. Heat transfers to the gasifier across the
dividing wall of the combustor and via the enthalpy contained
in the combustion product stream. The syngas produced is
extracted through a pipe from the top or side. The ash falls
to the bottom of the reactor where it is collected and removed
using standard equipment. The reactor has similar attributes
to a previously disclosed design for a combustion-gasification
system of wood chips.[121 Nonetheless, because the system
proposed for this project was conceived independently, there
are considerable differences to make this design unique. The
design possesses capabilities beyond existing technology
because it is prepared to handle fuel with a large heating


Figure 2. A rendering of the
proposed reactor. Tires and
airflow in an annular pipe
and combust on a
ceramic plate in the center. 1 Air
The combustion product combss
& liner
exits the inner chamber from cooling)
below and mixes with
additional water, sludge, and
tires. The syngas is extracted
from a pipe in the side or on
top and the ash falls
to the bottom. Baffles ensure
the flow patterns run along
the planned routes.
Syngas


m


value and high ash content, such as tires, while beneficially
using sludge material.

MATERIAL AND ENERGY BALANCES
Material and energy balances were performed for both the
combustor and the gasifier. For a reactor that consumes 10
million tires per year (164 thousand tons), combusting 30%
in stoichiometric air and gasifying the remainder with 87,600
m3 of water per year, the results of the calculations showed
the syngas to consist of 18.9% H,, 16.6% CO, 6.0% H20,
8.4% CO,, and 49.9% N,. The total energy output is 37 MW
of sensible heat and 103 MW of chemical energy. The tempera-
tures of the combustor wall and gasifier are 1,040 C, and 614
C, respectively. A flame temperature for the combustion of
the tires was calculated to be 1,469 C. Based on a material
residence time of one hour, a combustor size of 27 m3 was
calculated with a 2.8 m base diameter.
The gasifier was sized to allow a moderate flow rate of
the syngas produced, while enabling most of the large ash
particles to settle. This led to a unit of 77 m3 that is 5.0 m
diameter and 3.8 m in height.
The success of the process will be determined by the
ability to control the material and heat flow. Because com-
bustion temperatures can run higher than the limitations of
most metals, the wall must be kept cool, which will be ac-
complished by air flow augmented by the fresh tire/sludge
mixture entering the gasifier. The double wall will create a
unidirectional air-flow pattern and minimize the impingement
of hot combustion gases onto the wall. Both the temperature


Chemical Engineering Education


Water/Tire
Mixture


Double Walled
Combustion Unit
(Tire+ 02 CO2
+ H20 + heat)



Gasification
, Unit
(C + H20 + CO2 + heat -
CO + H2)


(combustion & strut cooling)










and the syngas quality will be regulated by the composition
of the feeds: adding more tires produces more energy and
higher temperature whereas more water yields more hydrogen
in the product stream.
LABORATORY WORK
While most, and probably all, parameters could be ob-
tained from the literature, one of the advantages of having
individual design projects is the ability for students to get
hands-on experience working in a laboratory. This enables
them to generate pertinent data needed for their design. It also
forces students to think of the experimental outcomes before
doing the work, thus preparing them to develop practical test
methods for efficiently generating data. To enable realistic
engineering, basic thermodynamic and kinetic parameters of
tires were required. Some of these parameters were obtained
from equipment readily available in the faculty advisor's
combustion laboratory. An oxygen bomb calorimeter (Parr)
yielded the heat of combustion of tires. The bomb calorimeter
adiabatically combusted the tire sample at constant pressure.
The temperature rise of a water bath correlated to the enthalpy
of combustion of the tires. The enthalpy of combustion of tire
was determined to range from -33.37 to -36.33 kJ/g which
was consistent with established data.I13 14 Kinetic informa-
tion was obtained from thermogravimetric analysis (TGA).
A Netzsch TG 409 PC instrument was used to record the
mass loss of a sample as the temperature was increased at a
constant rate. Constant flow rates (100 mL/min) of air (20%
02; 80% N2) and inert purge gas (20% CO2; 80% N,) flowed
over the sample. The air enabled evaluation of combustion
parameters while the inert atmosphere was selected to re-
semble the gasification zone. A plot of the fractional weight
loss, a, versus temperature, T, showed that the sample mass
decreased as temperature increased. A derivative plot of a
showed reaction rates as peaks. At a constant temperature


ramp, the reaction rate can be described by da/dT, which
follows an Arrenhius rate law.


da dT = AeEa/RT (1_a)n


where A, Ea, and n are the Arrenhius frequency, activation
energy, and reaction order, respectively; R is the universal
gas constant, and ( is the heating rate.
A representative TGA analysis is shown in Figure 2 in
which the student had to convert that raw data to usable data
for input into a model for design simulations. Combustion of
tires in air revealed five peaks in the derivative plot implying
an equal number of reactions, which have been proposed to
correspond to light oils, natural rubber, synthetic rubber, and
tars.15s Under inert atmospheres only the first three peaks
were observed, implying that the tar only combusts in the
presence of oxygen. At higher heating rates the peaks overlap
but maintain recognizable reactions.
Because Eq. (1) cannot be solved explicitly, quantitative
parameters were derived from a method from the literature.[16]
Plotting 1/T versus

-log l-(l-a)1-n /T2 (1-n) fornl (2a)


-log[-log(l-a)]/T2


for n= 1


(2b)


reveals a linear plot for the appropriately chosen reaction
order. The Arrenhius rate parameters can be related to the
slope, m, and the intercept, b, by

E = 2.3 m R (3a)

A= 10bE/R(1- 2 R TEa) (3b)

By this method, rate parameters for air and 20/80 CO2 /N
atmospheres could be obtained as shown in Table 1.


TABLE 1
Parameters Derived from TGA Experiments
Ascertained by the Derivative Method of Fristky, et al., (1994)
The "Weight" column refers to the amount of mass change that can be
accounted for by the reaction. Under air, there was a 5% residual,
whereas under an inert atmosphere there was a 38% residual.
Reaction under Air


Weight 15% 13% 23% 20% 24%
Ea (kj/mol) 120 180 187 325 258
A(hz) 5.5x 108 1.3x 1013 2.3 x 1011 4.0 x 1019 2.7 x 1013


Reaction under 20% CO2
Reaction 1 2 3
Weight 18% 12% 32%
Ea (kj/mol) 95 203 176
A(hz) 1.4 x106 2.5 x 1014 3.0 x 1010


Summer 2006


207










The outflow of the TGA was connected to anAgilent 3000
micro-GC gas chromatograph (GC). Figures 3 and 4 show a
typical species analysis of the product gases by the GC, and
allowed the student to conduct a material balance ensuring
the integrity of the data generated. Combustion under air
revealed that CO2 was the main constituent of the exhaust.
Pyrolysis of tires under CO2 indicated that the amount of CO2
in the product gas increased slightly during the reaction and
no CO was detected. The GC proved useful for determining
the product species and can be used in future work to identify
important constituents.

SIMULATION
Following the laboratory work, the student then reduced
the data and calculated the parameters needed to input into
a simulation. There were two sets of simulations done, one
was thermodynamic and the other used kinetics obtained
from the laboratory experiments to more accurately simu-
late the combined combustor-gasifier reactor. This allowed
the student to better understand the type of information
thermodynamics can provide versus an actual operating
system where the kinetics play an important role. Results
are shown in Table 2.
Thermodynamic Simulation
The thermodynamic data obtained from literature and the
student's experiments (bomb calorimeter) were input into
Aspen. Tires were defined as a mixture of the rubber mono-
mer (C sH), graphite, iron, and zinc so that the final assay
equaled the literature value.[8] Figure 5 shows the process
flow diagram of the equilibrium simulation. The combustor
and gasifier were represented as two Gibbs reactors con-
nected by material and heat streams. The ideal "separators"
surrounding the gasifier selectively remove solids from the
streams. A calorimeter module was programmed to combust
the syngas with stoichiometric amounts of oxygen only, to
enable the student to conduct an energy balance and compare
that with initial calculations. The model output of 170 MW
total power from 18,700 kg per hour (5.2 kg per second)
indicates an energy input for tires of -33 I I1. kg of tire,
which is consistent with bomb calorimeter measurements.
The product gas is composed of 24.0% H2 and 10.8% CO.
The temperature of the gasifier is 786 C, which suggests
that the temperature of the inner wall may be maintained
at an acceptable level. The Aspen equilibrium simulation
reflects the material and energy balance calculations from
preliminary assessments.
Kinetic Simulations
Upon completion of the thermodynamic simulations,
programming of the kinetic parameters began. For this task,
the student used his own data generated from the TGA and
compared them to literature values to ensure the integrity of
the data. The TGA data from Table 1 was used for the Aspen


kinetic simulations. The kinetic simulations first attempted to
model the TGA experiments to ensure the results were consis-
tent. The simulated TGA was a semi-batch reactor with a charge
of the tire sample and a constant flow rate. Due to the kinetic
parameters' sensitivity to the phase of the reactants, the tires
were modeled as a mixture of coal and graphite-both present
in the Aspen database. The Aspen-defined coal had many similar
properties to tires. The mixture was modified so that the final
analysis of the material would have an enthalpy of -33 N I. kg


Combustion Reactions 10 OC/min
--Alr .-- Enhanced 02 (28%) -- Deficient 02 (3%)
Air Dtg Enh 02 Dtg .--- Def 02 Dtg
150 ._- 0
125

100 l -0.02 -
75
S50 -0.04 6
25
0 -0.06
200 300 400 500 600 700 800
Temp (C)

Figure 3. The results of the TGA with derivative plots show
the combustion reactions under 02 and N2
atmospheres (Air 20% 02; enhanced 0 28% 02;
deficient 02 3% 02) heated at 10 OC/min.


TABLE 2
The results of one Aspen equilibrium simulation show a sys-
tem that handles 164,000 tons of tires (~10 million tires) per
year. The power output is 170 MW for a syngas that consists
of 34.8% by volume of useful material. The temperatures of
each reactor have reasonable values.

Reactant Flow Rates Product Flow Rates
tonr/hr Nm3/hrx 103 %

Tires (total) 18.7 H2 25.8 24.0
Combusted 5.6 CO 11.6 10.8

Gasified 13.1 CO2 11.5 10.6

Water 18.0 H20 11.2 10.4

Air 97,490 Nm3/hr N2 47.2 44.0

Temperature (C) Energy (MW)

Combustor 1626 Chemical 130 (eq)
Gasifier 786 Sensible 40

Chemical Engineering Education











TGA and GC under Air
-N2/30 --02/10 x CO CO2 --TIre(TGA)
Temp (C)
165 265 365 465 565 665
125
300 0
100
0 0
S75' 200 8

50
0o 100
)25 o 2

0o 0
15 25 35 45 55 65
Time (min)
Figure 4. GC flow rates (left axis) plotted with TGA mass
loss (right axis) forflow under air. The 02 and N2 curves
are scaled by 1/10 and 1/40, respectively. CO2 is produced
throughout combustion; CO increases in the tar region.


Figure 5. Flow diagram of the primary reactor. All feeds began at room
temperature (20 oC). The calorimeter was added to measure the heating
value (HV) of the final product gas (FINPROD).


and a total chemical assay of the literature."8] For the combus-
tion reactions, the tire was separated into six components
based on the reactions determined by TGA in Table 1. The
gasification simulation used four components. The simula-
tion results of Figure 6 show that the simulation closely ap-
proximates the measured data. The correspondence suggest
that the kinetic model can be developed further to simulate
the combustion-gasification reactor. Although the ultimate
plan was to simulate the design of the integrated gasifier
and combustor, time did not permit this step. The exercise
of conducting a thermodynamic simulation of the entire
design and programming the kinetics to simulate the TGA
experiments provided the student with sufficient experience
to complete the project had time permitted. Upon comple-
tion of this task, the student was keenly aware of the many
ways to arrive at designing a new technology or modifying
an existing technology. Moreover, the student was now
well prepared to understand the importance of
experimental data generation and how to attack
such a process in the future.


OUTCOMES
The undergraduate design project taught the
student how to address engineering challenges.
The literature research aided in the design of
the combined combustion-gasification system,
which was devised solely by the student. The
professor's role was to provide insights into the
merits and limitations of such a device. Using a
more experienced knowledge base, the profes-
sor was able to recommend calculations and
experiments that would prove the design.
The independence of the student allowed
greater opportunities for learning about the busi-
ness environment surrounding waste manage-


Combustion in Air


c 4 -
E
-3 Measured
3 -

2-

1 Aspen Tm

0
200 300 400 500 600 700


Gasification in 20% CO2
6

5 Measured

3 4 \
E
" 3 -

2 AspenTM

1 -


200 300 400


Temp (C)


500 600


Temp (C)


Figure 6. The
measured data
(solid line) and the
Aspen simulation
(dashed line);
the model shows
potential for
developing further
investigations.


Summer 2006


Combustor


Calorimeter


209











ment, industrial thermal processes, reactor designs, and the
engineering process. Performing the laboratory experiments
helped the student comprehend the operation of standard
analytical tools. Further, the student began to understand
how to devise a test plan and allocate adequate time to attain
a sufficient amount of data. The Aspen simulations were in-
structive in demonstrating the next level of design execution
and evaluation. Finally, to expose the student to the experience
of communicating the work, poster or lecture opportunities
were pursued with presentations at two academic departments
in Columbia University and Barnard College, at a university-
wide undergraduate research symposium, and at an American
Chemical Society meeting.[17]
The individual responsibility for the project encouraged a
greater commitment from the student and allowed a wider
platform for innovation. The development and successful
implementation of the project, however, may have benefited
from a larger undergraduate or graduate team. Nonetheless,
the student was able to maintain the directives assigned in the
predetermined schedule and provide a report at the end of each
term. The time frame of the project allowed ample time to
understand basic aspects of designing a reactor and carry out
the preliminary steps. The student was able to identify future
directions for the project in the final report. It is antici-
pated that the student will be better prepared for future work
as a professional engineer and that the project described
herein may be continued under other circumstances.

IMPLEMENTATION
A suggested implementation strategy is briefly presented for
those who wish to augment a traditional chemical or environ-
mental engineering capstone experience with a similar effort.
Provided the scope of the project is contained and a schedule
is put forth at the beginning of the semester, a project such
as this becomes very manageable. As evidenced in Figure 1,
all elements of engineering design are covered. An important
aspect of this type of project is to have the course span two
semesters to give students time to assimilate material, develop
and design processes, and possibly build devices or conduct
limited experiments.


So as not to risk skipping or eliminating any of the critical
areas of the design process, the real task, for the student and
faculty advisor, is not to spend extensive time on any one
area. All components should be developed to the extent that
the student can see a clear path to the outcome of each task.
Most engineering departments maintain a software license
to the common programs used in industry and typically have
some level of laboratory capabilities. This leaves only the task
plan and timing to be formulated and strictly followed. For
reference, the hours spent by the student and faculty advisor on
this project were no more than a typical design course and the
cost of everything except for the software license was under
$200. Typically, software licenses are heavily discounted for
educational institutions.

REFERENCES
1. E.A. Commission, Criteria for Accrediting Engineering Programs,
ABET, Inc., Baltimore, Nov. 1, 2004, pp. 1-19
2. Shaw, A., H.N. Yow, M.J. Pitt, A.D. Salman, and L. Hayati, Trans-
actions of the Institution of Chemical Engineers, Part A, 82, 1467
(2004)
3. Moller, J.C., and D. Lee, Journal of Engineering Design, 10 (1999)
4. EPA in Management of Scrap Tires (2005)
5. ASTM in Standard Practice for Use of Scrap Tire-Derived Fuel, Vol.
D6700-01 ASTM International (2005)
6. U.S. Scrap Tire Markets: 2003 Edition, Rubber Manufacturer's As-
sociation (2004)
7. Ulrich, G., A Guide to Chemical Engineering Process Design and
Economics, John Wiley and Sons (1984)
8. Reisman, J.I., and PM. Lemieux, Air Emissions from Scrap Tire Com-
bustion, (Ed.: C.A.T. Center), EPA(1997)
9. Belgiorno, V., G.D. Feo, C.D. Rocca, and R.M.A. Napoli, Waste
Management, 23, 1-15 (2003)
10. Bridgwater, A.V., Chem. Eng. J., 91, 87 (2003)
11. Tsouris, C., and J.V. Porcelli, Chem. Eng. Progress, 99, 50 (2003)
12. Susanto, H., andA.C.M. Beenackers, Fuel, 75, 1339 (1996)
13. Gonzalez, J.E, J.M. Encinar, J.L. Canito, and J. J. Rodriguez, J. of
Analytical and Applied Pyrolysis, 58-59, 667 (2001)
14. Jones, R.M., J.M. Kennedy, and N.L. Heberer, TAPPIJ. (1990)
15. Conesa, J.A., R. Font, A. Fullana, and J.A. Caballero, Fuel, 77, 1469
(1998)
16. Fritsky, K.J., D.L. Miller, and N.P Cernansky, J. of Air and Waste
Management Association, 44, 1116 (1994)
17. Weiss, B., and M.J. Castaldi, Novel, IntegratedProcessforBeneficial Use
of Waste Tires, Poster Presentation, ACS, Washington, D.C. (2005) 1


Chemical Engineering Education











[Ml1 outreach
----- s______________________________________


Demonstration and Assessment of


A SIMPLE VISCOSITY EXPERIMENT FOR


HIGH SCHOOL SCIENCE CLASSES


T.M. FLOYD-SMITH, K.C. KWON, J.A. BURMESTER+
Tuskegee University Tuskegee, AL 36088
The objective of this demonstration and assessment was
to develop an instructional model to inform and enthuse
students about chemical engineering. Figure 1 shows
the number of B.S. degrees granted nationally in chemical
engineering.J1 2] Rhinehart observed a 13-year-cycle period
for the production of B.S. degrees in chemical engineering at
Oklahoma State University, dating back to the 1930s.[1] It is
not clear at this time if the 13-year-cycle period for chemical
engineering degrees awarded will hold,P1] but it is clear that the
peak has dropped from approximately 7,500 degrees awarded
to approximately 6,500 degrees awarded, representing a
13% decline. Rhinehart attributes the cycling to B.S. chemi-
cal engineering supply/demand being out of phase but does
not discuss the magnitude of the peaks. Halford,1[3 however,
suggests the decline is due to a rising attraction of potential
chemical engineers to the environmental engineering and
bioengineering fields. The cause of the decline in chemical
engineering enrollment has not been determined conclu-
sively, but-regardless of the cause-the effect is that when
enrollment is low, administrators may question the benefit of
maintaining an expensive chemical engineering program.11]
B.S. chemical engineers are indirectly supplied by the
nation's high schools. Therefore, one potential approach to
positively impact enrollment in chemical engineering under-
graduate programs is to conduct outreach programs for high
schools. Ross and Bayles[4] describe a method for incorporat-
ing high school outreach into chemical engineering courses.
Their goal is to provide role models for high school students
by assigning chemical engineering students enrolled in their
courses to participate in an outreach project. In contrast,
this work describes an outreach program administered and
conducted by professors for the purpose of informing high
school students about chemical engineering and attracting
them to the profession.
Copyright ChE Division of ASEE 2006
Summer 2006


, F.F. DALE+, N. VAHDAT, AND P. JONESt


8000
7000
6000
5000
4000
3000
2000
1000
0


Rhinehart
ACS



170 1980 1990 2000
Academic Year


Figure 1. Annual national B.S. chemical engineering
degrees awarded.

Tamara M. Floyd-Smith is an assistant professor in the Chemical Engineer-
ing Department at Tuskegee University. She received her B.S. in chemical
engineering from Tuskegee University and her M.S .and Ph.D. in chemical
engineering from MIT.
K.C. Kwon is a professor in the Chemical Engineering Department at
Tuskegee University. He received his Ph.D. in chemical engineering from
the Colorado School of Mines.
Jeffrey. A. Burmester is a physics teacher at Martin Luther King, Jr., High
School in Lithonia, Georgia.
Frances F. Dale is the science department head at Martin Luther King, Jr.,
High School in Lithonia, Georgia.
Nader Vahdat is the head of the Chemical Engineering Department at
Tuskegee University. He received his B.S. in chemical engineering from
the Abadan Institute of Technology, his M. S. in chemical engineering from
the University of California, Berkeley, and his Ph.D. in chemical engineering
from the University of Manchester, England.
Paul Jones recently retired as director of product supply for the Snacks and
Beverages Division at the Procter & Gamble Corporation in Cincinnati.

+: Martin Luther King, Jr., High School, Lithonia, GA
t: Procter & Gamble Corporation, Cincinnati, OH










The overall objectives of this demonstration were twofold.
First, the authors wanted to develop a presentation giving an
overview of the field of engineering with emphasis on chemi-
cal engineering. Second, the authors wanted to conduct a
simple experiment with the high school students so that they
have an opportunity to learn a chemical engineering concept
and be exposed to principles and problems that practicing
chemical engineers will expect to encounter.

PRESENTATION DESCRIPTION
The demonstration was conducted at Martin Luther King,
Jr., High School in Dekalb County, Georgia, in November of
2004. Ajunior/senior-level physics course (a chemistry course
may also be appropriate) was chosen for an introductory
presentation followed by hands-on viscosity experimentation.
Twenty-six students participated in the demonstration during
a class period of 90 minutes.
General engineering, chemical engineering, and the concept
of viscosity were discussed first. In the general discussion of
engineering, the major engineering disciplines were described
in basic terms (e.g., civil engineering was described as the
branch of engineering responsible for designing municipal
structures such as bridges and roads).
After a general discussion on engineering, the presenta-
tion was focused on chemical engineering. The facilitator
discussed the kinds of jobs that chemical engineers are re-
sponsible for and the types of engineering fundamentals that
chemical engineers study. The job areas described included
petrochemicals, intermediate chemicals, food processing,
cleaning products, plastics, and pharmaceuticals. When de-
scribing what chemical engineers study, several core examples
were included. The list of what chemical engineers study
included accounting for material flows (material and energy
balances), how fluids move (fluid mechanics), how heat is
transferred, and how materials react to create new things
(reaction engineering). The students were informed that the
viscosity experiment for the day was related to fluid mechan-
ics. During the discussion on heat transfer, the example of an
egg cooling was introduced. As expected, the students had a
good idea about how long it would take for an egg to cool un-
der different conditions (free vs. forced convection, in air vs.
in cool water) but overall were surprised that it is something
that chemical engineers expect to predict theoretically and/or
empirically. During the discussion of reaction engineering,
the example of how an antacid helps indigestion was intro-
duced. The students were aware of acid/base reactions from
their chemistry class, but again didn't realize that chemical
engineers are involved in producing the antacids (bases) that
are administered to neutralize excess stomach acid.
The presentation ended with a discussion on viscosity.
Viscosity was described as a fundamental physical property
in the study of how fluids move or how "thick" and "slip-
pery" a fluid is. Several examples including paste, pancake
212


syrup, water, and motor oil were discussed. Viscosity was
not mathematically defined during the presentation, and a
discussion on Newtonian vs. non-Newtonian fluids was
not included because the facilitators thought that it was
beyond the scope of what was appropriate for a high school
science class.

APPARATUS AND THEORY
The viscometer used for the demonstration has been de-
scribed previously.P5 Briefly, the viscometer is a tank-tube
viscometer as illustrated in Figure 2. It consists of a tank and
a vertical drain tube attached at the bottom of the tank. In
addition, a balance, a thermometer, a stopwatch, and a bottle
of water at room temperature are required for the experi-
ment. The viscosity of a fluid is inferred from the drain rate
of the fluid through the drain tube of the viscometer tank.
The drain rate is dependent on the viscosity of the fluid and
follows the behavior described in Eqs. (1) through (4). The
detailed derivations of these equations have been described
previously.[5]

n h+L 8u R2L (t) (1)
In (h+L ( R g L


h=H- R (2)
KR2

m 2 p
-In l- ( R 1(t) (3)
(H+L)R2pj 8 R'' (3)2



m* n= i (H + L)tR2p (4)


where
H:

h:
L:
g:
R,:
p:
I:
t:
R:
m:
m:


initial height of the fluid in the tank (9.3 cm, illus-
trated in Figure 3)
height of the fluid in the tank
length of the drain tube (56.4 cm)
acceleration due to gravity
equivalent radius of the tank
density of the fluid
viscosity of the fluid
drain duration
radius of the drain tube (0.0509 cm)
accumulated amount of a fluid drained from tank 5
left-side value of the viscosity equation, as shown in
Eq. (3)


During the experiment a tank with a rectangular cross sec-
tion, illustrated in Figure 3, was used instead of a tank with
a radial cross section. This modification was made because
the tank with the rectangular cross section is easier to fabri-
cate. Thus, the equivalent radius Ro was computed with the
Chemical Engineering Education











following equation

Ro = (5)


where
W: width of the rectangle (25.4 cm)
D: depth of the rectangle (3.81 cm)
The experimental procedure for determining the viscosity
of water using the tank-tube viscometer is as follows:
Fill the reservoir with water.
Set up the balance with automatic data acquisition so
that the data from the balance are input directly into
Microsoft Excel in real time. Use a sampling rate of
1/s.
Remove the end cap on the drain tube and allow the
water to collect on the balance.
After 90 s, stop the data acquisition.
Plot m* [left-hand side value of the viscosity equation
as shown in Eq. (3)] vs. t (time) and obtain the slope of
the line.
Extract from the expression of the slope as illus-
trated in Eq. (6).

gR p 1
8Y R 2) slope 6)

Measure the temperature of the water used in the
experiment.
Compare the experimental u to the literature value.
Calculate the measurement error based on a percent
difference.

In addition to the experimental procedure outlined, brief
explanations on linear regression, Microsoft Excel features,
and standard deviation (o) were provided to the class. Units
were not discussed and, due to time limitations, only one
experimental run was performed.


RESULTS AND DISCUSSION
The viscosity experiment was demonstrated using water
at room temperature. The experiment was successful with
a measurement error of ~3% which is the typical result
obtained in a simple viscosity experimental setting with the
tube-tank viscometer in the absence of a temperature regulat-
ing circulator.

Facilitator's Perception
Overall, the students were enthusiastic and attentive, sug-
gesting that the activity was structured appropriately to main-
tain the interest of a high school student. The students were
also willing to interact with and participate in the presentation
and the hands-on viscosity experimentation. The experiment
Summer 2006


could be improved by structuring it for more student partici-
pation. Ideally, there should be one station per four students
so that the students can perform the experiment themselves.
Excluding the computer and the balance, the fabrication cost
is ~$ 100 so the concept is economically feasible. Also, if time
permits, it would be illustrative to measure the viscosity of
more than one fluid. For example, in addition to measuring
the viscosity of water, one could measure the viscosity of an
alcohol and its aqueous solutions or water with the viscosity
modified by adding a second component such as sugar.
Student Survey
Students were asked to rate the following five questions on
a scale of one to 10 before and after the demonstration, where


Figure 2. Set-up of a viscosity experiment.


-- -------------------
H .


S w
4--------
---------- --------- -
------- ------------------- ---

H
h






2R------ *---
L






... ---.. .. .. ..-- ...


Figure 3. Tank-tube viscometer.

213


Rectangular
Reservoir
Tank


Vertical
Drain
Tube


imputer










one is "no knowledge" and 10 is "very knowledgeable":
1. How much do you know about engineering?
2. How much do you know about chemical engineering?
3. How much do you know about viscosity?
4. How interested are you in engineering?
5. How interested are you in chemical engineering?
The students were also asked to comment on what they
liked most about the presentation/experiment and what could
have been improved.
Survey Results
Table 1 shows the results from the survey given to the
students. The results summarize the students' knowledge
and interest before and after the presentation followed by
the hands-on viscosity experimentation. The table also shows
the difference between the two values and the statistical sig-
nificance of the results. Using a paired-sample t test, it was
concluded that the students gained by at least 36% in the
knowledge of and interest in general engineering, chemical
engineering, and the viscosity topic. In the future, however,
a short test may be more informative than the student self-as-
sessment for determining how much the students learned during
the demonstration. Overall, the survey shows that the students
are more interested in general engineering, but their interest in
chemical engineering increased between 95% and 230%.
Survey Comments
Approximately half of the students indicated that the most
interesting part of the demonstration was the experiment.
The other half indicated that they enjoyed learning about
different types of engineering and/or learning about chemical
engineering. Most students didn't comment on potential im-
provements, but of those who did, the majority indicated that
more audience (i.e., student) participation was preferred.


SUMMARY AND OUTLOOK
An experiment/presentation appropriate for high school stu-
dents was developed and demonstrated. Based on the survey
results, the students gained by at least 36% in the knowledge
of and interest in engineering, chemical engineering, and
fluid viscosity. Furthermore, interest in chemical engineering
increased between 95% and 230%.
Based on the survey results and the facilitator's perception,
for any high school experimental demonstration, a significant
portion of the time allotted should be devoted to talking to the
students about engineering and chemical engineering. In the
future, the facilitators would like to contact high schools and
offer to send them simple tank-tube viscometer kits so that
a viscosity experiment can be incorporated into their exist-
ing curriculum. Also, the facilitators would like to develop a
program so that undergraduates can participate in the viscosity
experiment at local high schools as one of the department's
outreach efforts.[4]

ACKNOWLEDGMENTS
The authors would like to thank the Procter & Gamble Cor-
poration for funding this work under a curriculum grant.

REFERENCES
1. Rhinehart, R.R., ,... ,... i. ..; of Enrollment Cycling in ChE," Chem.
Eng. Ed., 35(1) 50 (2001)
2. The ACS Committee on Professional Training, "ACS Committee on
Professional Training Annual Report Tables: 1995-96 through 2003-
2004," (2004)
3. Halford, B., "Pursuing New Paths," ASEE Prism Online, Vol. 13, No.
3 (2003)
4. Ross, J.M., and T.M. Bayles, "Incorporating High School Outreach into
Chemical Engineering Courses," Chem. Eng. Ed., 37(3) 184 (2003)
5. Kwon, K.C., S. Pallerla, and R. Sanjeev, "Experiments on Viscosity of
Aqueous Glycerol Solutions Using a Tank-Tube Viscometer," Chem.
Eng. Ed., 33(3) 232 (1999) 1


TABLE 1
Student Survey Results


3J./ 2.0 1.0 4.2 Z.1

1.9 1.9 1.3 3.0 1.5

8.3 8.5 8.0 7.3 5.5

0.9 1.1 1.7 2.2 2.5

4.6 5.8 6.3 3.1 3.4

2.1 2.3 2.0 2.8 2.5

10.7 12.4 15.4 5.4 6.7
3.4-5.8 4.5-7.1 5.2-7.4 1.5-4.7 2.0-4.8
Chemical Engineering Education











I, laboratory













PLANT DESIGN PROJECT:

Biodiesel Production Using Acid-Catalyzed

Transesterification of Yellow Grease










RAFAEL HERNANDEZ, TRENT JEFFREYS*, ANIRUDHA MARWAHA, AND MATHEW THOMAS
Mississippi State University Mississippi State, MS 39762
ver the last 10 years, the chemical industry, federal Rafael Hernandez has a B.S. (1993) and M.S. (1995) in chemical
and state agencies, and the chemistry and chemi- engineering from the University of Puerto Rico, Mayaguez, and a Ph.D.
cal engineering profession have been increasing (2002) in chemical engineering from Mississippi State University (MSU),
cal engineering profession have een increasingly Mississippi State, MS. He worked for the U.S. Army Corp of Engineers'
investing intellectual, technical, and financial resources on Engineering Research and Development Center (1994-1997) on the
the research, development, and application of chemicals and development, design, and implementation of groundwater treatment
technologies. Presently, he works as an assistant professor in the Dave
fuels generated from renewable raw materials and sustain- C. Swalm School of ChemicalEngineering at MSU. His research interests
able processes. The main goal of the involved parties is to are the development of technologies for the remediation of contaminated
develop p s tt media and the use of biomass for producing value-added chemicals.
develop energy-efficient and cost-effective processes that
prevent pollution and decrease our dependency on foreign Trent Jeffreys has a B.S. (2004) in chemical engineering from Missis-
oil. The number of papers describing sustainable processes sippi State University. Currently, he works for Albemarle Corporation in
Orangeburg, SC as a process technology engineer. He has worked on
and renewable fuels that have appeared in the publications projects for cost reductions, automation and control system develop-
and conferences of the American Chemical Society (ACS) ment, and sample campaigns for new products in the Fine Chemicals
Division.
and American Institute of Chemical Engineers (AIChE)
have increased significantly over the last five years. The Matt Thomas received his B.S. in chemical engineering from Mississippi
State University in 2004. He is currently seeking his master's degree
2005 57th AIChE Institute Lecture was titled I I'l_.. Sup- in chemical engineering at Mississippi State University, studying the
ply Challenges and Opportunities." The ACS dedicated one remediation ofnitroaromatic contaminated groundwater. Afterreceiving
issue of Environmental Science and Technology, the society's his master's, he plans on continuing his education at Mississippi State
University by working toward a Ph.D. in chemical engineering and focus-
main publication on environmental research, to sustainable ing on producing oils from glycerol for biodiesel production.
processes.[ Presently, that journal includes a section on sus-
SAnirudha Marwaha has his B.S. degree in chemical engineering (2003)
tainable technologies in every issue. Additionally, numerous from Mississippi State University. He is currently pursuing a M.S. degree
papers were presented at the 2005 Annual AIChE Meeting in chemical engineering at Mississippi State. Since 2001, he has been
working for the E-Tech Lab in the chemical engineering department.
(Cincinnati) on biorefineries, sustainable technologies, and His thesis work is on the solidification and stabilization of high-level
renewable fuels, radioactive waste.

* Albermarle Corporation in Orangeburg, SC Copyright ChE Division of ASEE 2006


Summer 2006


215











The main contributing factor to the chemical engineering
and chemistry professions' focus on efforts to promote the
development of sustainable technologies and the production
of renewable alternative fuels such as ethanol, biodiesel, and
hydrogen has been the commitment of resources by the U.S.
Environmental Protection Agency (USEPA), the U.S. Depart-
ment of Agriculture (USDA), the National Science Founda-
tion (NSF), and the U.S. Department of Energy (USDOE).
One way to develop creative new production processes for
renewable chemicals is to educate future chemists and chemi-
cal engineers on the design, advantages, disadvantages, and
economics of current production techniques.
The last core course in the chemical engineering curriculum
at most universities in the United States is capstone design.
In this course, students have the opportunity to practice, for
the last time in an academic environment, the design and
economic evaluation of industrial chemical plants. To our


knowledge, most chemical engineering capstone design
projects focus on the use of petroleum-based raw materi-
als for producing specialty and commodity chemicals. To
broaden the students' perspective on the potential contribu-
tions of chemical engineering to areas such as new energy
sources, global warming, and environmental sustainability,
they should be introduced to the conversion of plants, natural
oils, microorganisms, and other types of biomass into alterna-
tive energy sources and value-added products. The capstone
course represents an excellent opportunity to assign projects in
which students synthesize and analyze renewable-chemicals
production facilities. The objective of this paper is to describe
a project entitled "Design of a Biodiesel Production Facility
Using Acid-Catalyzed Transesterification of Yellow Grease,"
assigned to the capstone design course at Mississippi State
University (MSU). Research on biodiesel is conducted by the
class instructor's research group. Thus, the design problem


Biodiesel is an alternative renewable fuel derived from vegetable oils or animal
fats, which conforms to ASTM D6751 specifications for use in diesel engines.[2]
Biodiesel utilization has increased significantly over the last 10 years, mainly
due to environmental benefits and government efforts to reduce dependence on for-
eign oil. The use of biodiesel reduces emissions of CO2, CO, SO2, and particu-
lates from operating diesel engines. Under a newly established sustainable energy
policy by the U.S. Department of the Interior, over 20 national parks operate
boats, trucks, heating systems, electricity generators, and other fuel related
systems on 100 percent biodiesel and/or biodiesel/petroleum diesel blends. Blends
of 20 percent biodiesel with petroleum diesel require no engine modifications.
Furthermore, biodiesel/petroleum diesel blends have demonstrated lubricity en-
hancements over the newly required low sulfur petroleum diesel. Numerous school
districts, transit authorities, public utility companies, and recycling companies
have also successfully used biodiesel. Recently, the U.S. military has begun to
procure biodiesel for use in on-base vehicles. These numerous experiences with
the use of biodiesel have clearly shown the environmental and high performance
characteristics of this alternative fuel.
In spite of the fact that Mississippi ranks 4th and 16th nationally in yellow
grease generation and soybean production (main biodiesel feedstocks), there are
no biodiesel production facilities in the state. The Swalm Engineering Design
Group at MSU was asked by the Alternative Energy Company to perform a prelimi-
nary design of a 2,240 lb/hr biodiesel production facility using acid-catalyzed
transesterification of yellow grease, and to evaluate the process economics. In
order to perform sensitivity analysis of process variables, the company requires
a simulation of the whole process using ChemCad. The company has acquired land
adjacent to a fertilizer manufacturing company in Yazoo City, Miss., at the cost
of $1,000,000 as the plant site. The design is to be based on a project life of
20 years. The major equipment, however, is to be depreciated in accordance with
applicable IRS regulations. In your final design report, you are requested to
provide estimations of the annual return on investment as well as the rate of
discounted cash flow taking into account the most recent laws and regulations on
corporate taxes. Design basis and specifications, available utilities, and other
information will be provided in further communications.


Figure 1. Project description.
216 Chemical Engineering Education











Some of the comments in the students'
course evaluations were..., "What I
like most about this course was the fact
that the project was broken into separate
portions over the whole semester," and
"I loved the layout of the class ... ."


also represented an excellent opportunity to integrate re-
search and education. As part of the course, invited speakers
and the course instructor presented seminars on ethics, job
interview preparation, entrepreneurship, and the social and
environmental implications of reducing our dependency on
petroleum. A workshop on ChemCad (chemical process
simulation software) was offered to students and faculty
by the software creators.

PROJECT DESCRIPTION
The statement of the problem submitted to the students on
the first day of class is presented in Figure 1. The class was
divided into five groups and each group had four members
selected by the instructor. The same project was assigned to
all groups. The open-ended nature of the problem statement
led to five different design configurations.
The design project was divided into three progress reports
(memorandums) and one final report. Several activities
and rules were established to maximize participation of
all students:
(A) Progress reports and the final report were accompa-
nied by an oral defense. The student in charge of present-
ing the oral defense was selected at the time of the pre-
sentation. Each member of a design group was questioned
extensively during each progress report presentation.
(B) Written peer evaluations were required after each
progress and final report. The evaluation forms were
similar to those suggested by I -.
C) A panel of industry and academic members judged and
selected the best final presentation. The presence of indus-
try representatives was additional encouragement for all
the students to prepare for the presentation. The instructor
selected the best report. The group or groups with the best
presentation and final report received plaques and cash
awards.

PROGRESS REPORTS
Division of the design project into progress reports had
two objectives. The first objective was to evaluate an induc-
tive approach to the teaching of plant design. This approach
consists of the presentation of a general problem or concept,
followed by closer focus on details and the solution of
component small problems. This method is applied by the


Summer 2006


chemical industry and during academic and industrial research
and development activities and it is an approach suggested by
chemical engineering educators.[4 5]
The second objective was to facilitate the organization of the
project and enhance students' time-management skills. Some of
the comments in the students' course evaluations were related
to the second objective. For example, "What I like most about
this course was the fact that the project was broken into separate
portions over the whole semester," and "I loved the layout of the
class-progress reports and the final presentation." The tasks
conducted for each progress report were as follows:
Progress Report 1: Literature survey, calculation of gross
profits, block diagram preparation, overall mass balance
calculations, and input of yellow grease components into the
ChemCad database.
Progress Report 2: Preparation of process-flow diagram and
simulation of the transesterification reactor and methanol
recovery system.
Progress Report 3: Simulation of all the biodiesel purifica-
tion steps: neutralization, solids removal, glycerol recovery,
and biodiesel and glycerol purification.

PROJECT SOLUTION
Yellow grease is the fat generated during animal rendering
activities. It is mainly composed of oleic, palmitic, and stearic
fatty acids attached to glycerol,[6] and contains a relatively high
percentage of free fatty acids (15%). Zhang (2000) used triolein
(triacylglycerol) as a test compound to represent yellow grease
during a Hysys simulation of a biodiesel production facility.']
The acid-catalyzed transesterification of this compound using
methanol as the alcohol results in methyl oleate and glycerol.
Zhang (2000) assumed that biodiesel could be represented by
methyl oleate.[7 To generate a mixture of transesterification
products with similar biodiesel chemical and physical proper-
ties, the students were encouraged to use several triacylglycerols
and oleic acid (free fatty acid) as representative of yellow grease
for the ChemCad simulation of the biodiesel production facility.
Figure 2 presents the reactions of acid-catalyzed transesterifica-
tion of the selected triacylglycerols and oleic acid. The acid-

C57H0406, + 3CH3OH H2SO4 )3C19H3, + C3H 03
Triolein Methyl Oleate Glycerol

C519HO + 3CH3OH H2SO4 3C1 H3402 + C3H03
Tripalmitin Methyl Palmitate Glycerol

C5,H342 + 3CH3OH H2SO4 3C19H32O, +C3HO03
Tristearin Methyl Stearate Glycerol

CH3402 +CH30H H2SO4 C19H3O + H2 0
Oleic acid Methyl Oleate
Figure 2. Acid-catalyzed transesterification reaction for
producing fatty acid methyl esters (FAME).
217





















































Figures 3a and 3b.
Process flowsheet
state.edu/hernandez/
biodieselproduction/fig-
ure 3.jpg>.


8 Chemical Engineering Education


EC-4M
eedist|
CMdae. -











catalyzed transesterification of the proposed components of
yellow grease results in a mixture of methyl esters of oleic,
palmitic, and stearic fatty acids. This mixture contains more
than 90% of the methyl esters found in commercial biodie-
sel from yellow grease. Phase behavior of triglyceride- and
alcohol-rich phases was ignored for simplicity. Students
recognized that they were making this simplification.
The complete process flow diagram (PFD) and stream
table are presented in Figure 3 (a and b) as well as in
Table 1. Both were prepared using the ChemCad process


simulation software licensed by Chemstation in Houston.
Some of the physical and chemical properties of the yellow
grease-assumed components were determined using the UNI-
FAC Group Contribution method in ChemCad. Other basic
properties, such as boiling point and melting point, were input
manually into the simulator.
The first main unit operation of the PFD is the transesterifi-
cation reaction system (R200). To determine reactor volumes,
it was assumed that the reactors were half full and the reac-
tions followed first-order kinetics. Reactor volume meeting


TABLE 1
Stream Properties Corresponding to the Process Flowsheet Presented in Figure 3


Stream No. 2 7 8 9 10 11 12 14

Name HSO4 MeOH YG RxOut MeOH Rcy Rx Feed MeOH Bot LLE Bot

Molar Flow, lbmol/h 3.93 7.85 3.35 178.07 162.57 178.07 15.5 32.32

Mass Flow, lb/h 385.12 251.52 2206.75 8200.12 5207.87 8200.09 2992.34 1057.53

Temperature, "C 25.12 25.16 25.19 80 64.41 59.02 140.35 139.57

Pressure, kPa 340 340 340 400 101 400 110 111

Vapor mole fraction 0 0 0 0 0.0032 0 0 0.7484

Enthalpy, MMBtu/h -1.3256 -0.80637 -2.7577 -21.14 -16.459 -21.501 -4.6005 -4.6009

Average mol. weight 98.08 32.04 659.29 46.05 32.04 46.05 193.03 32.72

Actual dens. lb/ft3 114.42 49.28 55.17 49.07 15.53 5031 53.42 0.09

Std liq. ft3/hr 3.38 5.03 39.63 154.84 104.21 155.02 50.63 13.52

Flow rates in lbmol/h

Triolein 0 0 1.17 0.04 0 1.2 0.04 0

Tripalmitin 0 0 0.67 0.02 0 0.69 0.02 0

Tristearin 0 0 0.34 0.01 0 0.35 0.01 0

Oleic Acid 0 0 1.17 0 0 1.17 0 0

Sulfuric Acid 3.93 0 0 3.93 0 3.93 3.93 3.9

Methanol 0 7.85 0 162.64 162.49 170.34 0.16 0.15

Methyl Oleate 0 0 0 4.67 0 0 4.67 0

Methyl Palmitate 0 0 0 2.06 0 0.04 2.06 0

Methyl Stearate 0 0 0 1.28 0 0.26 1.28 0

Glycerol 0 0 0 2.18 0 0 2.18 2.18

Water 0 0 0 1.25 0.08 0.08 1.17 26.09

Calcuim Oxide 0 0 0 0 0 0 0 0

Calcium Sulfate 0 0 0 0 0 0 0 0

Summer 2006


219












TABLE 1 CONTINUED
Stream Properties Corresponding to the Process Flowsheet Presented in Figure 3
Stream No. 16 17 18 19 20 21 22 23 25 27

Name LLE OH Biodiesel YG Rcy NeutRxOut Glyc Feed Waste H 0 Glycerol CaSO4 CaO Wash HO

Molar flow, Ibmol/h 8.44 8.07 0.38 36.21 29.76 27.71 2.05 6.45 3.9 25.26

Mass flow, lb/h 2389.81 2240.98 148.83 1276.07 686.56 501.17 185.39 589.52 218.54 455

Temperature, C 101.17 41.47 330 100 100.08 98.81 250 100.08 25 25.08

Pressure, kPa 101 8 20 110 340 101 110 340 101 340

Vapor mole fraction 0.001973 0 0 0 0 0 0 0 0 0

Enthalpy, MMBtu/h -2.6945 -2.6333 0.14603 -6.6235 -3.8985 -3.3341 -0.52741 -2.7249 2.1632 -3.1021

Average mol. weight 283.1 277.84 396.16 35.24 23.07 18.09 90.27 91.36 56.08 18.01

Actual dens. lb/ft3 42.82 52.63 42.82 86.96 63.16 59.66 68.13 154.95 155.46 62.22

Std. liq. ft3/hr 44.4 41.64 2.76 14.16 10.4 8.05 2.35 3.76 1.4 7.29

Flow rates in lbmol/h

Triolein 0.04 0 0.04 0 0 0 0 0 0 0

Tripalmitin 0.02 0 0.02 0 0 0 0 0 0 0

Tristearin 0.01 0 0.01 0 0 0 0 0 0 0

Oleic Acid 0 0 0 0 0 0 0 0 0 0

Sulfuric Acid 0.03 0.03 0 0 0 0 0 0 0 0

Methanol 0 0 0 0.15 0.14 0.14 0 0.01 0 0

Methyl Oleate 4.67 4.67 0 0 0 0 0 0 0 0

Methyl Palmitate 2.06 2.01 0.04 0 0 0 0 0 0 0

Methyl Stearate 1.28 1.02 0.26 0 0 0 0 0 0 0

Glycerol 0 0 0 2.18 2 0 2 0.17 0 0

Water 0.34 0.34 0 29.99 27.62 27.57 0.05 2.37 0 25.26

Calcium Oxide 0 0 0 0 0 0 0 0 3.9 0

Calcium Sulfate 0 0 0 3.9 0 0 0 3.9 0 0


the conversion requirement (97% the initial triglycerides)
was minimized by including two equal-size reactors in series.
The first and second reactors achieve an overall 83% and
97% conversion, respectively. The volume of each reactor
was 200 ft3 and the material of construction selected was
316 stainless steel. The reactions were performed at 80 C
and 400 kPa. The reactor influents were 3.35 lbmol/hr, 3.93
lbmol/hr, 170.42 lbmol/hr yellow grease, sulfuric acid, and
methanol. The reactors were simulated in ChemCad using the
equilibrium reactor. This reactor gives the user the capability
to simulate multiple reactions.

The purpose of the methanol recovery system (T210) is to
220


return excess, unreacted methanol to the reactor to save raw
material costs. The major challenge in simulating the metha-
nol recovery tower is to return as much methanol as possible
to the reactor, thus minimizing water in the recycle stream.
Conversion of triglycerides into biodiesel drops dramatically
if the reactants contain between 0.5% and 5% water. The
simulation was conducted using the ChemCad tower module.
The column was operated using atmospheric pressure for
the overhead stream and a bottom pump-out pressure of 110
kPa. The smallest number of theoretical stages that could
be obtained while keeping the water-weight percent of the
reactor feed below 0.10% was 13 with a feed stage of seven.
This resulted in a water concentration 0.064% by weight in
Chemical Engineering Education











the feed to the reactor. The distillation column recycles 99.9%
by weight of the methanol in the reactor effluent.

The remaining portion of the process after the methanol re-
covery system consists of separation and purification steps to
obtain purified biodiesel, glycerol, and yellow grease recycle
streams. The effluent of the surge tank (V-220) is pumped
into the bottom stage of the liquid-liquid extractor (T-300).
Water cascades down the column after being fed into the top
stage. The wash water extracts the entrained glycerol while
the biodiesel and unreacted yellow grease exit the top of the
column. The ChemCad simulation of T-300 resulted in four
theoretical stages for complete separation of biodiesel from
glycerol. The students conducted a sensitivity analysis using
ChemCad to determine the effect of reboiler operating tem-
perature of the biodiesel distillation column and wash-water
flow into the liquid-liquid extractor on cost and biodiesel
purity, respectively. Figure 4 shows the effect of wash-water
flow on biodiesel purity. It can be observed that water flows
in excess of 500 lb/hr have a negligible effect on biodiesel
purity. This type of analysis is essential to determine optimum
plant operating conditions to meet biodiesel quality.
The crude glycerol stream from the bottom of the extrac-
tor flows to the reactor (R-500) for the neutralization of the
sulfuric acid catalyst by the following reaction:
H2SO4 + CaO CaSO4 (gypsum) + H20

Calcium oxide (CaO) was the base choice due to low cost,


limited complications in regard to materials of construction,
and low solubility of its salts formed by neutralization. A
CSTR was selected to perform the transesterification reaction.
The reactor was simulated in ChemCad using the stoichiomet-
ric reactor module and assuming 100% conversion of sulfuric
acid. The reactor was maintained at 80 OC. Effective mixing
of solids is easily maintained by physical agitation in a CSTR.
Additionally, the CSTR should prevent any excess collection
of calcium sulfate in the neutralization reactor. A centrifuge
(CN-510) was used to separate the gypsum from the glycerol
and water. A solids effluent moisture fraction of 10% was
defined for the centrifuge. It was assumed that the gypsum
recovered was sold to a cement company at $56/ton.1
The liquid stream from the centrifuge flowed into the glyc-
erol purification tower (T-600) to achieve a bottoms product of
99.5% by weight purity glycerol. Four theoretical stages were
required to achieve the desired purity. To use high-pressure
steam, the column must be operated at a reduced pressure so
that the reboiler temperature is 250 C. The final reflux ratio
of 1.8 results in a reasonable reboiler duty while maintaining
the desired purity of 99.5%.
The biodiesel purification column (T-400) must produce
99.6% by weight biodiesel by separating the methyl esters
from the unreacted yellow grease. This column presented
several challenges in simulating its operation due to the lack
of experimental vapor-liquid equilibrium data for biodiesel
1 Communication with cement company.


1.000



0.980

o0
U
0.960

-)

0.940

[-
a
.a 0.920

m

0.900



0.880
0 100 200 300 400 500 600 700 800 900 1000
Wash Water (Ib/hr)

Figure 4. Effect of wash-water flow into the liquid-liquid extractor
on the biodiesel purity exiting the distillation tower.

Summer 2006


221











and yellow grease. Operation of the column at atmospheric
pressure required extremely high reboiler temperatures of
up to 600 C to achieve a sufficient biodiesel purity. There-
fore, the column was simulated under severe vacuum with
top and bottom operating pressures of 8 kPa and 20 kPa,
respectively. The necessity of low vacuum for the separation
of triglycerides from biodiesel also has been observed by
other investigators.6, 7] Under these conditions, the bottom
product temperature was 330 C. The vacuum necessary for
this separation can be achieved using multistage steam injec-
tors.81 Dowtherm G at 357 C was selected as the heating
medium for the reboiler.
Students prepared the final report following the format
suggested by Peters, et al.[9] Capital and operating costs
were determined using the Cap Cost software included in the
textbook by Turton, et al.,E101 Web sites, and communications
with vendors. Several scenarios were evaluated to determine
plant economics. For example, students evaluated return on
investment (ROI), taking into consideration the biodiesel tax
incentive ($1.00/gal) included in the current version of the
Energy Bill. ROI also was determined after increasing the op-
erating capacity of the plant. These scenarios helped students
understand the economics of scale and the current situation
of the biodiesel industry in the United States, which requires
government incentives to be economically feasible.11u
The results of the estimation of capital and total product
costs are presented in Tables 2 and 3. The information in these
tables was essential to determine net present value and the
ROI. The prices used for raw materials costs were: yellow
grease, $0.1175/lb; methanol, $0.6/gal; sulfuric acid, $67/ton;

TABLE 2
Estimation of Capital Investment
for the Proposed Biodiesel Production Facility
Estimation of Capital Investment
Cost components Direct Costs
Equipment (including service, installation, and instru-
mentation)
Distillation columns $206,100
Jacketed reactors $252,450
Liquid-liquid extractor $42,000
Heat exchangers $448,900
Pumps $93,892
Centrifuge $88,349
Tanks $153,263
Total equipment costs $1,291,642
Land (buildings and service facilities included) $1,000,000
Indirect costs (20% fixed-capital investment19]) $572,910
Fixed-capital investment $2,864,552
Working capital (15% of fixed-capital investment) $429,683
Total capital investment $3,294,235


and calcium sulfate, $56/ton. Except for the price of calcium
sulfate, all the other prices were obtained from the September
2003 issue of the Chemical Market Reporter.[12] The students
assumed that the solid recovered during the neutralization step
(calcium sulfate) was sold to a local cement company. They
contacted a local cement company to obtain a calcium sulfate
purchasing price. Utilities costs were the following: low-pres-
sure steam, $2.50/1000 lb; high-pressure steam, $5.50/1000
lb; natural gas, $2.70/1000 SCF; electricity, $0.04/kWh; cool-
ing water, $0.05/1000 gal; wastewater treatment, $56/1000
m3; and process water, $0.5/1000 gal.10 13]
Total income was calculated by adding biodiesel, glycerol,
and calcium sulfate sales. The prices used for biodiesel,
glycerol, and calcium sulfate were $2.411 gal (the price of
petroleum diesel at the time was $1.4'1- gal and the $1.00/gal
tax incentive was added), $0.72/lb,[12] and $15/ton,[14] respec-
tively. The mass and volume rates are presented in Table 1.
The total annual income was $7,083,700. Subtracting the
total product costs shown in Table 3 results in annual gross
earnings of $888,100. Assuming a35% tax, the after-tax profit
(ATP) was $577,200.
The after-tax cash flow (ATCF) is the sum of the ATP and
depreciation. ATCF is calculated for every year of plant op-
eration. The depreciation was calculated using straight-line
depreciation with 9.5 years recovery period. Thus, deprecia-
tion and ATCF were given by:

d original investment
9.5 years
d$2,291,642 $241,226
9.5 years year
ATCF = ATP + d
ATCF = $577,277 + 241,226 = $818,503

Only half of the depreciation was added in year 10 of
operation and no depreciation was added in the final years
of operation.110 ROI is a profitability measure defined as
the ratio of profit to investment. Average profit over the
20 years of plant operation and fixed capital investment
were used to calculate ROI for the biodiesel production
facility. This value resulted in:
20
SATP
ROI= 1 x100
20
$577, 277
ROI = x 100 = 20%
$2,864,552

This value of ROI is considered acceptable for a new
product entering into an established market.[9] The pay-
back period is the length of time necessary for the total


1 Communication with cement company.


Chemical Engineering Education











return to equal the capital investment. It was calculated using the
following equation:
FCI
PBP 20
ATCF

20
$2,864,552
PBP = $2,4 = 4.14 years
$13,837,187 / 20

To calculate ATCF, full depreciation was only added the first nine
years of operation and only half of the depreciation was added for
year 10. As mentioned above, no depreciation was added the final
years of operation. The value of PBP obtained is also acceptable
for a new product entering an established market.[91 The students
concluded that a biodiesel production facility is not economically
feasible without government tax incentives. This result gave the
students an understanding of the need for state and federal support
for developing new industries associated with renewable energy.

Some of the results presented above were taken from the class-best
final report. Student course evaluations and senior exit interviews
indicated that the application of research and teaching was an excit-
ing and motivating experience for the class. Some of the students'

TABLE 3
Estimation of Total Product Cost for the Proposed
Biodiesel Production Facility
Estimation of Total Product Cost
Cost Components Cost/Year
Manufacturing Cost
Raw Materials $2,407,199
Utilities $367,178
Labor (based on plant capacity 1. .1I '. $1,252,912
Maintenance (7% of fixed capital investment $90,415
minus land and indirect costs[9])
Operating (15% of maintenance costs) $13,562
Depreciation (straight line depreciation) $241,226
Local Taxes (1% of fixed capital investment"9) $28,645
Insurance (1% of fixed capital investment"9) $28,645
Overhead (56% of labor and maintenance"9) $749,202
Total Manufacturing Cost $5,178,984
General Expenses
Administrative (20% of operating labor and $273,186
maintenance'91)
Distribution and Marketing (7% of the total $433,695
product costs[91)
Research and Development (5% of the total $309,782
product costs[91)
Total General Expenses $1,016,663
Total Product Cost (Total Manufacturing + $6,195,647
General Expenses)


Summer 2006


comments about the project included:
"I liked the fact that the project was a real-life
application."
"I became more competent with ChemCad."

"This class helped with my teamwork skills."

Additionally the class benefited by:
Access to the instructor's extensive literature
collection on biodiesel production technology.

Excitement of working on the production of
renewable fuel with clear environmental, health,
and safety benefits.

Discussing contemporary issues associated with
the economic feasibility of a renewable fuel.

Visualizing the importance of lifelong learning
on the application of chemical engineering prin-
ciples to contribute solutions to society's dwin-
dling energy resources.

Determining the capital and P *."", cost driv-
ers of the acid-catalyzed transesterification
biodiesel production process.

CONCLUSIONS

The design project offered students the opportunity
to apply chemical engineering to the transformation
of a nontraditional raw material into a fuel. The
students gained a new perspective on the potential
contributions of chemical engineering to areas such
as new energy sources, sustainability, and policy. The
approach of presenting a general problem or concept,
followed by a closer focus on details and the solution
of component small problems using chemical process
simulation, was key to the successful completion of
the design problem.

ACKNOWLEDGMENTS

The course instructor is grateful for the excellent
presentations on ethics, ChemCad, and career options
provided by Robert Green, Stuart Schwab, and the
Mississippi State University Career Services. Funds
for the best report and presentation awards were pro-
vided by Tom Austin, a Mississippi State chemical
engineering alumnus.

REFERENCES
1. Environmental Science and Technology, 27(23) (2003)
2. USEPA Office of Air and Radiation, A Comprehensive Analy-
sis of Biodiesel Impacts on Emission Exhaust, EPA Report
Number: 420-P-02-001 October (2002)
3. Fogler, H.S., workrules.htm.>
4. Wankat, PC., The Effective, Efficient Professor: Teaching,
223












Scholarship, and Service, Allyn and Bacon, Boston (2002)
5. Dahm, K.D., R.P Hesketh, and M.J. Savelski, ( I.... I,, i.' 36(3),
192, (2002)
6. Canakci, M., and J. Van Gerpen, Trans. ASAE, 44, 1429 (2001)
7. Zhang, Y., M.A. Dube, D.D. McLean, and M. Kates, Bioresource
Technology, 89(1) (2003)
8. Perry, R.H., D.W. Green, and J.O. Maloney (Eds.), Perry's Chemi-
cal Engineering Handbook, 6th Ed., McGraw-Hill, Inc., New York
(1984)
9. Peters, M.S., K.D. Timmerhaus, and R.E. West, Plant Design and
Economicsfor Chemical Engineers, 5th Ed., McGraw-Hill, Inc., New
York (2003)


10. Turton, R., R.C. Bailie, WB. Whitting, and J.A. Shaeiwitz, Analysis,
Synthesis, and Design of Chemical Processes, 2nd Ed., Prentice Hall,
NJ (2003)
11. Tyson, K.S., J. Bozell, R. Wallace, E. Petersen, andL. Moens, Biomass
Oil Analysis: Research Needs and Recommendations, NREL Report
No. TP-510-34796 (2004)
12. Schnell Publishing Company, Chem. Marketing Reporter, 264(6)
(2003)
13. Seider, WD., J.D. Seader, and D.R. Lewin, Productand Process Design
Principles: Synthesis, Analysis, and Evaluation, 2nd Ed., John Wiley
& Sons, Inc., New York (2003) 1


Chemical Engineering Education











M, t1laboratory


EXPERIMENTAL INVESTIGATION AND

PROCESS DESIGN

in a Senior Laboratory Experiment





KENNETH R. MUSKE
Villanova University Villanova, PA 19085


Drying experiment for the senior unit operations labo-
ratory course at Villanova University is described in
this article. This experiment involves the determina-
tion of the drying rate of a solid material in a forced convection
drying apparatus and the scale-up design of this process. The
experimental drying rate data is used to determine the appro-
priate transport coefficients to mathematically describe the
drying process. The students then use this mathematical model
for the design of a large-scale dryer for a specified production
rate of the material under study. Various solid materials such
as sand, gravel, clay, sawdust, natural and synthetic fibers,
and agricultural products have been used in this experiment.
This variety of materials is intended to provide each student
group with a different experience that can be compared and
contrasted during student group oral presentations at the end
of the semester.
The main goal of the laboratory exercise documented in this
article is to provide the students with hands-on experience
in the analysis and design of drying processes. Drying is an
essential unit operation in the chemical process industries
with applications ranging from forest productsl and mineral
processing21] to food products31] and pharmaceuticals.[41 Al-
though this technology has been a key component of chemical
engineering since its inception as an academic discipline,
the science of drying continues to remain an active area of
research and development 1 Despite its widespread industrial
importance, however, it is not emphasized in the heat and
mass transfer courses due to time constraints. This laboratory
experiment provides an opportunity for students to apply
transport phenomena concepts presented in the classroom
to the process of drying, while becoming familiar with this
common unit operation.


A second goal of this experience is to provide students with
an opportunity to apply the results of their experimental study
to a process design. A similar approach to the unit operations
laboratory course is advocated in Reference 6. The emphasis
of this experiment is not simply to obtain data to determine
transport coefficients. The students must also use their results
in the scale-up design of the drying process. This addition of
a design element to the laboratory provides a more practical
objective for students and a more realistic application of their
experimental investigation. This experience also provides
additional learning objectives in the laboratory course, such
as the development of engineering awareness, mathematical
modeling, scale-up, and economic evaluation.[7
There have been a number of chemical engineering labo-
ratory drying experiments reported in the literature such as
microwave drying of sandE81 and convection drying of a
towel.[9] A bench-scale experimental drying apparatus10] and
the statistical treatment of drying data111" have also been re-
ported. The unique aspect to the experiment described in this
article is both the incorporation of a design element and the
study of a wide variety of materials with drastically different
drying properties for each group.


Kenneth Muske is an associate professor of chemical en-
gineering at Villanova University where he has taught since
1997. He received his B.S. ChE and M. S. from Northwestern
(1980) and his Ph.D. from The University of Texas (1990), all
in chemical engineering. Prior to teaching at Villanova, he was
a technical staff member at Los Alamos National Laboratory
and worked as a process control consultant for Setpoint,
Inc. His research and teaching interests are in the areas of
process modeling, control, and optimization.


Copyright ChE Division of ASEE 2006


Summer 2006


225










LABORATORY EQUIPMENT
The experiment is carried out using a batch cross-circula-
tion cabinet dryer. A schematic of the dryer system is shown
in Figure 1. Air is supplied at a rate of 0-440 ft3min by a
centrifugal blower with a gate-valve arrangement to adjust the
air flow into the cabinet. The air is heated by two steam coil
heaters located in the bottom half of the cabinet. The inlet air
is passed through the heaters in the bottom half of the cabinet
before being redirected to the drying section in the top half
of the cabinet. A baffle arrangement (not shown in Figure 1)
provides uniform air flow in the drying section. The volume
of the drying section is 5.6 ft3 and the cross-sectional area for
the air flow is 1.9 ft2. The material to be dried is contained
within a shallow tray or wire basket that is suspended in the
air stream.
Dryer air temperature is controlled by a valve on the inlet
steam line to the heaters with manual valves to isolate either
heater. Steam is supplied from the high-pressure building
main. Source pressure to the dryer is maintained at 30 psig
by a regulator on the supply line. This reduction in the heater
steam supply pressure is for both safety considerations and
improved temperature control by increasing the normal
operating range of the steam valve. The maximum dryer
temperature is restricted to 200 F for all experiments to
prevent the possibility of thermal decomposition or ignition
of the solid material.
Dryer process measurements include: the mass of the tray
and material using a load cell (Transducer Techniques EBB-5
load cell [0-5 kg] and DPM-3 digital panel meter with analog
output); the air temperature and relative humidity in the dryer
using a humidity probe (Omega Engineering HX94C relative
humidity/temperature probe); the surface temperature of the
material in the tray and a second dryer air-temperature mea-
surement using thermocouples (Analog Devices 2B52A type T
thermocouple transmitters); and the inlet air flow rate inferred
from the differential pressure across an orifice in the inlet
air header to the blower (Setra Systems C239D differential

I I r r


Vent f


Air
Inlet --


- -- Steam
Steam Valve


Condensate


Figure 1. Experimental cabinet dryer system.


pressure transmitter). The approximate cost of this instrumen-
tation at the time of installation was $400. In addition to the
electronic measurements, there is a thermometer inside the
dryer cabinet, a wet/dry bulb thermometer to determine the
conditions of the inlet air, and a water manometer connected
to the orifice in the inlet air line. The data acquisition and
control computer system displays the process measurements
in real time and also records these values in a data file for
later analysis by the students. A data sampling period in the
range of 0.25 to 1 min is suggested to the student groups for
data collection by the computer system.

LABORATORY EXERCISE
There are two three-hour laboratory sessions each week
for the experiments in the senior laboratory course. These
sessions are composed of two planning, experimental, and
analysis sequences. The students only have access to the
drying apparatus during the two experimental sessions. In the
first planning session, the group is introduced to the material
that they will be studying, the range of moisture content that
they must consider, and the specifications on the final dried
product. Because the students have essentially no practical
experience with drying processes, they are expected to re-
search the drying process and plan their experimental study
during this session. Presentations on drying can be found in
Perry's chemical engineering handbook"12 and other process
engineering handbooks such as Cooper, et al.,'131 along with
unit operation and mass transfer texts such as McCabe, Smith
and Harriot,[141 Geankoplis,11' and Treybal.J161
The initial drying experiments are carried out during the
first experimental session where the students determine the
character of the drying curve for their material. An initial estimate
for the tray loading is based on rules of thumb for the design of
drying processes such as presented in Reference 13. The drying
rate and moisture range for the constant rate period and the
transition to the falling rate period are determined from the
students' initial experimental results during the first analysis
session. An initial determination of the corresponding mass
transfer coefficients is also carried out dur-
1. ing this session.


In the second planning session, the group
develops its experimental plan for the sec-
ond experimental session. Depending on
the drying behavior of their material and
the results of their initial experimental ses-
sion, during this session the student groups
usually either concentrate on the falling rate
period or consider the effect of different tray
loadings. The incorporation of a second
experimental session allows for the unstruc-
tured experimental approach adopted in this
laboratory course. Benefits to the students of
an unstructured approach include exposure
to a more realistic experimental study, since
Chemical Engineering Education


Tray Load Air I temperature










in reality the precise details of a procedure are rarely known
in advance and the experimental plan often evolves as more
information is discovered.
The final analysis session is used to analyze the data ob-
tained from the second experimental session, and to design the
scale-up drying process. Each student group is informed of the
required production rate and initial moisture content of their
source material after completion of the second experimental
session. This practice is implemented in order to prevent the
students from specifically targeting their experimental study
to their scale-up process design requirements.

EXPERIMENTAL ANALYSIS
The process of drying can be described using an energy
balance determined by the heat transfer rate between the hot
gas and the moist solid, and a material balance determined
by the mass transfer rate between the moist solid and the hot
gas. The corresponding liquid component evaporation rate
can be calculated using the relationships 141

T,-T, (y -y,)
mh= M h\ -'-= \k 1 Y Yg) (1)
AH, (1- Y)LM

where ri is the evaporation rate of the liquid component
from the moist solid, M is the molecular weight of the liquid
component, A is the contact area between the moist solid and
hot gas, h is the heat transfer coefficient, T is the bulk hot
gas temperature, T is the solid-gas interface temperature, AH
is the latent heat of vaporization of the liquid component at
the interface temperature, k is the mass transfer coefficient
expressed on a mole fraction basis, yg is the bulk liquid
component mole fraction in the hot gas, y is the liquid com-
ponent gas-phase mole fraction at the solid-gas interface,
and (1 )LM is the log mean value of (1 y) and (1 y ).
For reasons of convenience and safety, the liquid compo-
nent is water and the gas is air for all of the experimental
studies in this laboratory. Due to the time constraints in
the laboratory sessions and the relatively slow evaporation
rates for the materials under study, the dryer system is typi-
cally operated at near maximum air flow rate to maximize
the evaporation rate. Under these conditions, constant air
temperature and humidity can be safely assumed. Because the
thermal changes to the system occur at a faster time scale
than saturation changes, it is also appropriate to assume
that the drying rate can be expressed in terms of the mass
transfer relationship for drying process design calcula-
tions.[17] This assumption is applied in both the experimental
data analysis and scale-up design. It should be noted that a
similar mass transfer expression to Eq. (1) in terms of partial
pressure and humidity driving forces can also be employed.
Although partial pressure has been the most popular, the
student groups have tended to be rather evenly divided
in their choice of driving force for their mass transfer
coefficient.
Summer 2006


The determination of the mass transfer coefficient depends
on the behavior of the drying curve for the material under
study. All of the materials considered in this experiment
exhibit a constant-rate drying period and some exhibit a
falling-rate drying period for the moisture range of interest.
The constant-rate period is characterized by a constant rate
of drying that is independent of the moisture content. Dur-
ing this period, a continuous film of water exists on the solid
surface that is constantly replenished as the surface water
evaporates. The falling-rate drying period occurs when the
moisture content of the solid falls below some critical point.
After this point, there is insufficient moisture present to
maintain a continuous liquid film on the solid surface and the
liquid mass transfer in the solid phase becomes limiting as
opposed to interfacial mass transfer during the constant-rate
period. This critical point is a function of the material, the
material thickness, and the driving force for mass transfer that
usually must be determined experimentally."141 The drying
rate typically decreases as the moisture content in the solid
decreases during this period.
Under typical experimental conditions during the con-
stant-rate drying period, all of the parameters in Eq. (1) are
relatively constant. The constant-drying-rate mass transfer
coefficient can then be determined from the slope of the total
tray mass vs. time, total material mass vs. time, or the free
moisture mass vs. time drying curves as follows

k = -a( y)LM (2)
S V\(y yg)

where a is the slope of the drying curve during the constant-
rate drying period, yg is the mole fraction of water in the
inlet air determined from the humidity probe in the cabinet
and checked with the wet/dry bulb thermometer, and y is the
mole fraction of water at the solid-gas interface. The interface
mole fraction is assumed to be the equilibrium saturation
value at the interface temperature, which can be determined
using steam tables or the Antoine equation. As discussed in
Reference 15, it is possible to apply the dilute gas-phase mole
fraction approximation (1 y)LM ~ 1 in the analysis of the
constant-rate-period mass transfer coefficient.
The determination of the mass transfer coefficient for the
falling-rate period is more problematic due to the changing
conditions of the material and the interface. Although many
approaches exist to describe the falling-rate period,"14 15 one
of the simplest is to assume that the drying rate is proportional
to the difference between the free moisture in the solid and
the equilibrium free moisture,

m= M \k (X-X*), X = (X-X') (3)
ms
where kx is the solid-phase mass transfer coefficient, X
is the bulk solid free moisture, X* is the equilibrium free
moisture, ms is the dry mass of the solid material, and the
227











other parameters are as defined previously. Eq. (3) can be
integrated yielding
m
t= In(X- X*)+b= aln(X- X*)+b (4)
JM \k

where a is the multiplicative constant and b is the constant of
integration. The solid-phase mass transfer coefficient can be


600


550


500


450


400


350


300


0 5 10 15


20 25 30 35 40 45 50 55 60 65 70 75
Time (minutes)


determined from the constant, a, obtained by a logarithmic fit
of the free moisture vs. time experimental data.

k = m (5)
aM \


This fit is easily a
fitting numerical pa


80 85 90 95 100


Figure 2. Total material mass drying curve.


149 -


148

147

146

145

144


141

140


Material Interface Temperature --oe--- ?-

Q
6 -

o

Q








<-s


0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100
Time (minutes)

Figure 3. Material interface temperature.


accomplished using Excel or any curve-
ckage. The use of the free moisture vs.
time curve, as opposed to the
drying rate computed by central
differencing the data as discussed
in Reference 8, provides a more ac-
curate estimate of the mass transfer
coefficients for both the constant
and falling rate periods because of
the noise inherent in the load cell
measurements.


. Measured Mass o
Linear Regression -


EXAMPLE EXPERIMEN-
TAL RESULTS
Figure 2 presents the total mate-
rial mass (load cell reading minus
the empty-tray mass) drying curve
for a coarse sawdust material ini-
tially composed of a 2:1 volumetric
mixture of sawdust (900 ml-156 g)
and water (450 ml-450 g) with an
average material thickness in the
tray of 1 cm. Figure 3 presents
the material interface temperature
for this system determined by a
thermocouple embedded in the
surface of the sawdust. Linear
regression on the material mass
for the constant-rate drying time
period between 5 and 80 minutes
resulted in a slope of -3.89 g/min
with a correlation coefficient of
0.992. Assuming an equilibrium
interface temperature of 140 F
for the constant drying period,
the resulting constant-rate drying
period mass transfer coefficient is
k = 11.2 gmol/m2-min. Although
not used in this calculation, the log
mean value was (1 y)LM= 0.895,
which is close to the dilute-gas
phase mole fraction approxima-
tion. The specification on the final
dried sawdust product was that
it must be free flowing without
any lumps. The student group
concluded that the dried sawdust
material met this criterion after

Chemical Engineering Education


S i i i 0


I I I I I I I I I I I I










85 minutes at the end of the constant-rate drying period with a
moisture content of 0.8 g H 0/g dry solid.
Figure 4 presents the free- moisture drying curve for a clay
absorbent material initially composed of a 3:1 mixture by mass
of absorbent (474 g) and water (158 g). The average material
thickness in the tray was 0.375 in. A logarithmic regression
on the falling-rate data after 31 min resulted in a value of the
multiplicative constant of -61.6 min with a correlation coef-
ficient of 0.953. The corresponding solid-phase mass transfer
coefficient is kx = 0.427 gmiii Ii -min. The specification on
the dried absorbent product was that it must be dry enough to
package. The student group concluded that the dried absorbent
met this criterion after 60 minutes with a moisture content of
0.075 g H 0/g dry solid. The justification of this decision was
that the drying rate essentially goes to zero after this time result-
ing in little further drying being possible without a very long
additional exposure.

SCALE-UP PROCESS DESIGN
The scale-up process design is based on manufacturing a speci-
fied production rate of some product from the wet solid material
with a specified initial moisture content. The final moisture
content of the material must be determined by the student group
based on a subjective performance criterion as indicated in the
initial laboratory handout. Examples of this criterion are that the
solid must be dried to the point that it is free flowing or dry to
the touch. This subjective criterion requires both experimental
data and engineering judgment to determine the final moisture
content. The intent is to demonstrate that product specifications
are often not directly measured physical quantities.
Because the mass transfer coef-
ficient is a function of the operating
conditions of the dryer, the scale-up 0.325 -4
process operation must not deviate 0.3
significantly from the experimental 0.
operating conditions if the experi-
mental mass transfer coefficient 0.25
is to be used in the design. This 0.225
restriction does impose limitations 0
0.2
on the scale-up design depending C
on the experimental conditions 0.175
that were considered. Specifically, 0.15
the material depth in the tray and 0.
the air velocity in the scale-up de- .125
sign must be representative of the 0.1
experimental conditions used to 0.075
determine the mass transfer coeffi-
cient. For example, the air velocity 0.05
is determined by the air flow rate 0.025
and cross-sectional flow area. To
0
change the driving force for mass '
0 5 10
transfer by changing the air flow
rate, the cross-sectional area must

Summer 2006


also be changed to maintain a similar air velocity. A benefit
of the design aspect in this experiment is the exposure to
this relationship between experimental investigation and
scale-up design through hands-on experience. Such expo-
sure is not available in the process design course because
of the lack of an experimental component.
A further objective of the design aspect is the develop-
ment of a physically realistic process design. Although
most student groups have little difficulty in determining
the required surface area and air rate for the scaled-up
process, the actual physical design of the dryer can often
be unrealistic. A common initial approach is to scale up the
experimental apparatus to handle the required production
rate. The result is a design with trays that are often too
large and heavy to be physically managed. More realistic
process designs evolve as the student groups are prompted
to consider the size and weight of the material that must
be handled. An example initial approach for the scale-up
design of a wood-chip dryer with a production rate of 1
ton/day consisted of a batch cabinet dryer using a single
50-ft-long tray containing almost one-half ton of wet wood
chips. The final design was also a batch cabinet dryer, but
instead consisted of ten 5-ft-long trays stacked on top of
each other where each tray initially contained approxi-
mately 120 lbs of wet wood chips. The consideration of
the practical aspects of a process design is an additional
benefit of this experience.
PRESENTATION OF RESULTS
The experimental and scale-up process design results
for each student group are reported in a formal written


15 20 25 30 35 40 45 50 55 60 65
Time (min)
figure 4. Free moisture drying curve.


229











report and an oral presentation. The formal written report
is due 10 days after the last analysis session for the experi-
ment. The oral presentations are scheduled over a series of
presentation days at the end of the semester. There are also
short memo reports required after each laboratory session to
document the planning, experimental, and/or analysis results
and conclusions. The different materials and scale-up require-
ments incorporated into this experiment provide each group a
slightly different experience that can be shared with the class
during the presentation sessions.
Each student group is composed of three students and there
are three experiments in the senior laboratory course. Therefore,
each student in the group takes on the responsibility of group
leader for one experiment in the sequence. Each group leader
is responsible for the formal written report, oral presentation,
and short memo reports on their experiment. The group leader
responsibility also includes coordinating the activities of the
other group members. When the class is not evenly divisible
by three, there will either be one two-member or four-member
student group. A two-member student group will not prepare
a formal written report and oral presentation for one of the
experiments although they will carry out this experiment and
prepare the short memo reports. A four-member group will be
given two objectives with a separate group leader for one of
the experiments. Because of the length of the drying experi-
ments, they have not been considered for two objectives in a
four-member group.

STUDENT RESPONSE
There are no formal course evaluations for laboratory courses
in the chemical engineering department at Villanova University.
Student response data for the senior laboratory course is ob-
tained from departmental surveys administered at the end of the
semester. Qualitative assessment of the students' experiences is
also based on their comments during and after the experiment.
This assessment indicates that the experience has been generally
well received by the students. Student comments concerning
the drying experiment reveal that the group spent more time on
this experiment because of the design aspect, which required
more use of the second planning and analysis sessions.

CONCLUSIONS

The laboratory experiment documented in this article has
been developed and implemented over the past two years in
the chemical engineering senior laboratory course at Villanova
University. Based on the results of informal course surveys,
the students have found the experience both challenging and
worthwhile, in addition to providing an applied mass transfer
and process design experience. The experiment has also pro-
vided valuable documentation of students' ability to design,
conduct, analyze, and interpret experiments for ABET18s
Criterion 3b and their ability to perform as part of a team for
ABET18' Criterion 3d.


ACKNOWLEDGMENTS

The collaboration with Prof. John Myers and Dr. Miriam
Wattenbarger in the development of the experimental exercise,
the contribution of Elena DeCandia of the class of 2002 in the
design and construction of the data-acquisition and control
system as part of her undergraduate thesis project, and a cur-
riculum revision grant to the Villanova University Department
of Chemical Engineering from Air Products and Chemical Co.
supporting the development of the senior laboratory course are
all gratefully acknowledged. I would also like to thank Gloria
Benson, Brett Gerasim, and Sebastian Houle of the class of
2005 for supplying their experimental results.

REFERENCES
1. Denig, J., E. Wenger, and W. Simpson. "Drying Hardwood Lumber,"
Technical Report FPL- GTR-118, USDA Forest Products Laboratory
(2000)
2. U.S. Department of Energy, Mineral Processing Technology Roadmap,
Office of Industrial Technologies (2000)
3. Clark, J., "Drying Still BeingActively Researched," Food Technology
56(9), 97-99 (2002)
4. Guerrero, M., C. Albert, A. Palmer, and A. Guglietta, "Drying in the
Pharmaceutical and Biotechnological Industries," Food Science and
Technology International, 9(3), 237-243 (2003)
5. Mujumdar, A., "Research and Development in Drying: Recent Trends
and Future Prospects," Drying Technology, 22(1/2),1-26 (2004)
6. McCallum, C., and L. Estevez, "Introducing Process-Design Elements
in the Unit Operations Lab," Chem. Eng. Ed., 33(1), 66-70 (1999)
7. Abu-Khalaf, A., "Getting the Most Out of a Laboratory Course," Chem.
Eng. Ed., 32(3), 66-70 (1998)
8. Steidle, C., and K. Myers, "Demonstrating Simultaneous Heat and
Mass Transfer with Microwave Drying," ( I,... i .t. 33(1), 46-49
(1999)
9. Nollert, M., "An Easy Heat and Mass Transfer Experiment for Transport
Phenomena," Chem. Eng. Ed., 36(1), 56-59 (2002)
10. Moor, A., "A Benchscale Drying Laboratory Illustrating Combined
Heat and Mass Transfer," ASEE Mid-Atlantic Section Conference,
Rowan University, April (2001)
11. Prudich, M., D. Ridgway, and V. Young, "Integration of Statistics
throughout the Undergraduate Curriculum: Use of the Senior Chemical
Engineering Unit Operations Laboratory as an End-of-Program Statis-
tics Assessment Course, "2003 ASEE Annual Conference, Nashville,
June (2003)
12. Moyers, C., andC. Baldwin, i ........1 i evaporative Cooling, and
Solids Drying," Perry's Chemical Engineers Handbook, McGraw-Hill,
New York, 7th Ed. (1997)
13. Cooper, J., W Penney, J. Fair, and S. Walas, Chemical Process Equip-
ment Selection and Design, Gulf Professional Publishing, Burlington,
MA, 2nd Ed. (2004)
14. McCabe, W., J. Smith, and P Harriott, Unit Operations of Chemical
Engineering, 5th Ed., McGraw-Hill, New York, (1993)
15. Geankoplis, C., Transport Processes and Unit Operations, 3rd Ed.,
Prentice Hall, Englewood Cliffs, NJ (1993)
16. Treybal, R., Mass Transfer Operations, 3rd Ed., McGraw-Hill, New
York (1980)
17. Keey, R., "Process Design of Continuous Drying Equipment," Water
Removal Processes, C. King and J. Clark, editors, AIChE Symposium
Series, 73(163), 1-11. AIChE, New York (1977)
18. ABET, "Criteria for Accrediting Engineering Programs," Engineering
Accreditation Commission, (2004) 1


Chemical Engineering Education











curriculum
-0


ENHANCING THE

UNDERGRADUATE COMPUTING EXPERIENCE








THOMAS F. EDGAR
corresponding author, CACHE Corporation


M any chemical engineering departments are I1 ling
with the following questions:

When should computing be introduced to the
chemical engineering student?
How should computer programming in chemi-
cal engineering be taught, and how much formal
programming instruction on languages such as C
should be provided (vs. usage of computing tools
such as MATLAB, Mathcad, spreadsheets, etc.)?
Is a numerical methods course required and
where is it in the course sequence? How many
credit hours are needed? Which department
teaches it?
Should every chemical engineering course include
some computing?
Since the mid-'80s two approaches have been taken to-
ward introductory computing for engineers: "CS 101" and
the engineering tools approach. The "CS 101" approach
was catalyzed by the growth of computer science programs,
which provided instruction in computer languages. Over the
years the "CS 101" courses have migrated through several
programming languages: Pascal, C, C++, and Java. In the
engineering branch, software vehicles such as spreadsheets
(first Lotus 123, then Quattro Pro, and now Excel), TK Solver,
Mathcad, and MATLAB have gradually pushed out program-
ming languages (primarily Fortran). Programming languages
are becoming endangered species in these courses.


The "CS 101" branch would claim a number of reasons
for existencea1:
engineers should learn fundamental concepts of
programming and computer science
computing should be taught by computer scien-
tists, not engineers
engineering faculty are not interested in teaching
computing languages to their students
these courses provide a i'. ,,'. ma number ofstu-
dent credit hours (SCH) and budgetary resources
There are concrete benefits to an engineering education that
incorporates the ability to write computer programs: Students
learn what assumptions go into the program, i.e.,
what the i-.1,t answer should be
what is the input, what is the output
clear ci .i,,,-.i -t, c:f i,. i.,.li. logic, andcalcula-
tions is required
that errors can exist in a program
that programming is unforgivingfor il....., I,,i,. .
and errors

Thomas F. Edgar is the Abell chairin the Department of Chemi-
cal Engineering at the University of Texas, where he has been a
faculty member for 35 years. He is also executive officer of the
CACHE Corporation, a nonprofit educational organization that
promotes development and distribution of technology-based
educational aids for the chemical engineering profession.


Copyright ChE Division of ASEE 2006


Summer 2006


231

















When it

comes to

assessing

computing

needs, faculty

often confuse

what is

important

for their

students

with what is

important for

themselves.


The "Engineering Tools Approach" branch believes:
engineering students need a solid ,' oundimn, in problem solving with mod-
ern computing tools
engineering students need the knowledge and tools required in their profes-
sions
engineering c. 'iii'. and problem solving are best taught by engineers
there is no room in the curriculumfor a separate three- orfour-SCH course
in programming
While the two branches are complementary and most engineering students could
benefit from both courses, most chemical engineering program curricula are too con-
gested to make room for both.
Many departments no longer require a course in a computer programming language
such as Fortran, C, or C++. It has been suggested that teaching computer programming
is analogous to teaching plane geometry. It is a way of thinking but you may not have
to use it. On the other hand, without some programming ability, engineers are limited
by the built-in capability of commercial software without any way to extend it. This
has led a number of departments to switch from teaching C++ to using MATLAB as
the programming tool. MATLAB is a structured programming language that incorpo-
rates many elements of Fortran, C, and C++. It allows for modularity, flow control,
and input/output control and has the following programming features.
1. Loops: like DO and WHILE in Fortran, MATLAB has for and while
2. Conditional statements: like IF in Fortran, C, and C++; MATLAB has if for
testing relational operations
3. Relational operations: like C and C++, MATLAB has the expected suite of
<, >, <=, >=, = =, ~=. And like C and C++, the result of the operation is 1
or 0, and can be used outside of a conditional statement
4. Logical operations: like Fortran, C, and C++, relational operations can be
strung;... d,. i. with AND (&), OR ( I ), and NOT (~)
5. Matching: like C, MATLAB has a switch/case syntax for matching string
variables, integers, or logical
6. 1/O: not only can a user bepromptedfor input and then have results output
to screen in formatted form (using fprintf, as in C), but MATLAB can read
(load) and write (save) to files, in binary or ascii format
7. Modularity: like SUBROUTINE in Fortran or function in C and C++,
MATLAB allows the user to create user-defined functions to be called by a
main program; any number of inputs and outputs can be associated with a
user-defined function
8. Error processing: using the try/catch syntax a user can attempt
calculations) and then gracefully continue execution if an error occurs
9. Array math: like Fortran90, MATLAB transparently accommodates scalar
and array math (i.e., implied FOR loops)
Reasons given by faculty for switching from C++ to MATLAB include ease of use
and widespread availability due to an inexpensive student version. Because MATLAB
is an interpreted rather than compiled language, the user can create (write), debug, and
run code in the same environment. The built-in editor can pass code directly to the
MATLAB application for execution. Also, MATLAB has a solid graphical interface
for creating 2-D and 3-D plots; and plots can be created using appropriate MATLAB
code from within a user's program.
Based on informal surveys many chemical engineering departments now introduce
programming and engineering problem solving in the freshman year. The view is that
Chemical Engineering Education










these subjects are best taught by an engineering department
in the context of an application. A typical introductory course
has the following outline.
problem-solving: engineering method, units, preci-
sion in calculations
symbolic (.. ni/i.. algebra, calculus
spreadsheet techniques: solutions to engineering
problems, Visual Basic for Application (VBA) in
Excel
programming fundamentals: data types, program
flow, modularity, object-oriented features
elementary numerical methods: linear and nonlin-
ear equation solving, linear .. ..., i'. -,
software tools: Mathcad, MATLAB, Excel

When it comes to assessing computing needs, faculty often
confuse what is important for their students with what is im-
portant for themselves. Faculty needs, more often than not,
align with their research interests and activities, and these
may be disconnected from the needs of their undergraduate
students. Also, faculty may have an incorrect impression of
the computing needs of professionals by either being out of
date or out of touch. Discussions on computing needs do not
always proceed on the basis of evidence from alumni and
employer surveys. Finally, computing is not part of the daily
professional existence of most faculty and is not expected to
be. Their computing skills can be oxidized, and most of their
computing is carried out by their students.
In the area of computing software, there is a noticeable dis-
connect between industry and academia. The appendix sum-
marizes a survey of computing practices of recent graduates
in chemical engineering, most of whom now work in industry.
Typically chemical engineering departments teach the use of
MATLAB, Mathcad, Mathematica, or Maple but not the use
of spreadsheets. Yet in industry, spreadsheet software (e.g.,
Excel) is the dominant computer package in use. Of course
this may reflect the nature of many calculations that need
to be performed by chemical engineers in industry, rather
than a need to de-emphasize the teaching of sound numeri-
cal approaches in universities. Some faculty resist teaching
spreadsheets, for example, because it is difficult to analyze
the logic in the code, but this appears to be changing with the
availability of VBA. Another objection is that a spreadsheet
approach can encourage the use of inaccurate or inefficient
numerical calculations (no error control, etc.) For complex
calculations, it may be better to program spreadsheets using
VBA, where programming logic is more transparent.
The survey of industrial usage of computing in the appen-
dix also indicates that less than 50% of recent graduates in
chemical engineering actually perform programming on the
job (although there is no clear definition of what constitutes
"programming" in industry). The use of spreadsheets in
chemical engineering practice appears to be increasing. The
Summer 2006


application of spreadsheets in university courses may be at-
tractive because of student-driven usage. David Clough (U.
Colorado) and Brice Carnahan (U. Michigan) have developed
many examples of spreadsheet applications, as presented at
the 2002 ASEE Chemical Engineering Summer School.

TEXTBOOKS AND AFFILIATED SOFTWARE
The fragmented nature of software tied to leading under-
graduate textbooks makes integration of computing through
the curriculum difficult, e.g.,
(a) material and energy balances: Felder and Rousseau"21
EZ Solve; Himmelblau and Riggs3' POLYMATH
(b) thermodynamics: Sandlert41 MathCAD; Kyle5'
POLYMATH; Elliott and Lira6/ various programs
(c) separations: Wankat71 Aspen
(d) process control: Seborg, Edgar, Mellichamp"8 MAT-
LAB; Bequette9l MATLAB; Riggs'01'- MATLAB/Ex-
cel
(e) chemical reaction engineering: Fogler"' POLY-
MATH; Ekerdt and I., a ,, ,/ Octave/MATLAB
(f) product and process design: Seider, Seader, Lewin'13
-Aspen, HYSYS, CHEMCAD, PROII
In addition to these courses, many departments are teach-
ing a statistics course, which involves still one more software
package such as JMP, SAS, or Minitab. Clearly, using a subset
of these textbooks sequentially through the sophomore,junior,
and senior years will require a student to learn up to five or
more different software packages. Adding software packages
from outside of chemical engineering can push the total num-
ber of packages beyond 10, which becomes problematic for
the typical student. It would be desirable to keep the number
of software packages below three or four if possible [note that
Excel is only mentioned once in (a)-(f) above althoughit is used
with many of the textbooks listed]. But usually textbooks are not
chosen because of the bundled software. In addition, departments
must address issues of software availability, licensing, cost, and
providing software in computer labs vs. student-owned comput-
ers. A textbook that is closely coupled to a software package, a
CD-ROM, or a Web site is clearly an attractive option.

TEACHING PROCESS SIMULATORS
THROUGH THE CURRICULUM
Several departments have found that the difficulty of
integration of computing tools mentioned above can be
avoided by more extensive use of process simulators. It is
quite common to expose students to a commercial simula-
tor in a thermodynamics or separations course. At Virginia
Tech, ChE undergraduates have been using Aspen Plus and
Aspen Dynamics to solve problems in all subjects, starting in
the sophomore year. It is fairly straightforward to convert a
steady-state model inAspen Plus into a dynamic model (with
PID control schemes) in Aspen Dynamics. The applicability
233










of Aspen Plus to mass and energy balances, thermodynamics
(physical and thermodynamic property analysis, estimation
and regression), multicomponent separations, reactor design,
and process flowsheet simulation is well known. In process
control Aspen Dynamics enables students to evaluate con-
troller tuning, process dynamics, startup and shutdown, etc.
HYSYS has similar features, and has been used at Rowan
University for analysis in freshman-senior years.
Recently, Version 2.0 of a CD-ROM, Using Process Simu-
lators in Chemical Engineering: A Multimedia Guide for the
Core Curriculum,14] has become available. Modules and
tutorials are provided for self-paced instruction in the use
of the process simulators to solve open-ended problems in
courses on material and energy balances, thermodynamics,
heat transfer, reactor design, separations, and product and
process design. A 110-page document has been prepared for
instructors suggesting the best instruction sequence and pro-
viding exercises and solutions, for each of the core courses
(first introduced at the 2002 ASEE Chemical Engineering
Summer School).

NUMERICAL AND ANALYTICAL APPROACHES
IN MODELING OF PHYSICAL BEHAVIOR
Historically many engineering courses have been taught
from an analytical viewpoint, but a transition is starting to
occur in which numerical experiments are being gradually
added in fluid flow or heat transfer courses. Problems and
experiments should not be so simplified that they are not
realistically formulated. Students are normally exposed to
idealized fluid flow cases in the curriculum, for which ap-
plication of theoretical concepts results in a solution of a
one-dimensional ordinary differential equation or an algebraic
equation. Therefore it is very easy for them to come away with
the notion that theory is useless for most real-life situations.
Students should be able to select either analytical or numeri-
cal techniques to solve a problem, hence they should learn
the advantages and disadvantages of either approach. Use of
more sophisticated numerical tools such as CFD (computa-
tional fluid dynamics) will reduce the need to make many
simplifying assumptions because you do not need as many
assumptions to solve the problem numerically. Chemical
engineering students should understand that there are both nu-
merical experiments and physical experiments. In some cases
we can make observations from numerical experiments that
you cannot see in physical data, but the converse is also true.
This does not suggest that all derivations should be replaced
by numerical simulation, neither should every experiment be
replaced with a simulator. There should, however, be a bal-
ance of experimental fluid dynamics (EFD), analytical fluid
dynamics (AFD), and CFD.
To prepare students for industrial practice, there should
be a department-level re-examination of the role of detailed
analytical solutions. Is the purpose of some of these exercises


the preparation of undergraduates for graduate school or in-
dustry? Today practicing engineers are not expected to carry
out complex derivations in project work. Once a fluid flow
situation is analyzed theoretically or the governing principles
are discussed, that same situation can be visualized using
the computer. This visualization of the flow phenomena can
significantly facilitate and enhance the learning process,
especially for the visual learner. CFD software makes flow
visualization easy. Students can simulate flow processes in a
transient or steady-state mode. How patterns can be displayed
via velocity contours, velocity vector plots, or graphs of ve-
locity profiles. A key element in flow visualization exercises
is exploring the effects of different parameters. Using CFD,
students can quickly change the size of the pipe, viscosity of
the fluid, size of the particles, velocity of the feed, etc., and
see the resulting changes in the flow behavior. This type of
parametric analysis also ties in nicely with a discussion of di-
mensionless groups and geometric and dynamic similarity.
While computing and visualization can increase understand-
ing, educators do not want students to view such simulators
as black boxes. In the fluid mechanics course, simulations
can become a mathematical exercise with little intuition,
unless the instructor has the students solve a simple problem
by hand first. More work on the software tools is needed,
and it is critical to match the software tool to the student's
knowledge base.
Two specific recent packages that have been developed for
educational usage are FlowLab (a finite volume-based code)
by Fluent, Inc., and FEMLAB (a finite element-based code)
by Comsol, Inc. Based on a survey by Professor Jennifer
Curtis at the University of Florida, about 20 departments
of chemical engineering in the United States expose their
undergraduate students to CFD software. FlowLab allows
students to solve fluid dynamics problems without requiring
a long training period. Using carefully constructed examples,
FlowLab allows students to get started immediately without
having to spend the large time commitment to learn geometry
and mesh-creation skills required by traditional CFD software.
Current exercises that have been developed include sudden
expansion in a pipe, flow and heat transfer in a pipe, flow
around a cylinder, and flow over a heated plate, among oth-
ers. In addition, professors can create their own examples or
customize the predefined ones.
FEMLAB provides ready-to-use application modes, where
the user can build his or her own model by defining the rel-
evant physical quantities rather than the equations directly.
The software also allows for equation-based modeling, which
gives the user the freedom to create equations. FEMLAB's
programming language is an extension of the MATLAB
language; this feature gives much flexibility to the user. FEM-
LAB's graphical interface includes functions for automatic
mesh generation of a user-defined geometry. Recently a k- e
turbulence model has been added to its menu of options.
Chemical Engineering Education









Seider, et al.,[13] present the design of configured consumer products, which usually
involves 2-D or 3-D simulations. In Chapter 19 (Product Design), momentum and spe-
cies balances in a 2-D plasma CVD reactor are employed to produce thin Si films using
CFD packages such as FEMLAB. This illustrates where it is very effective for students to
use CFD packages to optimize designs even without understanding all of the physical and
chemical interactions in the transport-reaction processes.
Even with these recent advances in educational CFD software, this computing technol-
ogy has been slow to penetrate undergraduate transport and reactor engineering courses.
A 2002 CACHE survey of all chemical engineering departments in the United States
on barriers to implementing CFD identified a lack of knowledge concerning available
CFD resources, a lack of professor training in CFD, the relative difficulty of use and
the long learning curve associated with using CFD software in a given course, and cost
of CFD software.

VIRTUAL LABORATORY EXPERIMENTS
Laboratory courses are evolving, and new directions are being examined at specific
universities, combining elements of simulation and also distance learning. In the chemical
process industries, the high cost of pilot-scale equipment and operating manpower has
led to more reliance on computer-based simulations rather than traditional pilot-scale
experiments. During a typical day, the plant engineer works from a control room, or at
least behind a computer screen. An engineer rarely is in the field adjusting valve posi-
tions, flow rates, and temperatures, because that is normally done using the computer
interfaces of distributed control systems.
The fourth-year unit operations laboratory at Texas Tech University is emulating
industrial practice, by providing computer-generated simulations based upon math-
ematical models for laboratory equipment."15 The unit operations laboratory can
familiarize students with safety concerns and operational issues regarding each piece
of equipment. Major pieces of equipment include a double-pipe heat exchanger, an am-
monia gas-absorber packed column, and a cooling tower. The Virtual Unit Operations
laboratory (VUOL) complements the existing laboratory to give students a realistic
experience with industrial operations. LabVIEW computer interfaces of the VUOL
permit students to control the equipment in addition to physically turning valves and
checking temperatures.
In the Texas Tech course each student operates two physical and two virtual experi-
ments. Based on preliminary assessment data, students reported that this type of labora-
tory class contributed either a great deal or considerably in all areas of ABET criteria a-k.
Virtual and physical experiments complement each other and enhance student learning.
In addition, there appears to be no significant difference in the student perception to
their learning in using virtual vs. actual unit operations experiments, in 18 out of 20
ABET-related skill areas. While students believe both types of experiments are valu-
able, a total virtual unit operations laboratory would apparently not be well-received by
the students. With the physical portion of the lab, students get a feel for the equipment
and how it operates. With the virtual portion, the students become familiar with the
computer interfaces that are similar to industrial control rooms, and learn to manipulate
the equipment via those controls instead of manually turning valves and knobs. They
can also explore operating scenarios which are not easily or economically investigated
with physical equipment.
Web-access of laboratory experiments enables real chemical engineering laboratory
equipment to be controlled and monitored interactively by computers that are connected
to the Internet, i.e., under the command of users over the Web. This capability is now
available in the labs at University of Tennessee-Chattanooga as well as other schools


Today

practicing

engineers are

not expected to

carry out

complex

derivations in

project work.

Once a fluid

flow situation

is analyzed

theoretically or

the governing

principles are

discussed, that

same situation

can be visual-

ized using the

computer.


Summer 2006


235










such as University of Texas-Austin, Columbia University,
University of Toledo, and MIT. Such labs permit faculty and
students from any university to run Web-connected experi-
ments at any time of the day or night, any day of the week. The
laboratory station computer operates the equipment (pumps,
valves, heaters, relays, etc.), collects the data (pressure, tem-
perature, position, speed, concentration, etc.)
and sends it to the Web user. The University
of Tennessee site is accessed through the i
Web address ,
and even includes audio and video of the compu
operating equipment. visual
,visual
All established chemical engineering pro-
grams are facing increased financial pressure can ii
to keep existing laboratory experiments up to
date and in satisfactory operating condition. un der
Major operating costs of unit operations lab- edu
oratories include maintenance and teaching-
assistant support. Using highly automated do n
experiments for remote operations will allow
a drastic reduction in TA time requirements stud
for those particular experiments. In addition, vier
by sharing the operation of the experiments
among several universities, there can be simul
a pro rata reduction in maintenance costs.
There is also the opportunity to add experi- black
mental assignments to a lecture class using
this technology. In a lecture class, it may be
desirable to have students individually or in small groups
carry out an experiment, much like a homework assignment;
in contrast, a traditional experiment would require continu-
ous supervision by teaching assistants (e.g., one week of TA
time for an entire class). Therefore, using an Internet-based
experiment can greatly reduce the time commitment by the
TA. It is clear that traditional experiments should remain in
the curriculum to give students "hands-on" exposure, but they
can be augmented with Internet labs.

PROCESS AND PRODUCT DESIGN
Historically there has been a process design emphasis in
the curriculum that is now transitioning to a dual product
and process design emphasis. This means that a framework
is needed to make process decisions to make structured
products. This has added a performance layer, i.e., not just
purity of the product. Given a structure, we can often predict
at some level what the properties of the material are likely to
be. The accuracy of the results and the methods used to treat
them depend critically on the complexity of the structure as
well as the availability of information on similar structures.
For example, various quantitative structure property rela-
tionship (QSPR) models are available for the prediction of
polymer properties. The inverse engineering design problem,
however-designing structures given a set of desired proper-


rhil

tin

iza

ncr

;ta

cat

ot 1

ent

v Si

ato

bo


Chemical Engineering Education


ties-is far more difficult. The market may demand or need
a new material with a specific set of properties, yet given the
properties it is extremely difficult to know which monomers
to put together to make a polymer and what molecular weight
the polymer should have. Today the inverse design problem
is attacked empirically by the synthetic chemist with his/her
wealth of knowledge based on intuition and
experience. A significant amount of work
is already under way to develop the "Holy
Grail" of materials design, namely, effective
g and and powerful reverse-engineering software to
solve the problem of going backwards from a
tion set of desired properties to the realistic chemi-
eae cal structures and material morphologies that
may have these properties. After this is com-
iding, pleted, a subsequent step would involve how
to manufacture the desired new product.
ors
A chapter on Molecular Structure Design in
vant Seider, et al., [13] contains simple optimization
procedures using GAMS to determine poly-
's to mer repeat units, refrigerants, and solvents that
ich have desired properties using group-contribu-
tion methods. Eventually, these will be replaced
irs as (and augmented) by molecular models.
Another subject related to product design is
the scheduling of batch processes, which can
be done using simple simulation techniques, as
inBATCH PLUS and SUPERPRO DESIGNER.
Hence design of optimal processing can be viewed as "product
design" for specialty chemicals. Clearly, spreadsheets and optimi-
zation packages can also be used for many of these computations.
Finally, the use of large databases and software systems, such as
ASPEN IPE, for equipment sizing and purchase and installation
cost estimation, is becoming common throughout the chemical
industries for product and process design.

MOLECULAR MODELING
A molecular-level understanding of chemical manufac-
turing processes would greatly aid the development of
steady-state and dynamic models of these processes. Process
modeling is extensively practiced by the chemical industry
to optimize chemical processes. One needs, however, to be
able to develop a model of the process and then predict not
only thermochemical and thermophysical properties but also
accurate rate constants as input data for the process simula-
tion. Another critical set of data needed for the models is
thermophysical properties. These properties include such
simple quantities as boiling points and also more complex
phenomena such as vapor/liquid equilibria phase diagrams,
diffusion coefficients, liquid densities, and the prediction of
critical points. A key role of computational chemistry is to
provide input parameters of increasing accuracy and reliability
to the process simulations.











Under the NSF grant, "World Wide Web-Based Modules for
Introduction of Molecular Simulation into the Chemical En-
gineering Curriculum," seven university experts in molecular
simulations have developed Web-based modules to facilitate
introduction of molecular simulation into the chemical engi-
neering undergraduate curriculum. These teaching modules
can be integrated directly into chemical engineering core
undergraduate courses, supplying for the instructor and the
student the appropriate linkage material between macroscopic
concepts currently taught in these courses and molecular
simulations designed to aid student understanding of the
molecular underpinnings of the phenomena. Modules are
centered around Java Applets that run the molecular simula-
tions and provide an "experimental" simulation platform for
students to explore concepts. In addition, modules contain
instructor materials, fundamental tutorials, student problems,
and assessment materials.
A consistent Web-based interface has been designed that or-
ganizes all of the material in each module and develops scripts
using perl; this eases the job of putting the written material
into this common format. The developer of a module must
construct simple text files, perhaps with HTML markup that
permits inclusion of figures and tables. Then he or she runs
the files through the perl script, which adds HTML formatting
and links to put the set of files into the common configuration.
The files are uploaded to the module site for anyone to access.
This site is perhaps best accessed through the Etomica site.
Etomica is a Java-based support environment developed for
the modules project, which has now been expanded for other
applications (, contact
is Professor David Kofke).
Following is a list of phenomena and concepts for which
modules are completed or planned:
Chemical reaction equilibrium
Osmosis
Diffusion
Molecular dynamics
Normal modes of a solid
Chemical reaction kinetics
Dissipative particle dynamics
Surface tension
Crystal viewer
Joule-Thomson expansion
Self-assembly
Chemical potential
Multicomponent phase equilibrium
Heat transfer
Atomic billiards
Viscosity

CONCLUSIONS AND RECOMMENDATIONS
One way to foster renewal of the curriculum is to identify
departments where curriculum revision is being carried out


and to evaluate best computing practices and current trends.
There may not be one answer because of different constraints
under which various universities operate, such as number of
faculty in the department and whether computing courses are
taught outside the department. Contributions to this article came
from nearly 20 universities, so we are aware of local issues.
CACHE makes the following recommendations to enhance
computing through the curriculum:
(1) There is increasing pressure on the total number of
hours in the curriculum, especially with the addition
of life science courses. Departments should continue
to re-examine whether a formal three- or four-credit-
hour computer programming course is required for the
chemical engineering degree (vs. teaching how to use
software or write m-files in MATLAB, for example).
The chemical engineering .. -- t i,, course also pro-
vides students with a valuable experience in quantita-
tive problem solving.

(2) The number of software tools that implement numerical
methods used by students should be minimized; de-
partmental agreement on software used in each course
should be reached within the faculty. Faculty need to
reach consensus on how student ,, it,,, skills can
grow systematically .-: ..1, 1. ,',in,, each course in
the curriculum.

(3) Courses such as transport phenomena and thermo-
dynamics offer new possibilities for introduction of
( -,, riis, physical and chemical behavior, such as
with computational fluid dynamics or molecular model-
ing. Process design can add a product design emphasis
by using such tools as well.

(4) Internet-based and virtual laboratories offer a new
means of 1,, vs 1, the student simulation experi-
ence in order to reinforce theoretical concepts.

(5) To prepare students to optimize process designs, it
helps to expose students to process simulators for solu-
tion of a problems) in the core courses of the chemical
engineering curriculum. Also, as software develops and
product design is added at the senior level, instructors
must select from among optimization packages (such
as GAMS), batch process simulators (such as BATCH
PLUS and SUPERPRO DESIGNER), and packages for
S1m1,, 1, equipment sizes and installation costs (such
as Aspen IPE). The use of comprehensive software
packages and databases is common in industrial de-
sign and needs to be introduced in design courses and
used for solution of design projects.


ACKNOWLEDGMENTS
Contributors to this paper include T.F. Edgar, W.D. Seider,
D.E. Clough, J. Curtis, D.S. Dandy, B.L. Knutson, PR. West-
moreland, J.J. Siirola, Chau-Chyun Chen, G.V. Reklaitis, R.
LaRoche, J.B. Rawlings, E.M. Rosen, D. Kofke, M. Cutlip,
M.J. Savelski, and M. Shacham.


Summer 2006


237











REFERENCES
1. Clough, D.E., "ChE's Teaching Introductory Computing to ChE Stu-
dents -A Modern Computing Course with Emphasis on Problem Solv-
ing and Programming," ASEE Annual Meeting, Montreal (2002)
2. Felder, R.M., and R.W. Rousseau, Elementary Principles of Chemical
Processes, 3rd Ed., Wiley, New York (1999)
3. Himmeblau, D.M., and J.B. Riggs, Basic Principles and Calculations
in Chemical Engineering, 7th Ed., Prentice-Hall, Upper Saddle River,
NJ (2003)
4. Sandler, S.I., Chemical, Biochemical, and Engineering Thermodynam-
ics, 4th Ed., Wiley New York (2006)
5. Kyle, B.G., Chemical andProcess Thermodynamics, 3rd Ed., Prentice-
Hall, Upper Saddle River, NJ (1999)
6. Elliott, J.R., and C.T. Lira, Introductory Chemical Engineering Ther-
modynamics, Prentice-Hall, Upper Saddle, NJ (1999)
7. Wankat, PC., Equilibrium-Staged Separations, 2nd Ed., Prentice-Hall,
Upper Saddle River, NJ (2006)
8. Seborg, D.E., T.E Edgar, and D.A. Mellichamp, Process Dynamics
and Control, 2nd Ed., Wiley, New York (2004)
9. Bequette, B.W, Process Control, Prentice-Hall, Upper Saddle River,
NJ (2003)
10. Riggs, J.B., Chemical Process Control, Fernet Publishing, Lubbock,
TX (2001)
11. Fogler, H.S., Elements of Chemical Reaction Engineering, 4th Ed.,
Prentice-Hall, Upper Saddle River, NJ (2006)
12. Ekerdt, J.G., andJ.B. Rawlings, Chemical Reactor Analysis andDesign
Fundamentals, Nob Hill, Madison, WI (2002)
13. Seider, WD., J.D. Seader, and D.R. Lewin, ProductandProcess Design
Principles, 2nd Ed., Wiley, New York (2004)
14. Lewin, D.R., etal., "Using Process Simulators in Chemical Engineer-
ing," CD-Rom for Seider, et al., textbook (2004)
15. Wiesner, T.E, andW Lan, "Comparison of Student Learning in Physical
and Simulated Unit Operations Experiments, "J. Engr. Educ., 195-204,
(2003); see also

APPENDIX
2003 Computing Survey of Recent Graduates
In 1997 the CACHE Corporation carried out a survey of re-
cent graduates in chemical engineering from three universities
to determine how that group used (or did not use) computing
in performing theirjobs. Since that time, there have been con-
siderable changes in the field of information technology. Four
universities volunteered to participate for the 2003 survey:
Carnegie Mellon University, Clarkson University, McMaster
University, and the University of Texas. AWeb-based form was
used to tabulate the responses using database software.
The four universities used different approaches for contact-
ing their recent graduates, defined as students who graduated
during the previous five years (1998-2003). Printed mail,
e-mail, and/or Web forms were used depending on the spe-
cific school. The response rate for the four universities was
estimated to be between 20% and 30%, which actually is quite
good given the complexity and length of the survey (which
took less than one hour to complete). The results of the survey
are available in PowerPoint form on the CACHE Web site,
.
The questionnaire asked for the nature of the work carried
out and the degree level of the respondents. No attempt was
made to remove current graduate students from the sample
even though they are technically not in the workplace. The


overwhelming majority of engineers value computing skills
as critical to industrial problem solving. About 75% of recent
graduates in the survey characterize their work as "technical."
Compared to 1997, there was a gradual increase in the use
of the computer as a general productivity tool. The personal
computer is ubiquitous in all business and engineering work,
including standard office tasks, with 70% of respondents using
a computer actively at least half the day. For the range of using
the computer 3/4 to all day, the percentages doubled from 19%
to 44% between 1997 and 2003.
Of the respondents, 99% report they use spreadsheets on
a daily basis. Faculty have observed that spreadsheets are
used by most if not all undergraduates, often with minimal
formal instruction in the department. Industry clearly values
the use of spreadsheets for a variety of applications based on
the percentage of respondents who use them: data analysis
(88%), numerical analysis (47%), material balances (25%),
economic studies (23%), and other tasks such as financial
modeling or emission calculations (17%).
Similar to spreadsheets, database software (70%) has the
same level of penetration in daily work usage. It is noteworthy
that even with continued improvement of packages such as
MATLAB and MathCAD, they are used much more heav-
ily in academia than in industry (26%). Numerical methods
libraries are only infrequently used (6%), which illustrates
their general decline in popularity since the 1970s.
Less than half of the survey respondents use a process simu-
lator in their work, probably because a growing percentage of
students are working in nontraditional industries (outside the
CPI). Even in the CPI, not all chemical engineers are actively
using simulators in the performance of their jobs.
In 2003 there was more emphasis on and time devoted
to training new engineers to use computing in their jobs
(compared to 1997). There is a continued reliance by recent
graduates on learning new computing skills on their own or
with the help of colleagues. This supports the notion that
universities should prepare their graduates to "learn how to
learn." The amount of formal training to use computing tools
continues to be fairly small.
A majority of the respondents (78%) replied that under-
graduates should be exposed to some form of programming.
This is not surprising even though a minority of engineers
write programs in the workplace. Most people agree that
use of programming logic is an important skill, whether it is
C++, VBA, or MATLAB m-files. Of the respondents, 38%
indicate they write computer programs at work (compared
to 20% in 1997), but it is not clear what actually constitutes
programming in the workplace today (is running simulations
considered to be programming?). Use of VBA along with
spreadsheets is a dominant practice. The growth of usage of VBA
to 34% of the respondents is unimportant development. C++ leads
the rest of the programming options (24%). 1

Chemical Engineering Education











M]j1n class and home problems


Design of a Fuel Processor System for


GENERATING HYDROGEN


FOR AUTOMOTIVE APPLICATIONS


PANINI K. KOLAVENNU, JOHN C. TELOTTE, AND SRINIVAS PALANKI
Florida State University and Florida A & M University Tallahassee, FL 32310-6046


Fuel cell power systems for automotive applications have
received increased attention in recent years because of
their potential for high fuel efficiency and lower emis-
sions.11] While there have been significant advances in fuel cell
technology, this technology has not seen widespread applica-
tion in the automotive industry due to the lack of an efficient
hydrogen distribution system.[2] One option is to develop a
system that utilizes a commonly available carbon-based hy-
drogenous fuel such as gasoline or methane to generate the
necessary hydrogen in situ on an "as needed" basis.
In this paper, the objective is to design a fuel processor
system that utilizes methane to generate sufficient hydrogen
of desired purity, generating 50 kW of power, or enough to
drive a small car [1]

PROBLEM STATEMENT
A schematic of the fuel cell system under consideration
is shown in Figure 1 (next page).[3] Methane enters the fuel
processing system and is converted to hydrogen. Hydrogen
enters the fuel cell where it reacts with oxygen to generate
electrical power, driving an electric motor.
The fuel processing system has a train of three packed-bed
reactors: (1) the reformer, (2) the water-gas shift reactor, and
(3) the preferential oxidation reactor.


Copyright ChE Division of ASEE 2006

Summer 2006


The object of this column is to enhance our readers' collections of stimulating problems in
chemical engineering education. Ideal problems, which may be "open-ended," are those that mo-
tivate the student either by the novel illustration of a particular principle, or by the elucidation of
a difficult concept in a more traditional setting. Practical relevance is encouraged. The text portion
of a manuscript (excluding figures) should not normally exceed 10 double-spaced pages (about
2,500 words). Please send manuscripts to Professor James O. Wilkes (e-mail: wilkes@umich.
edu), Chemical Engineering Department, University of Michigan, Ann Arbor, MI 48109-2136.
Preliminary ideas may be discussed with Prof. Wilkes before submitting a manuscript.


Panini K. Kolavennu has a B.Tech in
chemical engineering with specialization
in biotechnology from Andhra University,
Visakhapatnam, India. He is currently pur-
suing his Ph.D. in chemical engineering
at FAMU-FSU College of Engineering,
Tallahassee, Fla. His research interests
include fuel cell and fuel processor design
S and analysis, model predictive control, and
adaptive control.

John C. Telotte has a B.S. and M.S. in
chemical engineering from Tulane Univer-
sity. He also received a Ph.D. in chemical
engineering from the University of Florida.
He was a member of the chemical engineer-
ing faculty at the University of Wisconsin
and Louisiana Tech University before join-
ing the FAMU-FSU College of Engineering
in 1985. He is currently associate professor
of chemical and biomedical engineering.
His current research focuses on hydrogen
storage problems and design of fuel cell
systems.
Srinivas Palanki received his B.Tech. in
chemical engineering from the Indian In-
stitute of Technology, Delhi, and his M.S.
and Ph.D. in chemical engineering from
the University of Michigan, Ann Arbor. He
joined the faculty of the FAMU-FSU College
of Engineering in 1992. He is currently a
professor of chemical and biomedical engi-
neering. His current research interests are in
real-time optimization and nonlinear robust
control with applications in the fuel cell and
& biomedical areas.


239






























Figure 1. Schematic of fuel cell system.


Based on the Figure 1 schematic diagram, students are
required to complete the following tasks:
1. Write the mole balance equations for the reformer, wa-
ter-gas shift reactor, and preferential oxidation reactor.
2. Calculate the volume necessary for 75% conversion
in the steam reformer. Assume isothermal operation at
1000 K, with the reactor 71" ,ii, at 5 atmpressure.
The flow rate of methane into the reactor is 9 mol/min
at room temperature, and the ratio of steam to meth-
ane is 3:1.
3. Calculate the maximum conversion in the water-gas
shift reactor that can be obtained at 450 K and 600 K,
respectively, and the minimum volume required.
4. Calculate the volume of the water-gas shift reactor to
obtain 90% conversion if:
(a) 20% of the total volume of the reactor is at 600 K
and the rest of the reactor is at 450 K.
(b) 60% of the total volume of the reactor is at 600 K
and the rest of the reactor is at 450 K.
5. The input to the preferential oxidation reactor consists
of exhaustfrom the water-gas shift reactor and air.
The amount of air fed should be adjusted such that the
amount of oxygen in the air is 2.1 times the amount of
CO in the exhaust of the water-gas shift reactor. If the
preferential oxidation reactor is 1 ,'.i,i, at a temper-
ature of 473 K and a pressure of 2 atm, calculate the
volume required to bring the concentration of CO to
below 100 ppm. Assume 90% conversion in the steam
reformer 'j ',ii, at 1000 K and 5 atm and a conver-
sion of 90% in the water-gas shift reactor '' ,iin, at
500 K and 2 atm.
6. Calculate the flow rate of hydrogen ,-,,,, the pref-
erential oxidation reactor. How does this flow rate
change when the flow rate of methane entering the
reformer is changed?
7. Energy Balance:
Calculate the heat of reaction from the heat of
formationfor all the reactions, and list out the


exothermic and endothermic reactions.
Calculate the enthalpies of all the feed and product
streams and use this information to complete an
overall energy balance for the reactor system.
DATA
Steam reformer In this reactor, methane is converted to
hydrogen and carbon monoxide. Part of the carbon monoxide
reacts with water to produce carbon dioxide and hydrogen,
and some methane is totally oxidized to carbon dioxide.
CH4 + H2O 0 3H, + CO (1)
CO+ H,20 CO, + H2 (2)
CH4 + 2HO0 4H, + CO, (3)
Xu and Froment 41 developed intrinsic rate expressions
for steam reforming of methane, accompanied by the water
gas shift reaction on a \i g \ Ig 1 0 catalyst. The following
reaction rate laws were derived:
k p3 P Y
k, I_ 2H CO
p25 PH4PH2O K
H __2 IIK1
(1+ KoPco + K2PH2 + KCH4PH4 + KH20PH /PH2 )2

(4)


k2 p PH2PCO2
PH2 PCOPH20 K2,
r2 (1+ KcPo + KH2 PH + KCH PCH4 + KH20PH2 / PH )2

(5)


i ( 4 P I 1
k3 p Pp2 -H2CP2
p35 PcH4 0 K3

S(1 + KoPco + KH2PH + KCHPCH + KH20PH20 P )2
(6)
where r r and r3 are the rates of formation of CO, CO ,
and CO2 in the reactions represented by Eqs. (1), (2), and (3)
respectively. The P are the partial pressures of the reactants.
The values of the constants are given in Table 1.
The adsorption coefficients can be found using the follow-
ing relations for the respective species
-AH
K= A(K,)exp-- ,where i= H,,CO,CH4,H20 (7)
RT
The rate constants are given by a similar Arrhenius-type
equation.
-E "}
k = A'(k,)exp where j=1,2,3 (8)
RT

The equilibrium constants for the three reactions are given
by the following expression,

K,=exp Aj+- where j= 1,2,3 (9)

Chemical Engineering Education










Water-gas shift reactor In this reactor, most of the
remaining carbon monoxide is converted to hydrogen. The
following exothermic reaction occurs:
CO + H20 CO2 + H, (10)

Choi and Stenger5s] proposed a kinetic model for the
water-gas shift reaction on a Cu/ZnO/Al 20 catalyst operat-
ing between 400 K to 700 K. The following rate law was
developed.


PCO H2,Kq
r4 =k4PcoPHo1 PCOPHoKeq (11)


where r4 is the rate of formation of CO2 in the reaction repre-
sented by Eq. (10). The equilibrium constant Ke varies with
temperature as follows:

Kq exp4778 4.33 (12)

The rate constant k4 follows an Arrhenius type equation
as given below:

k4 = A'(k4) exp -. (13)

Other constant values used are given in Table 1.
Preferential oxidation reactor The stream exiting the
water-gas shift reactor may still have significant amounts of
carbon monoxide that can poison the Polymer Electrolyte
Membrane (PEM) fuel cell electrocatalyst. For this reason, it
is necessary to have a preferential oxidation reactor where the
carbon monoxide from the water-gas shift reactor is reacted
with air to form carbon dioxide. Some of the hydrogen reacts
with the oxygen to produce water.
CO + (1 / 2)0, C CO, (14)
H2 + (1/ 2)0, HO (15)
The following kinetic model was taken from Kahlich, et
al.[61


r5 k p0 42 [2P01
Sp CO J

r6 = 1.5kP 242 02P


where r5 represents the rate of formation of CO2 in the reaction
represented by Eq. (16), and r6 represents the rate of forma-
tion of H,2 in the reaction represented by Eq. (17). The rate
constant k5 follows an Arrhenius-type equation:

k= A'(ks)exp R5 (18)
5 ex p! RT

SOLUTION

Each reactor is modeled as an isothermal plug-flow reac-
tor. It is assumed that no axial mixing or axial heat transfer
occurs. This implies that the reactors are operating at high
Summer 2006


(17)


TABLE 1
Kinetic Parameters for the Three Reactors
Parameter Value
A, 29.3014
A2 -4.35369
A, 25.225
A' (k ) 9.886X 1016 [mol atm5/(m3 min)]
A' (k ) 4.665X107 [mol atm /(m3 min)]
A' (k3 ) 2.386X1016 [mol atms5/(ltr min)]
A (KH2) 6.209X109 [atm-1]

A(Kco) 8.339X10 [atm 1]
A (KH O) 1.77X105

A (KCH4) 6.738X104 [atm 1]

B -26248.4 [K-1]
B2 4593.17 [K 1]
B3 -21825.28 [K 1]
E, 240.1 [kJ/(mol K)]
E2 67.13 [kJ/(mol K)]
E 3 243.9 [kJ/(mol K)]
AHo -82.90 [kJ/(mol K)]

AHo -70.65 [kJ/(mol K)]
co

AH0o +88.68 [kJ/(mol K)]

AHo -38.28 [kJ/(mol K)]
CH4,

A' (k ) 6.195X108 [mol atm /(m3 min)]
A' (k ) 2.333X10" [mol atm 4/(ltr min)]
E14 47.53 [kJ/(mol K)]
E 71 [kJ/(mol K)]

Peclet numbers for both heat and mass transfer. A more de-
tailed analysis incorporating these diffusive effects has been
conducted by Bell and Edgar.[7] The automotive application
puts a constraint on the total volume of the reactor train,
since the entire system has to fit under the hood. The design
equations are solved numerically in MATLAB. The process
is also simulated in the process simulator CHEMCAD. The
results are given below.

Mole Balance Equations
The general mole balance equation for a PFR is given by:
dF
= r(19)
dV
where j represents the species present in the reactor. It is
necessary to determine the reaction rate for each species in
the three reactors.


241

















0.6 \


-H2
-- CH4
-CO
CO2
'- *20


0.5

0.4
o
0.3

0.2

0.1

0
0


-- ------------ --




A AA-A-A
t-0ttAA e --
0 3 4


2 3 4 5
Volume (liters)


(b) CHEMCAD


(a) MATLAB


Volume (liters)


Figure 2. Concentration profiles in reformer.


Inn, Eq. (11) as follows:


rco = -r4

rH2 = -r4

rco2 = r4
rH = r4


(25)
(26)

(27)

(28)


In the preferential oxidation reactor, the reactions taking
place are represented by Eq. (14)-(15). The reaction rates, in
terms of the species, involved can be expressed in terms of
the reaction rates represented by Eq. (16)-(17).


r 2 = -0.5rs 0.5r6

rco = -r5

rco, = r5

rH20 = r6
rH2 = -r6


(29)

(30)
(31)

(32)

(33)


In the reformer, the reactions taking place are represented by
Eq. (1)-(3). The reaction rates in terms of the species involved
can be expressed in terms of the reaction rates represented by
Eq. (4)-(6) as shown below.
rCH4 = -r r3 (20)

roo = r1 r2 (21)

rco = r2 + r3 (22)

rH0 =-r r2 2r3 (23)
rH2 =3r + r2 + 4r3 (24)

There is only one reaction occurring in the water-gas shift
reaction as shown by Eq. (10). We can express the reaction
rate of each Eq. (10) species in terms of the reaction rate of
242


Volume of Steam Reformer

Rate expressions for the different reactions are given in
terms of partial pressures of reacting species. The given mo-
lar feed rate of the gases, F, should be converted to partial
pressures. Using the molar feed rates, we can calculate the
mole fraction of the feed used to calculate the partial pres-
sures as follows:


Xj = j
S=PT
iXJ
1 T


(34)

(35)


The partial pressure obtained is substituted into the rate
expression to calculate the change in flow rate along the
volume of the reactor.
Chemical Engineering Education


0.6

0.5
C
S0.4

i.

(I

I I


- 600 K
..- 450 K


200 250


0 50 100 150
Volume (Liters)


Figure 3. Conversion of carbon monoxide in
water-gas shift reactor.




Full Text

PAGE 1

ta 5 ::j teaching tips ) r This one-page column will present practical teaching tips in sufficient detail that ChE educators can adopt the tip. The focus should be on the teaching method, not content. With no tables or figures the column should be approximately 450 words. If graphics are included, the length needs to be reduced. Tips that are too long will be edited to fit on one page. Please submit a Word file to Phil Wankat , subject: CEE Teaching Tip. TEACHING TIP: ELEVATOR TALKS PHIL WANKAT Purdue University West Lafayette, IN 47907 Both industry and ABET require that engineering gradu ates can communicate. Clearly the best way to achieve this is to have frequent assignments throughout the curriculum requiring writing and oral presentations. Unfortunately, oral presentations tend to require a significant amount of class time. An alternative oral presentation is the "elevator talk." The scenario: a student steps into an elevator with someone she needs to persuade or sell. For example, the student may want to convince the person to hire her. She has from one to two minutes to do this. minutes, the elevator door opened anyway and they had to summarize very quickly. The students saw the relevance of elevator talks and were well prepared. Grading the talks with the scoring rubric was straightforward and I was able to finish the grading while the next pair walked to the front. Since it takes less than 30 seconds to change speakers, 20 two-minute talks can be done in a SO-minute period. While not eliminating the need for more formal presen TABLE 1 I assigned the topic to the stu dents ( ask for a job), gave them the time (two minutes), gave them a copy of the scoring ru bric (Table 1), and told them to prepare a talk that they will present extemporane ously, without visuals. There was no written assignment. In class, I assigned the "boss" for each person. Scoring Rubric for Elevator Talks. Adapted from Mitchell and LawYl tations, eleva tor talks can provide an easy way to include oral communica tion in courses that normally would not have time. Grading all of the talks with the scoring ru bric and then savmg cop ies provides evidence for ABET that all students have been assessed and can do oral presentaAttribute Not Barely Meets Exceeds Acceptable Acceptable Expectations Expectations Logical topic Disjointed; no Parts out of Organized by Superior; order organization order guidelines enhances communication Appropriate Far too long or Somewhat Appropriate time use too short long or short length Objective Not stated Poorly stated Clearly stated Background & Neither stated Only one Both stated Both clearly Significance stated stated Conclusions None Present, but Logical & Logical & supenot logical clearly stated rior explanation Presentation Many Some No distractions Superior mechanics* distractions distractions presentation Response to Not responsive Incomplete Clear and Complete questions (if any) direct Focus on person Not focused; Some focus; Focused with Totally speaking to distracted, no some eye good eye focused; exceleye contact contact contact lent eye contact students were voice, poise, mannerisms told to assume that they knew the boss well enough to talk to. Presenters and bosses went to the front of the room and stood in the elevator. Talks were timed for a strict two min utes. Since two minutes is actually fairly long, most students finished early and had to do something-perhaps just stand therefor the remaining time. If they weren't finished at two tions, at least at the barely acceptable level. REFERENCES 1. Mitchell, B.S., and V.J. Law, "Community-Based Presentations in the Unit Ops Laboratory," Chem. Eng. Ed., 39(2), 160 (2003) Copyright ChE Division of ASEE 2006

PAGE 2

Author Guidelines for the LABORATORY Feature The laboratory experience in chemical engineering education has long been an integral part of our curricula. CEE encourages the submission of manuscripts describing innovations in the laboratory ranging from large-scale unit operations experiments to demonstrations appropriate for the classroom. The following guidelines are offered to assist authors in the preparation of manuscripts that are informative to our readership. These are only suggestions, based on the comments of previous reviewers; authors should use their own judgment in presenting their experiences. A set of general guidelines and advice to the author can be found at our Web site: . Manuscripts should describe the results of original and laboratory-tested ideas. The ideas should be broadly applicable and described in sufficient detail to allow and motivate others to adapt the ideas to their own curricula. It is noted that the readership of CEE is largely faculty and instructors. Manuscripts must contain an abstract and often include an Introduction, Laboratory Description, Data Analysis, Summary of Experiences, Conclusions, and References. An Introduction should establish the context of the laboratory experi ence (e.g., relation to curriculum, review of literature), state the learning objectives, and describe the rationale and approach. The Laboratory Description section should describe the experiment in sufficient detail to allow the reader to judge the scope of effort required to implement a similar experiment on his or her campus. Schematic dia grams or photos, cost information, and references to previous publica tions and Web sites, etc., are usually of benefit. Issues related to safety should be addressed as well as any special operating procedures. If appropriate, a Data Analysis section should be included that concisely describes the method of data analysis. Recognizing that the audience is primarily faculty, the description of the underlying theory should be referenced or brief. The purpose of this section is to communicate to the reader specific student-learning opportunities (e.g., treatment of reac tion-rate data in a temperature range that includes two mechanisms). The purpose of the Summary of Experiences section is to convey the results of laboratory or classroom testing. The section can enumerate, for example, best practices, pitfalls, student survey results, or anecdotal material. A concise statement of the Conclusions (as opposed to a summary) of your experiences should be the last section of the paper prior to listing References.

PAGE 3

EDITORIAL AND BUSINESS ADDRESS: Chemical Engineering Education Department of Chemical Engineering University of Florida Gainesville, FL 32611 PHONE and FAX: 352-392-0861 e-mail: cee@che.ufi.edu EDITOR Tim Anderson ASSOCIATE EDITOR Phillip C. Wankat MANAGING EDITOR Lynn Heasley PROBLEM EDITOR James 0. Wilkes, U. Michigan LEARNING IN INDUSTRY EDITOR William J. Koros, Georgia Institute of Technology PUBLICATIONS BOARD CHAIRMAN E. Dendy Sloan,Jr. Colorado School of Mines VICE CHAIRMAN John P. O'Connell University of Virginia MEMBERS KristiAnseth University of Colorado Pablo Debenedetti Princeton University Dianne Dorland Rowan University Thomas F. Edgar University of Texas at Austin Richard M. Felder North Carolina State University Bruce A. Finlayson University of Washington H. Scott Fogler University of Michigan CarolK.Hall North Carolina State University William J. Koros Georgia Institute of Technology Steve LeBlanc University of Toledo Ronald W. Rousseau Georgia Institute of Technology Stanley I. Sandler University of Delaware C. Stewart Slater Rowan University Donald R. Woods McMaster University Summer 2006 Chemical Engineering Education Volume 40 Number 3 Summer 2006 DEPARTMENT 146 The University of Sherbrooke J. Peter Jones, Bernard Marcos, Gervais Soucy EDUCATOR 154 Susan Montgomery of the University of Michigan Scott Fogler, Lara Zielin CLASSROOM 165 Using A Commercial Simulator To Teach Sorption Separations Phillip C. Wankat 203 A Tire Gasification Senior Design Project That Integrates Laboratory Experiments and Computer Simulation Brian Weiss, Marco J. Castaldi RANDOM THOUGHTS 173 How To Teach (Almost) Anybody (Almost) Anything Richard M. Felder, Rebecca Brent CURRICULUM 175 Hyper-TVT: an Interactive Learning Environment Marina Santoro, Marco Mazzotti 181 Integrating Biological Systems in Process Dynamics and Control Curriculum Robert S. Parker, Francis J. Doyle III, Michael A. Henson 231 Enhancing the Undergraduate Computing Experience Thomas F. Edgar LEARNING IN INDUSTRY 189 The Role of Industrial Training in Chemical Engineering Education Mamdouh T. Ghannam LABORATORY 159 An Agitation Experiment with Multiple Aspects Jordan L. Spencer 195 Validating The Equilibrium Stage Model for an Azeotropic System in a Laboratorial Distillation Column B.P.M. Duarte, M.N. Coelho Pinheiro, D.C.M. da Silva, M.J. Moura 215 Plant Design Project: Biodiesel Production Using Acid-Catalyzed Trans esterification of Yellow Grease Rafael Hernandez, Trent Jeffreys, Anirudha Marwaha, Mathew Thomas 225 Experimental Investigation and Process Design in a Senior Lab Experiment Kenneth R. Muske OUTREACH 211 A Simple Viscosity Experiment for High School Science Classes T.M. Floyd-Smith, K.C. Kwon, J.A. Burmester, F.F. Dale, N. Vahdat, P. Jones CLASS AND HOME PROBLEMS 239 Fuel Processor System for Generating Hydrogen for Automotive Applications Panini K. Kolavennu, John C. Telotte, Srinivas Palanki CHEMICAL ENGINEERING EDUCATION (ISSN 0009-2479) is published quarterly by the Chemical Engineering Division,American Society for Engineering Education, and is edited at the University of Florida. Correspondence regarding editorial matter, circulation, and changes of address should be sent to CEE, Chemical Engineering Department, University of Florida, Gainesville, FL 32611-6005. Copyright 2005 by the Chemical Engineering Division, American Society for Engineering Education. The statements and opinions expressed in this periodical are those of the writers and not necessarily those of the ChE Division,ASEE, which body assumes no responsibility for them. Defective copi,es replaced if twtified within 120 days of publication. Write for infonnation on subscription costs andforback copy costs and availability. POSTMASTER: Send address changes to Chemical Engineering Education, Chemical Engineering Department., University of Florida, Gainesville, FL 32611-6005. Periodicals Postage Paul at Gainesville, Florida and additional post offices. 145

PAGE 4

Chemical Engineering at the University of Sherbrooke J. PETER JONES, BERNARD MARCOS, AND GERVAIS Soucy T he University of Sherbrooke is a French-language university in the city of Sherbrooke, Quebec. It is 100 miles east of Montreal, about 30 miles from the Ver mont border, and approximately due north of Boston, Mass. The Main Campus, which houses the university administration and eight faculties, and the separate Health Campus-with the Faculty of Medicine and Health Sciences, and forming part of the Sherbrooke University Hospital Centre Complex -are all located in Sherbrooke, at the heart of the beautiful Eastern Townships region of southern Quebec. The area is well known for its many rivers, lakes, and mountains. Part of the Northern Appalachian chain of mountains, it is a favored cottage, ski ing, and recreation area for Montrealers, among others. THE UNIVERSITY The university includes nine faculties and offers more than 260 study programs, at both the undergraduate and graduate levels. The university has been experiencing unprecedented development: Since 2001, 650 people have been hired. In fact, more than 40% of its currently employed professionals joined the university within the last four years. In keeping with this growth, the university is presently investing some $310 million to renovate and make additions to existing buildings. Close to 35,000 students attend courses at the University of Sherbrooke, some 85% coming from outside the region; more than 1,300 international students are enrolled on a yearly basis. A study, published annually for the past three years in The Globe and Mail national newspaper, reveals that the Univer sity of Sherbrooke has consistently been the most appreciated university in Quebec and is among the three top-ranked uni versities in Canada. According to MacLean magazine's annual study, the University of Sherbrooke enjoys the best overall reputation in Quebec. The university is first in Canada for the excellence of its cooperative system, which allows students to alternate between study terms and paid-work terms, and for the high quality of services offered to its students. The co-op education system provides a large number of students with paid-work terms, giving them an opportunity to combine the theoretical notions acquired in the formal classroom with their practical experiences as received in the workplace. Students Copyright ChE Division of ASEE 2006 146 Chemical Engineering Education

PAGE 5

Known as The Agora, the campus fountain is a frequent fair-weather gathering spot of Sherbrooke students and visitors alike. involved in Sherbrooke's cooperative education system earn more than $30 million in salaries annually, for more than 4,000 paid-work terms. A LEADER IN RESEARCH AND CREATION The University of Sherbrooke has identified the fields of expertise it intends to develop, in both teaching and research, to meet tomorrow's requirements both nationally and inter nationally. As well, the university is addressing increasing numbers of requests for "partnerships" from institutions in Europe, Latin America, North Africa, and Asia concerning its master's and doctoral degree programs, especially in the fields of education, administration, cooperative management, and applied ethics. It also receives the most royalties of any Canadian university from past inventions by professors and researchers. Thus, it has received more than $79 million royalties to date, including $14.3 million in the 2002-2003 academic year alone. In addition, some 22 spinoff companies have been created by the University of Sherbrooke over the past 20 years. In fact, the university holds title to 300 patents (both established and pending)-51 % having been transferred to businesses. Notable among these innovations is the ACELP technology, developed at the University of Sherbrooke, which Summer 2006 has become the standard in mobile telephony (with more than a billion users) and on the Internet (with more than 500 million users). RECENT EARLY HISTORY Sherbrooke's Department of Chemical Engineering started its existence as a process engineering section in the Depart ment of Mechanical Engineering. Originally, there were three professors: Bernard Coupal, Andre Marsan, and a French military co-operant, Bernard Koehret. They were later joined by Maurice Ruel, Esteban Chornet, and Normand Therien. In December 1971, under the determined leadership of Coupal, the department was established as a full depart ment in its own right, with a distinct program of studies in chemical engineering. The battle to become a separate department was difficult as other departments were wary they would lose scarce re sources, but Coupal advised them that the establishment of a Department of Chemical Engineering could be done at "zero cost." Some of the older professors from other departments long remembered his statement, and used it liberally when later the department was fighting for an increased share of resources. The department continues to be the only one in a Quebec university to have a fully cooperative program. 147

PAGE 6

148 Sherbrooke's Chemical Engineering Faculty Nicolas Abatzoglou's activities are in the areas of thermo-catalytic chemica l reactors and the behavi or of particulat e systems in reactive and nonreactive indus trial processes. His work is being used by companie s for the gasification process commercialization, for conditioning of industrial gases, and for pharmaceuti cal product form ulati on. He has extensive scientific and indu st rial R&D experience in fields at the junctnre of energy and env ir o nment as well as in dry formulation o f pharmaceutical products and ha s initiated a substantial program o n Process Analytical Technology for the pharmaceutical indu stty He teaches va rious courses includin g the capstone design course reaction eng ine ering, pharmaceutical engineer ing processes, and separation and purification in biotechnology. Maher Boulos is a leading figure in the thermal plasma field, along with colleagues in Minnesota, France, Switzerland, and Japan. His work through the Sherbrooke Plasma Research Center has been broadly based and includes many novel experimen tal stud ies but it has also included industrial scale development studies modeling and-with collabo.__._ ..... ...., _. rators-more in depth"th eore tical work. Esteban Chornet i s a l ea ding international figure in re sea rch whose wo rk ha s led to the producti o n of chemicals a nd energy va lues from biomat eria l s and organic wastes. The first chemical engineer in Canada to obtain the prestigious Steacie Fellow ship for research, he is the founder of the research center L..--.::li~-....1 on the transformation of biological materials. Much of his recent work has been applied to environmental concerns (e.g., "green" chemical engineering biomaterials recycling, and environment al concerns). He has initiated a number of high-tech spinoffs in Quebec and has also developed spinoffs in his native Cata lonia. Nathalie Faucheux a biochemist with a Ph.D. in biomedical engineering plans to determine how biomaterials and cells can s har e information with one another. She i s one of the few r esea rch ers in the world working to gain a deeper und erstanding of how biochemical signals, trigg ered by contact with a bioL. ___ .1..;:...,....i,:i material, activate a cell 's capacity to survive, multiply, and function. The cuttingedge materials she is using are based on grafts of small molecules called peptides, which, among other things, promote cell adhesion. She holds a Tier 2 Canada Research Chair. Fran~is Gitzhofer was recruited for his expertis e in materials engineering from the U niversity of Limog es France He has become director of the departm ent's Plasma Research Gro up which is continuously evolving t oward a broader role in fuel cell develop ment a nd other ene rg y -int e n s ive applications. His focus i s on creating new ways of making coatings on various substrates, using both established and newly emerging plasma technologies. Denis Gravelle, one of the earliest professors to join the department, is an expe1t in the application of thermodynamics to thermal plasma system s and in the use of spectroscopic methods/techniques for a fuller understanding of plasma torch dynamics. A key teacher of thermodynamics at both the graduate and ______ undergraduate levels he i s also inv o l ved in the lab courses aiming at the int egra tion of basic engineering concepts of thermodynamics trausprnt phen omena and reaction kinetic s Michele Heitz the a s socia te dean for students (Engi neering Faculty) and a full professor, teaches introduc tion to chemical engi n eering, thermodynamic s, and chemical thermodynamics. She is also in charge of integrating projects for both the chemical and the biotechnological first-year students. Since 2002, she has also taught air pollution control and design, introduction to biochemical engineering, and chemical kinetics and reactor engineering. Her current research projects include air treatments by biofiltration as well as biodiesel production and biomass and whey valorizationboth topics involving various chemical a nd biotechnological approaches. Peter Jones arrived at the She rbrooke shortly after the depa1tm e nt was created. He ha s developed a research area in indu st rial water treatment and th e application of stat i st ical m e thods to environmental problem s, and more generally, to experimental re search. A past chairman of the department he was director of the environmenta l engineering and sci ence mast er's program and vice dean for research. Jerzy Jurewicz is a specialist in the area of plasma generation, using either direct or high-frequency AC currents. He apr---=----, plies this expertise to the development of new reac tors for the synthesis of new products, especially for nanometric powders. Jer zy i s responsible for courses in safety in the first yea r and also in process sa fet y courses in the fomth yea r and is a key member of the ftr,'l'T':S"'l'T'TT"""'1T1 design course team. Bernard Marcos has done much work in the field of expe rt systems and neural networks as well as pursuing educational research such as the use of a system of intelligent tutorials. He was th e first director of the new program in biotechnologic a l engineering. Pierre Proulx has been involv ed in mathematical modeling of thermal plasma s si nc e his graduate studies under the supervision of Maher Bou l os in the P la s ma Research Center Hi s current projects entail m at h e matical modelin g of comp l ex r eactors. He teaches tran s port phenom e na and process control. Joel Sirois th e most r ece ntl y hired .__...__.._ _. profes sor comes with a very strong background in biotechnol ogy and was recently the chief techn ology officer at a sta1t -up company in this field. His focus is in the areas of characterization, modeling and optimization of cell metabolism and the design and scale-up of bioreactors. Gervais Soucy is current chair of the department. His teaching is concentrated in unit operations His research field is in the application of n ew technolo gies for various processes in th e aluminium indu stry. He ha s also de v eloped expert i se in thermal plasma technology to produce carbon nanost ructur es. Normand Therien r ece ntl y retired from the depart ment i s a n expert in the application of mathem atica l techniqu es to env i ro nmental problems. His most important work has been in the area of modeling of hydroel ectric reservoirs, where he has been au important figure in determining how mercury can accumulat e in reservoirs and render fish unfit for human consumpti on. Patrick Vermette is a research er at the Research Centre on Aging and i s an engineer and professor in the department with a joint appoint m ent in the Service of Ort hopa e dics. He ha s built a state-ofthe-art laboratory for s urfac e sc i ence and tissue eng in ee ring s tudies. He and co-workers are in vo lved in fields including biomaterial s angiogene s i s colloids and int erface science drug delivery systems bioreactors, tis sue engineering, and haemat opoietic stem cells. C hemical Engineering Education

PAGE 7

THE COOPERATIVE PROGRAM The Faculty of Engineering at the University of Sherbrooke was a very early adopter of the cooperative system of engi neering in North America. The University of Waterloo had initially adopted the cooperative system in 1956, and the University of Sherbrooke followed in 1966, before the De partment of Chemical Engineering was formed. The co-op system seeks to prepare students for their future career(s) by providing the practical experience that meets employer requirements in the workplace. Thus, the work term offers students the opportunity to acquire practical experience and to develop competencies (knowledge, skills, attitudes, values, etc.) relevant to their future careers. Cooperative education, a pedagogical approach whereby students spend alternate trimesters studying in the classroom and earning wages in the workplace, also offers many valuable features to potential employers. As a pioneer of cooperative education in Quebec, the University of Sherbrooke is proud to be a leader in this expanding field. A result is that the University of Sherbrooke now ranks second in Canada-and is among the top five advanced learning institutions in North America-for the importance given to its cooperative education system. In the Department of Chemical Engineering, students are able to achieve this gradual integration by switching alternate trimesters between their paid terms in the workplace and their study terms at the university. When students graduate, they will have served five work terms (15 weeks/ term) in an industrial company (90%) or in a research laboratory (10% ). Since there are eight academic sessions, the student obtains his B.Ing. degree after a period of 52 months. THE UNDERGRADUATE PROGRAMS Chemical Engineering The Department of Chemical Engineering has always kept up to date with the needs of employers in Quebec and the rest of Canada. The studies program in chemical engineering was com pletely overhauled for students set to begin in September 2001 and graduate in December 2005. Table 1 presents the curricu lum of the reformed chemical engineering program. We have since initiated five cohorts, or graduating classes, to this new TABLE 1 Chemical Engineering Session Description S-1 Introduction to chemical engineering The role of the engineer, safety and risks, chemistry, communications S-2 Measurement techniques for use in the laboratory and the factory/plant Instrumentation, chemistry, chemical analysis techniques, reports, controls T-1 Workterm#l At the end of the first year, the student should be capable of describing the chemical engineer's role and of undertaking control actions and performing analyses, both in the laboratory and at the plant; thereby displaying, at an early stage, a satisfactory competence level in performing the necessary tasks. S-3 Transport and exchange in fluids Fundamentals of chemical transport/transfers in processes T-2 Workterm#2 S-4 Design of the basic units employed for a chemical process Advanced chemical transfer/transport, chemical reactor and associated units calculations At the end of the second year, the student should be capable of modeling the operations of several parts of a functioning chemical process plant. T-3 Workterm#3 S-5 Industrial scale plant operations Control methods, techno-economics, orocess control laboratories S-6 Design basics of industrial-scale chemical processes Types of processes, process simulation, environmental and safety aspects At the end of the third year, the student should be capable of designing the unit parts and creating the basic overall process concept for an industrialscale process. T-4 Workterm#4 S-7 Combining process design skills and experiences I Integration of all aspects required to establish, modify, and operate a chemical industry installation at an important scale T-5 Workterm#5 S-8 Combining process design skills and experiences II Integration of all aspects required to establish, modify, and operate a chemical industry installation at an important scale At the end of the fourth year, the student should be capable of designing the unit parts and creating the basic overall process, taking into account relevant aspects of process economics as well as social and environmental issues. Summer 2006 149

PAGE 8

Sherbrooke students pitch in on a project in one of the university's well-equipped labs. regime. This overhaul was dictated by our knowledge of the companies that hire our graduates. We were also influenced by the tradition of innovation in pedagogical methods used in the university's Faculty of Engineering. We have aligned our developments with the goal that engineering graduates of our programs will be responsible for the development of new products and processes. Most of these new products and processes are not even mentioned in traditional course materials for aspiring chemical engineers, yet our students must now develop ability to work creatively in these areas. We have therefore developed three distinct avenues for go ing forward: Students are responsible for their own development, which is central to the educational process. They must combine their technical development with the simultane ous improvement of their leadership skills, their entrepre neurship, their teamwork skills, and their respect for their profession. Students must now take full and early advantage of the new computing and information technologies at their dis posal, especially the software products specific to their profession and their technical competence as chemical engineers. Courses cannot be approached separately. We and the students must find the commonalities through the session al projects program, activated during each of the first, third, and fourth years. The result has been an immersion in the practice of chemi cal engineering from the very first session. Students were often ill-prepared for some of the challenges they faced, but through their initiative and determination they were able, nevertheless, to produce very interesting results. The projects performed in those early sessions required a lot of "digging" to find pertinent information. This information was then ap150 plied to experimental setups, which they also had to design. They did, however, have considerable help from professors leading the course(s) for each session, along with assistance from departmental technicians. The project is a capstone design project, with which we have now had considerable success for a number of years. These projects are presented at the CSChE competitions. We have won the SNC-Lavalin prize a number of times. The 40 courses offered in the program are distributed in the following manner: 11 courses as general engineering courses in mathemat ics, thermodynamics, and materials 17 courses in chemical engineering, transport phenom ena, unit operations, and reactor design six courses in humanities and social science, law, ethics, and engineering economics six courses in the students' chosen major Students are mentored by their more-senior peers, who help them become accustomed to the range of department opera tions and also provide them with professional contacts at the very beginning of their professional careers. The unifying projects chosen for each session are an excellent initiation to the later work terms, following two sessions spent in the department. Through the projects students also learn about the human and societal aspects of their chosen profession. Biotechnological Engineering The numerous recent developments in biotechnology and in the medical sciences have led the Engineering Faculty to readapt its curricula under the belief that the most precious asset of a profession is its intellectual core. In an era of rapid evolution in biotechnology-based industry, it is imperative that the biotechnological engineering discipline define its own core. It must strengthen its core through scholarly activities and diverse applications. Considering that biotechnology constitutes a broad field, biotechnological engineers need to integrate skills in engineering principles, process engineering, and biological sciences, without being restricted to particular applications. Biotechnological engineering programs must take into account the complexity of living systems with their discrete and nonlinear relationships. The integration of complex engineering principles is not a simple task, and the biology-engineering barrier is an obstacle that has to be overcome. It is not sufficient to incorporate biological science courses into a chemical engineering curriculum hoping that students will be capable of integrating both concepts. Bio technological engineers must eliminate the present gap and correct misunderstandings between traditional engineers and biologists. They must accept the fact that living organisms are not entirely predictable. Consequently, they must master basic knowledge of living organisms and bioproducts, and of the fundamental unit operations and simulation tools used Chemical Engineering Education

PAGE 9

by engineers. They must understand the physiology of pro karyotes and eukaryotes as well as the engineering concepts used in bioprocesses. Biotechnological engineers must also be able to operate and control smalland large-scale culture systems of cells and micro-organisms for the production of products of commercial potential (e.g., proteins, antibiotics) as well as the downstream processing, including separation and purification of biomacromolecules. Finally, they must be skilled in project management and quality control. Broadly speaking, biotechnological engineers will be called upon to solve problems through the development of bioproducts and bioprocesses that use living organisms or the products they synthesize. The biotechnological engineering program at the University of Sherbrooke was originated by the departments of biology and chemical engineering. It took four years to build the program, which now offers an integrated training in biotech nological engineering. Table 2 presents the curriculum of the new Biotechnological Engineering program at Sherbrooke. The program is divided into eight terms that include labora tory studies, applied projects, and lectures. Creating a new discipline may present some drawbacks for the employment of new graduates. Industry will need to learn what a biotechnological engineer is-just as it understands what chemical engineers and biologists are. That is why the biotechnological engineering program was developed with industrial partners. These industrial partners are regularly updated on the curriculum's development. Another area of concern is that biotechnological engineers may be too nar rowly trained and too application-oriented. As explained previously, the biotechnological engineering program is a science-based program that should ultimately alleviate a too-narrow perspective. The goal of the training is to prepare a generalist engineer who is able to manage the evolution of the biotechnology industry. The wide breadth of knowledge of tomorrow's biotechnological engineers will be an important advantage to them. Several segments of the bio-industry are relevant to the employment of future biotechnological engineers: bio pharmaceutical and drug companies; agribusiness and food companies; environmental biotechnology companies; bio medical instrumentation companies; biomaterials; and the tissue-engineering sector. TABLE2 Biotechnological Engineering Session Description S-1 Introduction to Biotechnological Engineering Role of the engineer, safety and risks, biochemistry, information technology S-2 Introduction to Biology Biology, cell biology, functional biology, microbiology, laboratory techniques At the end of the first year, the student should be able to describe the role of the engineer and to make measurements and perform laboratory analysis. S-3 Genetics and transport phenomena The fundamentals of genetics and chemical transport T-1 Workterm#l S-4 Design of basic units Advanced transport, unit operations, experimental protocols At the end of the second year, the student should be able to model the behavior of a number of the units that make up a biotechnological process. T-2 Workterm#2 S-5 Operation of industrial unit processes Bioreactors, process control, engineering economics, biological polymers T-3 Workterm#3 At the end of the third year, the student should be able to design a process and the functional units to make it operate. S-6 Downstream operations Separation and purification, materials and biomaterials, biomolecular engineering T-4 Workterm#4 S-7 Design of biotechnological processes GMP-GLP standards, process simulation, design S-8 Integration of the abilities required to design a chemical process Integrate all the aspects related to the building, the modification, and the operation of a large biotechnological industrial installation At the end of the fourth year, the student should be able to design a process to include all of the economic, environmental, safety, and societal aspects for functioning in today's marketplace. Summer 2006 151

PAGE 10

Double Degree Program with Bishop's University Sherbrooke students may choose to earn a double degree in engineering and liberal arts in a joint program with Bishop's University. The engineering program includes four, four month, paid-work internships. The remainder of the liberal arts program is undertaken by taking selected courses at Bishop's University, located a few kilometers away from the University of Sherbrooke. The liberal arts degree provides for a broad education in the social and human sciences, allowing students to develop fully as individuals in a chosen humanities specialty: history, literature, philosophy, fine arts, or theater. This exclusive program is the result of close cooperation between the University of Sherbrooke and Bishop's Univer sity. It allows students to study in both of Canada's official languages while also experiencing the two unique and distinct university cultures. Project-Based Learning Approach The project-based learning approach is used to integrate coursework within an academic session in first-year projects and the third-year and final-year design project. Project-based learning is a new approach in the educational field. It was cho sen by the department to fill new professional and industrial requirements. The main features of our approach are: The project involves solution of a problem taken from a "real world" case. For instance, the last-year design project was an oil sands exploitation situation; the first year project was the design of a process for the valoriza tion of lactoserum. The student is responsible for this "on the job" learning and the initiative is to be taken by the students. The class is divided into small teams and collaborative learning principles are used to find and share new knowledge. At the end of these projects, teachers have often noted that their students have improved appreciably in the applica tion of self-learning skills. The project results in a "deliverable" (i.e., a process, product,jlow sheet, report). In the last year, the deliv erable (for the last-year project) was the design and specifications for the oil sands plant; the end product (for the first-year project) was the design and monitoring of a pilot plant for lactoserum valorization. The work lasts for a realistic amount of time. Each proj ect is spread over two sessions. Professors involved are considered an advisory commit tee and the approach is student-centered. The teacher team performs the "follow-up" each week and provides advice, if necessary. The team realizes the assessment of the project. The project is also evaluated by other students. Students subsequently present the project at public con152 ferences and the annual Canadian Chemical Engineering Congress. During the last 10 years, Sherbrooke students have won many awards at the Congress of the Canadian Society for Chemical Engineering for the quality of their first-year and final-year project presentations. These projects enable students to substantially improve their mastery of both oral and written communications. RESEARCH Research has been a very significant part of the department's activities from its beginning. The department received fund ing of more than 4,000,000 CAD last year. This funding is primarily in the form of grants, so there are fewer overhead charges than American schools are likely to bear. This fund ing has been principally used to support graduate students and researchers, as well as to build and maintain very well equipped laboratories. Research in the department is conducted in a number of areas, including aluminium production technology, biomass conversion, biotechnology, environmental engineering, fuel cell technology, flow modeling, and plasma technology. The research has evolved over the years as a result of recruitment of professors and substantial research funding. In 1973, Bernard Coupal obtained a very substantial grant to develop applications for peat moss during the first year of the department's existence as an independent entity. This grant provided a very important impetus to work in the department and led to a very substantial and lasting work program on biomass transformation. Environmental engineering developed because we started a master's degree program in environmental engineering and science. Biotechnology research was developed as we recognized the need for specialists in our new biotechno logical engineering program. The interest in this research has progressed hand in hand with the hiring of new professors working in the field. Thermal plasma technology was not common in chemical engineering departments before the 1970s but the arrival of Maher Boulos led to very substantial growth in this area at Sherbrooke. The need to model gas/plasma flows in plasma torches led to general interest in the modeling of flow, heat trans fer and kinetics, and particle behavior in a variety of systems. Fuel cell research evolved because of our specialized knowledge, which already existed in the plasma research center for materials, especially deposition on surfaces using thermal plasmas. Other research "specialities" were taken up because of their importance to certain industrial sectors existing in the Quebec economy, notably in aluminium production. The pharmaceutical industry is very developed in Quebec, and we are consequently developing very active research in this area, including PAT (Process Analytical Technology). Chemical Engineering Education

PAGE 11

Testimonial of a doctoral student's experience in a ioint program with a universit,y in Europe "The Department of Chemical Engineering at the Uni versity of Sherbrooke provided me with an opportunity to do my doctoral research in collaboration with a Belgian laboratory at the Catholic University of Louvain (UCL). This collaboration lies within the scope of an agreement of joint direction established between the two universities. "This European collaboration enabled me to work for some two-and-a-half years within the Bioengineering Unit at UCL. I discovered innovative enzymatic methods for the elimination of recalcitrant phenolic compounds. This experiment also allowed me to establish a network with other European laboratories working in fields related to my research task." GRADUATE EDUCATION The department has provided a strong graduate program from its earliest days. Professors who were in the process section of the mechanical engineering department had large contingents of graduate students. Upon creation of the chemi cal engineering department, we were accorded, in addition to our existing undergraduate programs, the master's and doctoral programs. There are presently approximately 50 post-graduate stu dents working in the department. We graduate about eight master's degrees and five Ph.D. degrees per year. Summer 2006 Winter visitors to Sherbrooke will see some of the prettiest scenery snowy climes have to offer, such as this view of nearby St. Benoit-du-Lac Monastery. INFLUENCE OF OUR PROFESSORS IN CREATING HIGH-TECH SPINOFFS The University of Sherbrooke generally and the Depart ment of Chemical Engineering in particular have been very successful in obtaining licensing fees and creating spinoff companies. Esteban Chomet, with the contribution of Nicolas Abatzoglou, has created a number of companies, many very successful in the tasks of coproducing useful products and energy from biomass and organic wastes. Maher Boulos has created a company, based in Sherbrooke (Tekna), with more than 40 full-time employees, specializing in the area of thermal plasma technologies. This company exports its products all over the world, achieving particularly strong results in Japan. THE FUTURE The future is bright. We are recruiting large numbers of students into the biotechnological engineering program. The incoming class recruitment in chemical engineering is stable, with approximately 35 new admissions per year. We are in the process of recruiting additional professors for both chemical and biotechnological engineering. The University of Sherbrooke provides a distinctive edu cational experience because of its "French" character. We encourage those predominantly English-speaking students who also have knowledge of "basic French" to come to Sherbrooke to simultaneously "perfect" both their chemical engineering and their French language skills for use in their future careers-in Canada, and beyond! 0 153

PAGE 12

dj i] =1 educator ) 111111-1111-iil..------Susan Montgomery ofthe University of Michigan ScoTI FOGLER AND LARA ZIELIN The University of Michigan Ann Arbor, Ml 48109 U ndergraduate program advisor Su san Montgomery was three years into a tenure track position at the University of Michigan (UM) in education research when she realized that teaching and advising were her true passions. Driven by those passions, Susan did the unconventional thing and became a lecturer in 1999. Since then, she has been a "mom" for over 1,000 ChE students, who appreciate the warm and supportive community she helps create within a big university atmosphere. "Susan Montgomery literally holds together the undergraduate curriculum at UM," says Ron Larson, chair of UM's chemical engineering department. "She is the most appreciated faculty member among the undergrads. The key element that makes her successful is her singular focus on the students and their needs. While other faculty members also care deeply about students, their research and administrative portfolios limit the extent to which they can involve themselves in the concerns of the students. There is simply no substitute for having a member of the faculty who is devoted ex clusively to the students." This is no small feat in a ChE program as large as UM's. Total enrollment hovers around 350 students. "UM can be daunting," explains Larson. "Even within the 'com munity' of a department, students can get lost." Susan works hard to combat this by maintaining a connection with each chemi cal engineering class. "Not only does she know every undergraduate that comes up through the program, she keeps track of them as they move on to their future careers," says Larson. 154 While she loves the classroom and has taught a wide range of undergraduate classes from introduction to engineering to process design Susan has a hard time choosing a favorite topic. "The real fun is seeing students transition from one phase of their careers to the next," she says. "The thrill is watching students blossom." Susan works tirelessly to make students' growth and advancement a reality. In addition to her advising and lecturing, Susan is the principal author of the Visual Encyclopedia of Chemical Engineering Equipment, a CD-ROM designed to help beginning ChE students understand how chemical engineering equip ment works. The CD-ROM stemmed from her research in the Multimedia Educational Labo ratory (MEL) at UM, which focused on studying the diverse learning styles of chemical engineering students, and developing multimedia educational software to address those learning styles. Susan then analyzed student use of this software to discern what types of interactions were preferred by what students. The goal was to help future educational software developers better understand the role that different interactions could play in addressing the needs of a variety of learners. Copyright ChE Division of ASEE 2006 Chemical Engineering Education

PAGE 13

CYCLO N ES / HYDROCYCLO N E S: Equipment Des i gn The picture shows the parts of a cyclone For more infonnation on a.given pati, click on its name. For a given particle size, a portion of the particles will exit out the bottom with the underflow, and the rest wilt exit with the o'l/erflow. The heavier the particle, the greater the chance that it exits out the bottom. The cutpoint size, a measure of a cyclone's petfonnance. is defined as the particle size for which half the particles exit at the top and half at the bottom A coarser O.arger) cutpoint means that on l y larger particles can be separated A finer (smaller) cutpoint means that smaller particles can be removed PAGE10F2 CHEMICAL ENGINEERING EQ.UIPMENT: Main Menu Click on eac h of the catego 1 ies be l ow to access the corresponding pi e ces of equi p ment, o r click o n the index button to see a com p lete alphabetical list PROCESS HEAT TRANSFER REACTORS PARAMETERS FLOWMETERS SEPARATIONS : POLYMER CHEMICAL PROCESSING TRANSPORT SEPARATIONS : MATERIALS ANDSIURAGE MECHANICAL HANDLING QUIT HELP INDE X Figure 1. A screenshot from Susan s encyclopedia CD-ROM showing an oversize view of equipment, left, and the main navigation page, right. The CD-ROM includes animations and pictures of real equipment as well as examples of applications of the equip ment. From the main menu, the user can branch into a variety of different topics and operations, such as the corresponding pieces of equipment for processes including heat transfer, reactors, materials handling, and more. The CD-ROM encyclopedia has been extremely well re ceived both in academia and industry. Figure 1 shows a screen capture from the CD illustrating a sample introductory over view of an item of equipment-cyclones and hydrocyclones in this instance. The possibilities for illustrating chemical engineering practice are almost endless, and many of the screens in the CD show dynamic operation of the equipment, as in the case of bubble-cap distillation columns, screw ex truders, filters, and cyclones, to mention just a few. When a visit to a chemical plant or refinery cannot be made, the Visual Encyclopedia is an excellent substitute for illustrating various operations that might otherwise have to be described more passively. It has been used for numerous industrial training courses. The CD is also included in two of the most popular textbooks in the field: Rich Felder's Elementary Principles of Chemical Processes; and ScottFogler's Elements of Chemical Reaction Engineering. In addition to the Encyclopedia CD-ROM, Susan and her colleagues have developed two other CDs. The first, titled Engineering Fundamentals in Biological Systems, provides real-world applications in fundamental processes, such as material balances on an artificial kidney. The second CD, Material and Energy Balances, provides interactive problem solving in real-world environments including the car pre painting system in Ford Motor Company's Wixom Assembly Plant, and Ann Arbor's wastewater treatment plant. The CD also includes tutorials on Pxy-Txy diagrams, psychrometric charts, and enthalpy-concentration diagrams. The CDs are distributed through the CACHE Corporation. Summer 2006 FAMILY TIES Susan's background and upbringing have much to do with her success in the field. She was born in Peru, of a Peruvian father and an American mother. Her father is a civil engineer who shared his passion for engineering with his daughters. He used to take Susan and her sister to job sites, showing them how things worked. Susan recalls going for a walk with her dad when they chanced upon a street replacement project. Her dad showed her all the layers that made up a street, and they met the construction workers. Susan and her family lived for a year and a half in Ni caragua, where her father was an Agency for International Development consultant with their ministry of public works. Forced to leave suddenly in 1978 when a civil war broke out, Susan, her sister, and her mother moved to Ann Arbor, Mich., where her American grandmother lived. The summer after her junior year in high school, Susan attended a Women in Science and Engineering program at Carnegie Mellon University and came home announcing that she intended to be a chemist. Her father corrected her, "reminding" her that she would be an engineer. Susan graduated from high school at age 15, a feat she attributes to an excellent kindergarten that allowed her to complete first, second, and third grades in one year. She completed her undergraduate work at UM, then went on to Princeton University for her graduate work. Susan's advisor at Princeton was Professor Ludwig Rebenfeld, of the Tex tile Research Institute, where her research focused on flow through porous media. But her passion was in teaching, and her goal was a faculty position. Susan had doubts, however, about her ability to complete the Ph.D. Thrice during her time at Princeton, she announced to Prof. Rebenfeld that she was quitting the program. Prof. Rebenfeld offered unwavering support, but two other factors 155

PAGE 14

Right, Susan celebrating her fourth birthday. Far right, a beaming Susan on her graduation day at Princeton. Below, Susan and her older sister Betsy proudly display a snow man they "engineered." made her tough it out and complete the program. The first factor was knowing that without a Ph.D., she would not be able to fulfill her goal of becoming a faculty member. The second was the scolding of her grandmother, Margaret Hampshire, an independent woman with whom she lived while an undergraduate at UM. Where oth ers, upon hearing that she was considering dropping out, offered condolences, her grandmother replied with comments such as, "We didn't work this hard for you to drop it all now. You get back there and get that Ph.D.!" Perhaps this is why Susan can now speak so well to the number of women students who are struggling to find their place in the ChE program. "After a month, they may come to me feeling like maybe they're not good enough," she says. "I encourage them not to focus on whether or not they're 'good enough' but on whether or not this path is taking them where they want to go." The approach seems to work. Susan says a number of alumni have thanked her for the advice she offered early on in their academic careers. "They're out there doing what they love now, so they can look back and be glad they stuck with the program," she says. Susan has the strength to help many, but when she needs someone at her side she calls on her sister, Betsy Vera. Susan says Betsy has stood by her continuously. They remain best friends and stay close, even though Betsy now lives in Chicago. "I have learned so much from my family," says Susan. For ex ample, she recalls watching her mother complete her undergradu ate studies in her 40s, and then go on to earn a master's degree in Latin American Studies from Georgetown University. "Now that I 156 am in my 40s, too, I appreciate having my mother as a role model for living the life you are meant to live," she says. Susan also says her mother's death at the young age of 52 solidified the importance of living your life now, versus waiting until retirement to do things you've always dreamed of. ENTERING THE FIELD Susan got her first job on a fluke. Like many under graduate students, she was struggling to find funding for spring semester of her third year. She was just leaving the dean's office, disappointed, when she ran into Scott Fogler, a UM professor of engineering, whose class she had just finished taking, and acing-she had been the top student. He inquired about how things were going and, upon learning of her situation, he immediately of fered her a summer job working in his laboratory and helping him with his textbook. This marked the begin ning of more than 20 years of collaboration in various educational projects. Susan's first teaching position was as a TA for the junior-level laboratory course at Princeton, where she caught the bug for education. "I could see the light bulb go off in students' minds," she says. An internship at a local community college teaching pre-algebra at night made her intrigued about learning styles, and the deChemical Engineering Education

PAGE 15

velopment of learners through their college careers. She was quite intimidated when asked to be responsible for the whole laboratory course at Princeton the following year, but that experience only cemented her decision to become a faculty member. During this time, Susan also recalls going to the engi neering library at Princeton to search for a research article in Chemical Engineering Science. Instead, she discovered Chemical Engineering Education, stacked right next to the journals she was supposed to be reading. All plans for the afternoon were scrapped as she spent long hours browsing through CEE instead. Also during her time at Princeton, Prof. Rebenfeld supported her attendance at the 1990 ASEE National Con ference in Toronto, back in the days when few graduate students attended the conference. Participants at that time stayed in college dorms, which really helped colleagues get to know one another better. "I always speak of this conference as the time in which I 'found my people'," she says. "Their dedication and passion for teaching matched my own." After graduate school at Princeton, Susan returned to UM to complete a two-year postdoctoral appointment developing educational software for chemical reaction engineering and problem solving. Once again, she col Summer 2006 Left, Susan poses with her sons and her father at the Plaza de Armas in Lima, Peru. Above, Susan (middle) and her sister, Bet sy Vera,flank Prof Dale Briggs at a 1984 undergraduate gradu ation reception; to her left are her grandmother and mother. laborated with her former professor, Scott Fogler, along with two dozen outstanding chemical engineering students. It was these stu dents who taught her how energizing it can be to supervise teams of undergrads. Again, Susan experienced a sea change. "At the time, my plans were to teach at a small undergraduate institution, or community college," she says. "But my love of UM and Ann Arbor meant I eventually accepted a tenure-track faculty position focusing on academic research, the first in a chemical engineering department." Susan's research focused on the use of multimedia to address diverse learning styles, and once again she supervised teams of undergradu ate students in developing educational software. Susan has a long list of accomplishments to her name. In addition to those already mentioned, in 1994 she started a student chapter of ASEE at UM, the third such chapter and one of the few long standing active chapters. The chapter has remained strong through the years, creating real change in the culture and appreciation of teaching through activities such as workshops and panels on engi neering education issues. "I strongly feel that theASEE student chapter changed the culture of teaching and graduate student training at the College of Engineer ing," she says. "Many of the activities we organized, which centered around preparing students for faculty positions, have been adopted by the college." These activities include academic job-search work shops and panel discussions on working toward tenure. To further prepare students, Susan regularly teaches a graduate course, "Teaching Engineering," that draws 50 graduate students and trains them for academic positions. The students learn to develop syllabi and course materials, practice presentation and teaching skills, and are introduced to different learning styles. They also learn to deal with student issues that may arise. "This has become 157

PAGE 16

an invaluable class for would-be future faculty members," says Sharon Glotzer, UM professor of chemical engineering. "It attracts students from around the College of Engineering, including postdoctoral students." Susan's concern for the well-being of graduate students and staff extends beyond the academic arena. Susan works to assist students with both academic disabilities and psychological issues. "Many times these issues manifest themselves during college years," she explains, "and bad grades are some of the early warning signs." To aid students, Susan says she tries to remove any stigmas around the topic of mental illness by sending out e-mail messages to students about depression, educating faculty about the issue, and encouraging those who need assistance to seek professional help. Her important work has not gone unnoticed: Recently she was asked to take part in a video titled "Depression on Col lege Campuses." canoeing down the Huron River," Susan says. For Ian, watching his mom forge her own path in engineer ing may have inspired him. He has aspirations to one day be a robotics engineer. Nicky thinks that might suit him as well if he doesn't make it as an NBA player first. Susan has not forgotten her Hispanic heritage, and has in stilled this pride in her sons. Ian and Nicky's friends have come to look forward to her alfajores Peruvian treats she prepares for any and all occasions. She also serves as faculty advisor to the Society of Hispanic Professional Engineers, participates in numerous sessions organized by the Minority Engineering Program Office (MEPO), and recently started "Ingenieros," an informal Spanish conversation group. "Dr. Montgomery has been an integral part of the diversity ef fort here in the College of Engineering," says MEPO program direc tor Derrick Scott. "She rarely turns down a re quest to participate in our initiatives to attract and retain underrepre sented minority engi neering students." In this sense, many faculty think of Su san as a pioneer. "She forges new ideas and utilizes new resourc es to make the cur riculum more effec tive," says chemical engineering lecturer Susan, right, with sons Ian, 12, and Nicky, 7, on a vacation to the Canadian Rockies. Susan is careful, however, not to get too busy. "My boys are the loves of my life and I want to share in every Barry Barkel. And it's not just students who benefit from her tireless work-so, too, do alumni. "Susan occupies the unique position of being the primary focal point of the department for both undergraduate students and alumni," says Barkel. "She is the face of the department for many people." BEYOND RESEARCH AND TEACHING Susan's family members likely think of her as a pioneer, too. She and ex-husband Sean Montgomery, whom she met when they were both undergraduates at UM, have two boys Ian, 12,andNicky, 7,-whomshehas taken on summer excursions to places such as the Grand Canyon and the Canadian Rockies, in keeping with her philosophy to take adventures now and not wait until retirement. This summer, they will embark on a trip to Peru with Susan's sister, Betsy, where they will visit family members as well as journey to Machu Picchu. Even when not traveling, life with two boys-and two cats, Smokey andAten-is understandably very active. Both boys are involved with karate and Ian plays on various team sports. "We also enjoy going for walks and bike rides, and going 158 aspect of their growing up," she says. After missing Ian's first soccer goal while out of town to be an ABET observer, Susan was determined that she would forgo traveling for business for a few years. "Ron Larson, the department chair, supported my efforts to create a balance between my academic position and my family life," she says. Despite this, Ian loves to point out that Susan missed his second soccer goal, which he scored while she was video taping Nicky and his friends playing on the sidelines. She did bend her own rules and take one business trip in 2002 to the ChE division Summer School for ChE faculty in Boulder, Colo. As Susan told the participants at the welcom ing session: "It doesn't matter what city you are in, if you are surrounded by your friends and colleagues of the ASEE ChE division, you'll always feel at home." It's hard not to feel at home around Susan, whether in Colo rado, Michigan, or Peru. Her excellence, determination, and love for people translates into every activity she performs. "In short, every department needs a Susan Montgomery on its faculty," says Ron Larson. "But they can't have ours." 0 Chemical Engineering Education

PAGE 17

JI i] =1 laboratory ) 11111-11111------=------AN AGITATION EXPERIMENT WITH MULTIPLE ASPECTS JORDAN L. SPENCER Department of Chemical Engineering Columbia University, New York 10027 A gitation and mixing are important in a wide variety of areas in both the traditional and modem process industries,lll and it is appropriate that chemical en gineering students see related experiments in the unit opera tions laboratory. This paper describes a teaching experiment involving both agitation and mixing that illustrates using a quite simple and relatively inexpensive apparatus-a number of aspects of this field. In particular the experiment involves not only simple and direct measurements, for example of torque and power as functions of stirring speed, but also more sophisticated data-acquisition and processing methods, for example indirect measurement involving the use of a model and a parameter estimation method. APPARATUS The apparatus is shown schematically in Figure 1 (next page). The major components of the apparatus are: o A torque table, consisting of a 12-inch-diameter alumi num circle mounted on a tapered roller bearing set in a 12-inch-square aluminum base plate. Even with a load of about 20 kg, the torque needed to set the upper plate in motion is less than 0.00706 Nm ( 1 oz inch). An arm attached to the upper plate bears a load cell (Omega Engineering, Model LCGC) connected to a panel meter (Omega Engineering). The data are acquired by a Lab View program at 0.2-second intervals. o A variable speed DC motor (Cole-Parmer) with speed controller and torque indication. The speed can be varied from 60 to 2400 RPM, and the maximum torque is 45 oz-in (0.318 Nm). o Two six-bladed turbines ( of diameter 14.4 and 7.5 cm and blade width 3.32 and 1.83 cm), and two three bladed propellers of diameter 14.4 and 7.5 cm. All are mounted on 3/8-inch glass-epoxy shafts inserted in a chuck on the motor shaft. o A polycarbonate tank of JD 29.2 cm and volume about 18 L. This tank is equipped with four removable stain less baffles of width 2.43 cm, mounted on an acrylic top plate. A stainless funnel is mounted near the top of the tank, to be used for adding a conductive tracer, for example NaCl or KCl at 30 g/L. A platinum electrode conductivity cell is mounted near the bottom of the tank 180 degrees from the funnel. The cell is connected to a conductivity meter (Amber Sciences) that sends a 0to Jordan L. Spencer is emeritus professor of chemical engineering at Columbia Univer sity. He received his B.S. in 1953 and his Ph.D. in 1961, both from the University of Pennsylvania and both in chemical engi neering. His research and teaching interests involve control and optimal control, and the development of chemical engineering teach ing experiments, including Web-operable experiments. Copyright ChE Division of ASEE 2006 Summer 2006 159

PAGE 18

DC MOTOR FUNNELi LAMP I t 0 0 BEAD IMPELLER 0 0 MOTOR CONTROLLER TANK WITH BAFFLES CONDUCTIVITY CELL c:::==p==o=j==::1~ TORQUE TABLE Figure 1. Schematic diagram of the apparatus, showing baffled tank on torque table with load cell, motor and impeller, conductivity probe, funnel for tracer addition, and optical sensor and lamp for bead detection. I-volt signal to the computer. A photo transistor ( OPF-703) mounted on the side of the tank 3 cm from the bottom is used to detect the approach of almost neutrally buoyant 0.9 cm toroidal plastic beads. The fluid near the sensor is illuminated by a 20 Watt desk lamp. o A second polycarbonate tank of inner diameter 17.5 cm, also equipped with four stainless baffles, and with volume about 4 L. All dimensions of the smaller tank/impeller combination are propor tional to the dimensions of the larger tank/impeller combination, with a factor of0.551. o A heat transfer probe constructed by inserting a 90 W resistance heater in the center of a 5-cm-diameter by 7.5-cm-long aluminum cylinder. A transistor-based temperature sensor (LM35-CAZ) is also mounted, off-center, in the cylinder. The probe could be suspended near the middle of the 20 L tank, about 5 cm from the wall. RESULTS Presented below are some typical results, with comments related to what the results illustrate for the student. Torque and Power The 18-liter tank was filled with water to a depth of 25 cm. With the baffles in place and the 7 .5-cm-diameter turbine mounted in the motor, the stirring speed was varied over a range of RPMs. The 0.30 ~--------------------0.25 -,.....---------------------, torque was measured by the load cell and averaged by a Lab View program. Typical results are shown in Figure 2, where the torque (N m) is plotted as a function of RPM. Also shown is a curve of the form Torque= k RPM 2 which fits the data almost perfectly, as expectedYl E 0.20 -t---------------------------1 2:.. 0.15 -t-----------------:a'1"""'--------t e0 0.10 +----------------,w,:::;_--------1 I0.05 +----------:--=.,...::z.... ________ ---1 0.00 .__ ... =:::::::::~------,-------.-------1 0 200 400 RPM 600 800 Figure 2. Torque as a function of turbine speed for water in bafjled 18-litertank. Also shown is a curve of the form Torque: k RPM2. 60 50 40 i 0 30 D.. 20 10 0 0 I\ V 5 I" r \ V \ _f\A 10 15 20 25 30 Frequency (dimensionless) Figure 3. Power spectrum of torque signal for 7.5-cm turbine in 18-liter bafjled tank at 800 RPM. 160 Since the torque signal was found to vary with time (with an amplitude about 10% of the average torque), especially at high RPM where the flow in the tank is quite turbulent, it was of interest to look at the power spectra of the data. A typical spectrum (Figure 3) shows that most of the power is at low frequencies, less than 0.15 cycles/sec. It is probable that the fluctuations reflect the presence of eddies or vortices generated by the impeller. Non-Newtonian and High Viscosity Fluids In order to examine the behavior of anon-Newto nian liquid, the 4-liter tank (with baffles) was filled to a depth of 15 cm with commercial ketchup. The 7.5-cm-diameter flat-bladed turbine was used to stir the ketchup. At 200 RPM the surface of the ketchup was stationary, clearly a non-Newtonian behavior. At 400 and 800 RPM the ketchup flowed smoothly at the surface. The data were well, but not perfectly, fitted by a curve of the form Torque= k RPM 2 Data were also collected under the same conditions using com syrup (Karo), a Newtonian fluid with a viscosity about 2,500 times that of water. For these runs the torque was a linear function of the RPM, as expected. Chemical Engineering Education

PAGE 19

Tracer Studies of Mixing Times-Modeling and Parameter Estimation Mixing times have been estimated[!, 2 i by visual ob servation following the addition of a dye or conductive tracer to the agitated liquid. This method, while satis factory in some cases, has the disadvantage that it is inherently subjective, and cannot be used with cloudy or opaque fluids. A more objective method is based on acquiring and processing conductivity data following the addition of a conductive tracer (e.g., NaCl solution) to the agitated fluid in the baffled or unbaffled 18-liter tank. The data are acquired by a Lab View program, typically 100 baseline points in 20 seconds, followed by 400 conductivity points in 40 seconds. The tracer data are saved from Excel as a space-delimited (.pm) file readable by a QuickBASIC parameter estima tion program. The model used to fit the data, shown in Figure 4, consists of six well-mixed tanks. Three, of equal volume, correspond to downward-moving fluid in the core of the tank. The other three tanks (all of the same volume) correspond to the fluid moving upward along the walls of the tank. Symmetry around the propeller shaft is assumed, so that only three shell tanks are shown. Salt solution injection is assumed to take place at the upper right, and a conductivity probe is located at the lower left. The model contains two undetermined parameters, denoted b 1 and b 2 with b 1 the fraction of the known tank volume in the three core tanks, and b 2 the volumetric flow rate (Lis) downward through the core tanks and upward through the shell tanks. The volume of a core tank is b 1 V /3, and the volume of a shell tank is (1b 1 )V/3. An objective mixing time is three times the tank volume, denoted V, divided by br Figure 5 shows the normalized conductivity vs. time data for salt solution injection into the 18 L tank, with agitation provided by a 7.5-cm-diameter downward-driving propeller rotating at 60 RPM. Also shown is the best-fit curve corresponding to the model discussed above. The best-fit parameter values were b 1 = 0.484 and b 2 = 1.368 Lis. The conductivity does not rise for about five seconds, corresponding to the time needed for the salt solution to move from the wall of the tank to the center, down to the bottom of the tank, and then over to the conductivity probe located opposite the injection point. Then the conductivity rises and drops rapidly as the bolus of salt solution passes the probe. After some further oscillations the conductivity reaches a constant value. The curve based on the six-pool model fits the data fairly well, but certainly not perfectly. This is because Summer 2006 z, s: :::s "C C: 0 u CORE TANKS B1 V/3 B1 V/3 SHELL TANKS 82 INJECTION 1+----(1-B1)V/3 POINT (1-B1)V/3 CONDUCTIVITY B1 V/3 ------1(1-B1)V/3 PROBE 82 SIX-POOL MODEL OF FLOW PATTERN Figure 4. Six-pool model of the flow pattern in the tank, used for analysis of conductive tracer-injection data. The core pools (tanks) represent the downward-flowing liquid in the center of the tank, the shell pools represent the upward flowing fluid near the tank wall. Parameter b 1 is the fraction of the tank volume Vin the core, and b 2 is the volumentric flow rate (Lis) through the array of well-mixed tanks. The state variables are the tracer concentrations in the six tanks. 2.5 2.0 DATA 1.5 1.0 0.5 0.0 0 10 20 30 40 Time (seconds) Figure 5. Conductivity data vs. time, and best-fit response from model, for tracer injection into 18-liter baffled tank at 60 RPM using 7.5 cm propeller. Parameters: b 1 =0.484, b 2 =1.368 Lis. 161

PAGE 20

At high Reynolds numbers the flow in the tank is, at least in some sense, strongly turbulent. This implies that particles suspended in the tank move in a chaotic and uncorrelated way, and thus move independently. the model is much too simple to correspond exactly to the highly turbulent, three-dimensional, and complex flow pattern actually existing in the baffled tank, even at low RPM. The parameter b 1 (the fraction of tank volume) was 0.484, and b 2 (the circulation rate) was 1.368 Lis. When the stirring speed was doubled to 120 RPM the conductivity rose much earlier, overshot less, and settled out about three times more rapidly-cor responding to more vigorous mixing. The value for the circulation rate parameterb 2 was 2.765, more than double that for the 60 RPM run. Similitude In theory, for tanks that are geometrically similar, at the same Reynolds number the power numbers should also be equal. As a test of this theory, the torque is measured at 600 RPM using water and a flat-bladed turbine in the 4-liter baffled tank, and the Reynolds number is calculated. Then the torque is measured in the geometrically similar 18-liter tank, at an RPM corresponding to the same Reynolds number. When the experiments were performed, the calculated and measured torques typically agreed within 20%. Heat Transfer Coefficient It is well documented[!, 2 i that heat transfer coeffi cients in agitated tanks depend strongly on the intensity of agitation. In order to demonstrate this, the heat transfer probe was immersed in water in the baffled 18 L tank. The power to the probe was turned on, the stirring speed was set, and the probe temperature was allowed to come to steady state, which occurred in a few minutes. The difference between the probe and water temperature was plotted as a function of RPM. As expected, the temperature difference was highest at 0 RPM, dropped monotonically as RPM increased, and approached a nonzero constant as the RPM ap162 proached high values. This reflects the fact that the resistance to heat transfer is the sum of a constant resistance due to the aluminum wall of the probe, and a film resistance at the probe surface that varies with the 2/3 power of RPM. From the high RPM asymptote the first resistance can be calculated, and from a second point the heat transfer coefficient can be found as a function of RPM. In general the results were consistent with Eq. (1): k = a No67 = a N213 (1) where k is the heat transfer coefficient (Wm 2 / 0 C), a is a constant, and N is the stirring speed (s 1 ). 0.50 cu 0.40 ui ... 0 0.30 1/j C: a, (/) 0.20 cu 0 '[ 0.10 0 0.00 0 5 10 15 20 Time (seconds) Figure 6. Optical sensor voltage for beads in 18-liter tank at 400 RPM. Each drop in voltage corresponds to the entry of a bead into the field of view of the phototransistor. Only drops below a selected voltage are counted as events. 6.0 z: "iii 5.0 C: cu 1J oi 4.0 C: 2 3.0 C: ...J 2.0 0.0 1.0 2.0 3.0 4.0 Interval Length (seconds) Figure 7. Semilog plot of distribution of bead arrival intervals. The straight line corresponds to a Poisson distribution. Chemical Engineering Education

PAGE 21

Note that all real-world measurements involve both signal and noise. In most cases the information is contained in the signal and we attempt to minimize the noise. But in some cases the noise also contains useful information. The results .. illustrate this. Particle Dynamics At high Reynolds numbers the flow in the tank is, at least in some sense, strongly turbulent. This implies that particles suspended in the tank move in a chaotic and uncorrelated way, and thus move independently. In order to test this prediction about 200 plastic toroidal beads of approximate diameter 0.9 cm were added to the tank. The 7 .5-cm turbine was run at 400 and 800 RPM. A typical sensor signal is shown in Figure 6, corresponding to a sample rate of 100 per second. Each drop in voltage represents the entries of one or more beads into the illuminated region. The program that acquired the data also used simple logic to identify close-bead approaches, to calculate the time interval between entry, and to construct an interval distribution function. If the beads in fact move independently, this is expected to be a Poisson distribution. Typical results, shown in Figure 7 in a semilog plot, corre spond reasonably well to a straight line and thus to a Poisson distribution. The slope of the line is related to the average frequency of bead events, but also depends on the efficiency of detection of bead approaches, which is not easily known. The lower values at low time intervals probably correspond to the fact that the data acquisition program is not able to differentiate between two or more bead entries that occur at almost the same time. Note that all real-world measurements involve both signal 0 20 40 Time (seconds) Summer 2006 and noise. In most cases the information is contained in the signal and we attempt to minimize the noise. But in some cases the noise also contains useful information. The results above illustrate this. Salt Crystal Dissolution When a crystal of a soluble salt, for example NaCl, sits in water that is not moving, it dissolves relatively slowly. But the opposite holds for a crystal suspended in strongly turbulent water. Approximately one gram of a coarse (about 150 crystals per gram, the crystals being of variable size) kitchen salt was added to water in the baffled 18 L tank. At 0 RPM full dis solution required more than 60 minutes. The signal from the conductivity probe recorded by a Lab View program is shown in Figure 8. Using the 7.5-cm-diameter turbine, dissolution was almost twice as fast at 800 RPM as at 400 RPM. These results demonstrate that the apparatus described above can be used in a wide variety of studies of the effect of stirring speed, impeller design, and other parameters on the rate of dissolution of ( or extraction from) solids, as long as the solids release a conductive tracer. CONCLUSIONS The agitation and mixing experiment described above is based on a relatively simple and inexpensive apparatus. But it illustrates a number of aspects of the subject, including the 60 80 Figure 8. Conductivity as a function of time following addition of 1 gram of NaCl crystals to the agitated 18-liter baffled tank at 400 RPM. 163

PAGE 22

dependence of torque on stirrer speed, impeller design, baffle design, and nature of the fluid involved. The principle of similitude can be tested. The experiment also illustrates very general and more sophisticated concepts, including the use of modem data acquisition software and nonlinear regression methods to estimate the parameters of a model of the flow pattern in an agitated vessel. The apparatus is well adapted to studying the rate of dissolution of salt crystals, and the effect of agitation intensity on heat transfer from a solid surface. 164 Finally, students are able to acquire and process essentially stochastic data to obtain some information on the turbulent flow in the vessel. REFERENCES 1. McCabe, WL., J.C. Smith, andP Harriot, Unit Operations of Chemical Engineering, 4th Ed., McGraw-Hill, New York (1985) 2. Uhl, V.W, and J.B. Gray, Mixing, Theory, and Practice, Vol. 1, Aca demic Press, New York (1966) 0 Chemical Engineering Education

PAGE 23

5 =i classroom ) 11111-1111111---------USING A COMMERCIAL SIMULATOR TO TEACH SORPTION SEPARATIONS PHILLIP C. WANKAT Purdue University West Lafayette, IN 47907-2100 S ince modern practice of chemical engineering uses specialized process simulators extensively (e.g.,Aspen Plus, CHEMCAD, HYSIM, and PROSIM), chemical engineering departments need to prepare students to use these tools. For example, distillation columns are designed almost exclusively using process simulators, and if the equilibrium data is deemed reliable, the column will be constructed with out any laboratory or pilot data. Most chemical engineering departments now use one of the steady-state process simula tors in separations and/or design courses_[i,zi The steady-state simulators do not include adsorption, chromatography, and ion exchange (collectively, sorption), which are normally operated as unsteady-state processes. Formerly, sorption systems were designed by a combination of data and rules of thumb. Recently, it has become more common to use a more fundamental design procedure based on solution of the partial differential equations governing the heat and mass transfer in the column and the algebraic equations for equilibrium and pressure drop. In industry, the detailed simulations are always accompanied by laboratory and often pilot plant data. Chemical engineering graduates who understand the fundamentals of sorption processes and are familiar with sorption simulators will have a competitive advantage. This paper discusses the use of the commercially available Aspen Chromatography simulator to teach sorption separations. The Summer 2006 course outline, grading procedure, assignments, computer laboratory operation, and testing procedure are delineated. Student survey results and the author's opinion of the effec tiveness of teaching with this simulator are presented. THE COURSE ChE 558, "Rate-Controlled Separation Processes," is a three-credit, dual-level elective course that has been taught off and on for almost 30 years_[ 3 l The topics covered always include sorption separations, and depending upon the profes sor, might also include crystallization, electrophoresis, or mem brane separations. I have used Rate Controlled Separations [ 4 l although this book is currently difficult to obtain. This course has always been taught in a lecture style with homework and often a course project. Phil Wankat is a Distinguished Professor of chemical engineering at Purdue University who earned his degrees from Purdue and Princeton. His technical research is in separation processes, and he recently finished his textbook, Separation Process Engineering 2nd Edition of Equilibrium Staged Separations, Prentice-Hall, 2006. He is also co-author of the book Teaching Engineering, available free at . Copyright ChE Division of ASEE 2006 165

PAGE 24

Three considerations led me to change the teaching method. First, since I believe that the sorption separation processes will eventually be designed almost entirely using simula tors, proper preparation of graduates will require teaching with simulators. Second, the understanding of an average ChE 558 student was too low. Since I had observed student improvement in a distillation course when a simulation lab was incorporated,[ 2 J I expected an increase in understand ing if a similar change was made in ChE 558. Third, I had proposed in the educational part of two NSF proposals to teach ChE 558 with a simulator, and now I had to deliver on these promises. In spring 2005 I changed ChE 558 to focus entirely on sorp tion separations. The nominal schedule had a one-and-a-half hour lecture on Tuesdays and a one-and-a-half-hour computer laboratory using the Aspen Chromatography simulator on Thursdays (see Table 1). This schedulehadfewerlectures on sorption separations than in previous years, but tests covered the same amount of material on these topics. The total amount of material in the course was reduced by removing the mem brane separation material, which is now often included in the required undergraduate course on separations. The course was taken by four undergraduates and three graduate students. Only one of the students had previous experience with an unsteady-state simulator, but all had previous experience with Aspen Plus, which has a somewhat similar graphical user interface to Aspen Chromatography. The grading scheme used a straight scale (85-100 = A, 75-85 = B, 60-75 = C, 50-60 = D) as guaranteed grades, but I reserved the right to use lower cut-offs if that was appropri ate. The two regular tests were each 25% of the grade, the lab exam was 20%, lab attendance 9%, lab assignments 6%, homework 5%, and the group course project was 10%. Stu dents were encouraged to work together on lab assignments TABLE 1 Schedule ChE 558, Spring 2005. Readings are from Reference 4. Date Class Room Subject Reading T, Jan 11 1 110 Intro. Adsorption & Chromatography 207-228 Th, Jan 13 2 111 Lecture -Adsorption: thermo/phys. prop./flow; start solute movement 228-251 T, Jan 18 3 110 Lecture Solute movement 239-251, 296-305 Th, Jan 20 4 111 Lab 1 Intro to Aspen Chromatography Skim 268-274 T, Jan 25 5 110 Solute movement/thermal effects-focusing 251-268 Th, Jan 27 6 111 Lab 2Chromatography/adsorption basics 288-296 T, Feb 1 7 110 Heat & Mass Transfer, local equilib r ium solut i on 268-277, 296-305 Th, Feb3 8 111 Lab 3 Convergence T, Feb 8 9 110 Chromatography Linear solutions 305-316 Th, Feb 10 10 111 Lab 4 Chromatography 316-321, 336-347 T, Feb 15 11 110 Chromatography Linear solutions 316-331, 334 Th, Feb 17 12 111 Lecture Constant pattern and scaling 365-393 T, Feb 22 13 110 Plateaus & Nonlinear behavior, start MB and SMB 393-400, 521-533 Th, Feb 24 14 111 Lab 5 Thermal effects 405-412 T, Mar 1 15 110 Test 1 Th, Mar3 16 111 Lab 6 Flow reversal systems 405-418 T, Mar8 17 110 Moving Beds and SMB; review test 499-537 Th, Mar 10 18 111 Lab 7 TMB and SMB 521-533 SPRING BREAK T, Mar22 19 110 Ion Exchange 452-484 Th, Mar24 20 111 Lab 8 Ion exchange 475-481 T, Mar29 21 110 Ion exchange 475-491 Th, Mar31 22 111 Lab 9 LAB EXAM T, Apr 5 23 110 PSA/Gas separation 400-418, 421-438 Th, Apr7 24 111 Lab 10 Lab demo -ADSIM PSA Aspen Chromato. obtaining data from article Read article T, Apr 12 25 110 PSA/Gas separation 421-431 Th, Apr 14 26 111 Lab 11 Project T, Apr 19 27 110 Lab 12 Projects Th, Apr 21 28 111 Test 2 T, Apr 26 29 110 Work on projects Th, Apr 28 30 111 Lab 13 Project reports and demos 166 Chemical Engineering Education

PAGE 25

View Tools Flowsheet Run Window Cl liil I ii [:?l I J!, rn I "" C, I t(/ Dynamic !El l ljP Process Flow sheet Window A / A I> 0 '1 ..:1 1 f'll r;m,sN,wRom,n isf io E) B I u 81 I and homework. The complete course syllabus is available from the author at . Homework assignments were problems from the textbook plus one straightforward simulation. The textbook problems were similar to the test problems; of course, new problems were written for the tests. Since the students were all able to come to class early, they were given two hours for each test. Unfortunately, due to a mistake in solving an ion exchange problem on the second test, this problem, although solvable, was about an order of magnitude too difficult. I adjusted scores based on the performance of the second-best student in the class (the best student appeared to be an outlier whose performance was not representative of the class). The students appeared to be satisfied with the fairness (or generosity) of this procedure. ASPEN CHROMATOGRAPHY COMPUTER LABORATORY Aspen Chromatography is an algebraic-differential equa tion-solving program with a user interface for the solution of liquid adsorption and chromatography problems (see Figure 1). This simulator is very powerful and a trained user can often solve in a few hours a problem that used to take months. Aspen Chromatography uses the method of lines to solve the partial differential equations. The user can select both the differencing method to be used and the integration method Summer 2006 Figure 1. Screen shot of Aspen Chromatography Interface with flow sheet for a simple chromatography system. to solve the resulting ordinary differential equations. Aspen Chromatography licenses are expensive for companies, but are reasonably priced for universities and can be bundled with other Aspen Technology programs. It cost $400 to add an As pen Chromatography license for 60 users to Purdue's Aspen Technology order for University Lifecycle Package Bundle #1 (60users) that cost $2,000. The current Version 12 is quite stable and reasonably user friendly, but not as user friendly as Aspen Plus. My experience withAspenPlus is that 98-99% of the difficulties students have are due to operator error. With Aspen Chromatography about 80% of the students' difficul ties are caused by operator error. As expected, the numerical integration routines, which use the method of lines to solve the partial differential equations, have difficulty converging when the profiles are steep and the isotherms are nonlinear. In general, the resources and expertise that have been developed for teaching with steady-state simulators[!, 2 5 l are not avail able for sorption separations. More troubleshooting and more computer assistance will be needed. Since much of my current research involves simulation of chromatography and simulated moving-bed systems with Aspen Chromatography, I am familiar with this simulator and my graduate students are very familiar with it. The graduate students and post-doc supported by the NSF grants were en listed to help with the computer laboratory. With their aid, I developed 10 laboratory assignments including a laboratory 167

PAGE 26

test. Each of the first eight laboratories showed how to build a flow sheet for a new aspect of Aspen Chromatography in a cookbook fashion, and then had the students solve simula tion and design problems. Excerpts from the first laboratory assignment are presented in Table 2. All lab assignments are available from the author at . As the semester progressed the amount of detail in the in structions was decreased. Most of the students stayed in the lab after the nominal closing time to finish the take-home as signments that accompanied the labs. The material covered in TABLE2 Excerpts from First Lab Assignment A complete set of instructions for all labs is available from . The goal of this lab is to get you started in Aspen Chromatography. It consists of a cookbook on running Aspen Chromatography and some helpful hints. We will also simulate a real separation. Keep this lab assignment. You will want to refer back to it. 1. Log in to the computer. Go to Start, Programs, ChE Software, AspenTech, Aspen Engineering Suite, Aspen Chromatography 12.1, Aspen Chromatography. This opens a window if you are at a station that allows you to access the hard drive. Otherwise, you will get a message that essentially says, 'The working folder is unavailable." In this case, change working file to your N drive. Click on OK, and window should open. If not, run in circles, scream and shout, and ask for help. 2. We will first develop a simple chromatography (or adsorption) column system. To do this, go to the menu bar and on the left side, File. Click on File and go to Templates, and in that window click on "Blank trace liquid batch flowsheet," and click on Copy. It will ask for a file name. Use something like "columnl." This will be saved in your working file. NOTE: In all file names and names for compo nents, columns, steams, and so forth there must be NO spaces. 3. In the "Exploring simulation" box (LHS), click on "component list." Then in box below (Contents) double click on Default. This lists A and B. Change these names to the names of the components to be separated (fructose and dextran T6). First, click "Remove all" button. Then in window below type in first component name (e.g., fructose) and click on "add" button. Do the same for all other components. Then click OK. 4. Now draw the column. Click on the+ to the left of "Chromatog raphy" in the "Exploring Simulation"box to open other possibilities. Click on the word "chromatography. "This should give "Contents of Chromatography" in box below. Double click on the model you want to use (Reversible since it is most up-to-date). Click and drag the specific model you want: in this case "chrom_r_column," and move to the center of the Process Flowsheet Window. This gives a column labeled B 1. Left click on B 1, then right click to open a menu. Click on Rename. Call the block something like "column." Click on OK. 19. If you have time, do this next step. If not, save your file (remem ber the file name), exit Aspen and do this step outside of class. The two peaks are not completely separated. There are a number of ways they can be separated more completely. Double the value of L, to L = 50 cm. Click on Rewind, change Lin the column dimen sions table, and then rerun the one-minute pulse input. When you run pure solvent, a pause time greater than 10 minutes is needed since doubling column length will double time for material to exit. Do this run and look at the result. Separation is better, but still not complete. Print your plot and label it. This plot will be handed in with the lab assignment. Save your file (remember the file name) and exit Aspen. 168 TABLE3 Handout on What to Expect in Lab Exam The exam will be open book and open notes. You may not open or use any of your old Aspen Chromatography files. Part A. (50 points) Generic Problem. This is a demonstration that you can do a basic Aspen Chromatography simulation. Open up Aspen chromatography and use a "Blank trace liquid flowsheet tem plate." Set up a chromatographic column with one feed, a column, and a product. Use specified models for the column, feed, product, and connecting streams. Set up the system to process compounds that will be specified in the test. Have Aspen do discretization with xyz procedure with NN nodes (these will be specified in the test). Use a model with convection plus a specified form of dispersion, constant pressure, and velocity. If needed, the dispersion coefficients will be supplied. Use a linear, lumped parameter model with a specified driving force and constant mass transfer coefficients (they will be given). The isotherms will be given and the units for q and c will be specified. Operation is isothermal. The column length and diameter will be given. The adsorbent has the following properties: Ee =0.aa, Ep = 0.bb, KD = 1.0, QS = cc kg/m 3 The following feed values will be specified: flow rate, pressure, and all component concentrations. Use a specified integrator with a specified fixed or variable time. Use default values of the tolerances. Develop a graph of the product concentrations ( on the same scale) versus time. 1. Run a breakthrough curve for zz minutes. Print, label, and turn in your plot. Use the history to accurately determine the center of the breakthrough curve and the -rz for one of the components where ~TZ is measured from 0. 05 times the feed concentration to 0. 95 times the feed concentration. These calculations should be shown on the plot. 2. Input add-minute feed pulse and develop with pure solvent for a total time of zz minutes. Print your plot, label, and turn in. There should not be any convergence problems in Part A. PartB. (50 points) The second part of the lab test will be a design problem for one of the other processes that we have studied (e.g., flow reversal, adiabatic operation, SMB, TMB, ion exchange). TABLE4 Homework Assignment 4 1. Use the Lapidus and Amundson solution with E,ffotive to predict the behavior of fructose in a column packed with silica gel. The feed is 50 g/liter, the feed pulse lasts for eight minutes, and then it is eluted with water. The flow rate is 20 ml/min. The other values are: Value Units Description L = 200.0 cm Length of adsorbent layer in column Dcol =2.0 cm Ee = 0.4 m 3 void/m 3 bed Ep = 0.0 m 3 void/m 3 bed dp = 0.01 cm ED = 0.15 cm 2 /min Internal diameter of column Inter-particle voidage Intra-particle voidage particle diameter (needed to find E,ff,ctiv.) Constant Dispersion Coefficient Lumped parameter with concentration driving force. k a = 5.52 1/min Constant mass transfer m,c p coefficient Isotherm is linear K' = 0.69 dimensionless Isotherm parameter (q and c both in g fructose/liter) 2. Solve problem 1 using Aspen Chromatography. 3. Compare your solutions for problems 1 and 2 at the peak center time predicted by the local equilibrium solution, peak center time minus four minutes and peak center time plus four minutes. Chemical Engineering Education

PAGE 27

the laboratory was cumulative, and by the end of the semester the students were able to simulate rather difficult problems without detailed instructions. Part A of the lab test was a demonstration by the students that they had learned how to use Aspen Chromatography for simple simulations. Two weeks before the test the students were given the generic form of part A (Table 3). They were encouraged to supply data and parameter values to generate their own form of the test and then practice solving it. Part B, a design problem, proved to be more difficult. The lab test divided the class into groups of 3, 3, and 1. I used this unusual procedure because one of the graduate students is doing his thesis research with me and during the course of the semester he had much more practice with Aspen Chromatography than the rest of the class. He agreed to be a group of one, and the class accepted my rationale when the groupings were pre sented. The other two groups were made as equal as possible based on grades in the course. EXAMPLE PROBLEM was given during a normal lab period that was extended to Students solved a number of chromatography and adtwo hours. Since there were only seven students in the class sorption problems during the semester. The real strength of and I knew them all well, no special precautions beyond numerical analysis is it can solve problems with complicated proctoring the exam were taken to ensure honesty. (When nonlinear isotherms that cannot be solved analytically. To I gave an Aspen Plus lab test in a core .------------------, avoid the "black box" effect, benchmarkjunior class with 95 students, I wrote Aspen Chromatography ing of numerical solutions with analytical a different test for each of the five lab solutions was done for linear problems sections and disabled both e-mail and is an algebraic-differential where analytical solutions exist. One access to student files.) equation-solving program convenient analytical solution is the During the 10th lab, students first h ,I'. Lapidus and Amundson solution[ 4 l with wit a user interJace watched a computer demonstration of an effective dispersion coefficient that the use of ADSIM for pressure swing for the solution includes the effects of dispersion and adsorption. Gas separations can involve of liquid mass transfer_[7J Homework assignment large changes in flow rates which are not 4 ( see Table 4) illustrates benchmarking modeled by Aspen Chromatography. adsorption and of analytical solutions. This assignment Then the students did a simulation with requires students to solve a simple, sinchromatography Aspen Chromatography that required gle-component chromatography problem them to determine the parameters needproblems. This simulator is with a large pulse of feed by the Lapidus ed for the simulation from a literature and Amundson method and numerically very powerful and paper. In the earlier labs the students had with Aspen Chromatography. been given all the necessary parameters a trained user can often since that makes troubleshooting of stu dent difficulties much easier. Students were told that the purpose of learning how to extract parameters from the literature was to prepare them for the solve in a few hours a problem that The effective dispersion coefficient that lumps all dispersive effects into axial dispersion and assumes negligible mass transfer resistance was estimated to be 8.062 cm 2 /min. This is much greater than the axial dispersion coefficient value 0.15 cm 2 /min because mass-transfer reused to take course project. months. The course project was to develop a new Aspen Chromatography problem and solution suitable for one lab period. This is a form of Felder's generic quiz.[ 6 l Students were required to use equilibrium and mass transfer data from the literature and/or the Internet, not from the text book or from Aspen Chromatography demonstrations. They were told that projects that considered operational methods not taught in the lab or that combined different operational methods would be most impressive. Student groups presented an oral report, including a computer demonstration, and turned in a written report. As a treat for the students, I ordered pizza to be delivered after the oral reports were presented. The student projects-nonisothermal ion exchange, ion-exchange with flow reversal, and SMB separation-were quite well thought out. Since seven students do not divide evenly into groups, I Summer 2006 sistance controls dispersion. The Lapidus and Amundson solution requires the use of superposition as a step up followed by a step down eight minutes later. The same problem was solved numerically with Aspen Chromatography using two of the higher-order differencing schemes, Buds (Biased Upwind Differencing Scheme, a 4th order method) and QDS (Quadratic Differencing Scheme), and the default UDSI (Upwind Differencing Scheme 1) with 50 nodes. The solutions all used the Gear method with a fixed time step for integration. The QDS solution was done first with the actual value of the mass transfer coefficient and axial dispersion coefficient, and then with a very high mass-transfer coefficient (essentially no resistance) and the effective dispersion coefficient. A screenshot of the Aspen Chromatography solution using 169

PAGE 28

Buds with 200 nodes is shown in Figure 2 and a screenshot of the solution using UDS 1 with 50 nodes is in Figure 3. The Lapidus and Amundson solution and the higher-order nu merical solutions were bell-shaped curves and looked almost identical. UDSI with 50 nodes also produced a bell-shaped curve, but it is much more spread out and has a lower peak concentration than the other curves because of significant numerical dispersion. The curves are different enough that students can easily see the differences by comparing Figures 2 and 3. Thus, the use of UDS 1 with 50 nodes is numerically inappropriate for this problem. .. ...,, The students' responses to the survey (Table 6) show that previous knowledge of different computer applications varied from no knowledge to comfortable. General comfort levels with computers were high. With the exception of the speed of the Distributed Academic Computing System (DACS), which allowed remote access to Aspen Chromatography, laboratory operation was rated as about right. The students thought that both the computer labs and the lectures helped them learn sorption separations and that combining lecture and lab was an appropriate way to teach this material. Most of the comments are positive and reinforce the advantage of 10 15 Fructose_Concentralion 20 nmeMinutes 30 !l!.I 2!l 35 40 _J j Since the Lapidus and Amundson and the higher-order numerical analysis curves are so similar, differences can only be ascertained by looking at exact values of concentrations and times (Table 5). The concentrations predicted by the Lapidus and Amundson solution are: t = 25.575, c = 25.0 g/liter; t = 29.575, c = 48.54 g/liter (peak maxi mum); and t = 33.575, c = 24.975 g/liter. The Lapidus and Amundson solution has its peak center at exactly the time predicted by the local equilibrium solution (29.575 minutes). The peak concentration, peak time, and the predicted times for concentrations of 25.0 and 24.975 g/liter are given in Table 5 for the five different solutions. Since the two QDS solutions are quite close to each other, the use of an effective dispersion coefficient is valid for this linear system. All of the rea sonable solutions ( excluding UDS 1) are quite close, with a small shift in times. Although the Buds solution with 200 nodes is the best fit to the analytical solution, in practical terms it doesn't matter which is used. One of the lessons students learn from this and other benchmarking exercises is that they must pay close attention to numerical convergence. Figure 2. Screenshot of Aspen Chromatography solution for problem 2 in Table 4 using Buds with 200 nodes. RESULTS A survey on the computer laboratory was developed, and a research exemption was obtained from the Purdue Institutional Re view Board for Human Subjects Research. The students all responded to the survey (Table 6) on the last day of class. To avoid biasing any of the responses, the survey was administered by the undergraduate secre tary; I was not in the room while the students filled out the survey, and the process was completed before the students knew there would be a pizza delivery. 170 F Iie View Tools Run Window HetJ 10 Fructose_Concentr.ation 20 a 30 ...... J jQynami,: s oo14S.0"""'-t"'5 45 ----'jl .,a Figure 3. Screenshot of Aspen Chromatography solution for problem 2 in Table 4 using UDS1 with 50 nodes (an inappropriate choice). Chemical Engineering Education

PAGE 29

Peak time Peak cone. TABLES Comparison of Solutions for Problems in Table 4 L&A Soln. 29.575 48.54 Buds 200 nodes 29.6 48.63 Aspen Chromatography Solutions QDS QDS (100 nodes) 100 nodes [E=E di' MTC=l00.000] 29.5 29.5 47.84 47.87 UDSl 100 nodes 28.7 34.30 having a computer lab. At the same time they filled out the survey, the students responded to the standard course evalu ation questionnaire required in all ChE courses. Course evaluation questions that ask for global ratings correlate positively with student learning. [SJ These core ques tions were, 1. "Overall, I would rate this course as:" and 2. "Overall, I would rate this instructor as:" The choices were: Time. min 25.575 25.47 @ upward curve. c=25.0 Time. min 33.575 33.44 @ downward curve. c= 24.975 25.43 25.38 33.43 33.38 24.90 32.46 Excellent=5, Good=4, Fair=3, Poor=2, and Very Poor=l. The scores obtained for these questions-4.1 and 4.6, respec tively-collaborate the impression that TABLE6 ChE 558 Computer Laboratory Survey (The average values and comments in italics are based on student responses.) I. Computer experience before taking ChE 558. Rate your experience with the following applications (name package used where asked) using the following scale: 1 = Never used it before 558. 2 = Knew a little about it before 558. 3 = Used it some before 558. 4 = Was comfortable with it before 558. AYg_ Spreadsheets ................................................................ 1 .............. 2 ............... 3 .............. 4 ............ .4.0 Internet ......................................................................... 1 .............. 2 ............... 3 .............. 4 ............ .4.0 DACS .......................................................................... 1 .............. 2 ............... 3 .............. 4 ............. 2.3 Aspen Chromatography ................................................ 1 .............. 2 ............... 3 ............. .4 ............. 1. 1 Aspen Plus ................................................................... 1 .............. 2 ............... 3 ............. .4 ............ .3.1 Other steady-state simulator ......................................... 1 .............. 2 ............... 3 ............. .4 ............. 1.6 ........... Pkg? _P_r_o~l-l_l _____ {Mathlab. Mathcad. Maple.Mathematica} .................. 1 .............. 2 ............... 3 ............. .4 ............ .3.1 ........... Pkg? Mathematica4 Matlab5 DEQ-algebraic eqn solver ............................................ 1 .............. 2 ............... 3 ............. .4 ............. 1.0 ........... Pkg? ________ Data Base ..................................................................... 1 .............. 2 ............... 3 ............. .4 ............. 2.0 ........... Pkg? ~A~c-c_e_s,_v 2 _____ Statistical package ........................................................ 1 .............. 2 ............... 3 ............. .4 ............. 2.9 ........... Pkg? IMP 2 Crystal Ball I Programming language(s) ............................................ 1 .............. 2 ............... 3 ............. .4 ............. 2.1 ........... Pkg? FORTRAN I C++ IC 2 Other ............................................................................ 1 .............. 2 ............... 3 .............. 4 ............. 1.0 ........... Pkg? _______ II. Computer comfort level. Rate your comfort level with the computer: 1 = Uncomfortable 2 = Neither comfortable nor uncomfortable 3 = Reasonably comfortable General comfort level using computer before class ...... 1 ............. 2 .............. 3 ............. .4 General comfort level using computer now ................. 1 ............. 2 .............. 3 ............. .4 Comfort level using Aspen Chromatography now ........ 1 ............. 2 .............. 3 ............. .4 Comments: III. Computer Laboratory Operation. Please circle the appropriate response. AYg,. ............ .3.7 ............ .3.7 ............ .3.4 4 = Very comfortable Avg. The computer speed with direct installation (without DACS) was: ................ 1. slow ........... 2. about right ........ 3. fast ....................... 2.1 Computer speed using DACS was: ................................................................. 1. slow ........... 2. about right ........ 3. fast ....................... 1.4 The laboratory assignments were: ................................................................... 1. too long ...... 2. about right ........ 3. too short ................ 2.0 Computer lab should be scheduled for: ........................................................... 1. less time ..... 2. same time ......... 3. longer time ............ 2.0 The assistance available during lab from the graduate student and the professor was: Comments: On one survey the term "graduate stwlent" was underlined. IV. Learning. Please answer these questions with the following scale: 1. inadequate ... 2. adequate ............. 3. very good ................ 2.4 1 = Strongly disagree 2 = (Between 1 & 3) 3 = Neither agree nor disagree 4 = (Between 3 & 5) 5 = Strongly agree. Avg. The computer labs helped me learn adsorption and chromatography ...................................................... 1 ..... 2 ... 3 .. .4 ..... 5 .............. .4. 7 The lectures and homework on the theory helped me learn adsorption and chromatography ................. 1 ..... 2 ... 3 .. .4 ..... 5 ...... ....... .4. 9 The format of ChE 558 (combining lecture and computer laboratory) is appropriate for this subject. ... 1 ..... 2 ... 3 .. .4 ..... 5 .............. .4.6 Comments: "Because of the complexity of solving chromatngraphic problems, being able In see what actually happens in a column was quite nice." "Two-day Tues./Thurs. schedule worked great!" "Without lab a lot of material would be lost" "More classroom time would be helpfal In reinforce some material" V. Suggestions for improving 558 computer lab: "Run the simulations before the stwlents run them." "More labs with more lab time, cover a little more material." Summer 2006 171

PAGE 30

the students thought this was a good course. I believe the students learned sorption separations in more depth in spring 2005 than in previous years. This seemed to be true across the spectrum of student abilities (good students learned more than good students previously, average students learned more than average students previously, and struggling students learned more than struggling students previously). Since in previous years the course also covered membrane separations, the breadth of coverage was less in 2005; how ever, the students learned sorption operations better despite less lecture time spent on this topic. Obviously, the 2005 students are also prepared to use the simulator. DISCUSSION AND CONCLUSIONS The students generally liked the format of lab and lecture and thought it helped them learn; however, these students all volunteered to take this elective knowing there would be a computer lab. Students who feel uncomfortable using the computer probably took other electives. During the semester a faulty installation of Windows caused difficulties running Aspen Chromatography in the computer laboratory. For several weeks the students needed to log into DACS, which was slower than the direct installation. Once the problem was identified and Windows was reinstalled, we had no difficulties with the direct installation of the software. The comment in Table 6, "Run the simulations before the students run them," probably referred to this difficulty running some of the labs on DACS. Lab 6, with flow reversal, ran without problems when I tested it using the direct installation of Aspen Chromatography in my office, but would not run on DACS. The student group that later did its course project with flow reversal had no difficulty following the original lab instruc tions and obtaining solutions with a direct connection. It is important to have reliable computer support before scheduling use of any simulator. 172 If there are transferable skills in learning how to use simula tors, students who become skilled with, for example Aspen Plus, will learn to use another simulator faster. This appeared to be true for Aspen Chromatography. Thus, even if they never use simulators taught in the curriculum, the experience of learning to use these simulators will probably help graduates efficiently learn to use simulators on the job. ACKNOWLEDGMENT The assistance of Nadia Abunasser, Jin-Seok Hur, Weihua Jin, and Dr. Jeung-Kun Kim is gratefully acknowledged. The computer personnel in ChE-George Bailey and Eric Pratt-were extremely helpful in making the computer lab run smoothly. This project was partially supported by NSF grants CTS-02112008 and CTS-0327089. This paper was presented orally at the AIChE Annual Meeting, Cincinnati, November 2005. REFERENCES 1. Rockstraw, D.K., "Aspen Plus in the ChE Curriculum," Chem. Eng. Ed., 39(1), 68 (2005) 2. Wankat, PC., "Integrating the Use of Commercial Simulators into Lecture Courses," J. Eng. Ed., 91(1), 19 (2002) 3. Wankat, PC., "An Elective Course in Separation Processes," Chem. Eng. Ed., 15(4), 208 (1981) 4. Wankat, PC., Rate-Controlled Separations, Kluwer, Amsterdam (1990) (There were earlier printings by Elsevier and Blackie. Kluwer is now part of Springer.) 5. Seider, WD., J.D. Seader, and D.R. Lewin, Process Design Principles, Wiley, New York (1999) 6. Felder, R.M., "The Generic Quiz: A Device to Stimulate Creativity and Higher-Level Thinking Skills," Chem. Eng. Ed., 19(4), 176 & 213 (1985) 7. Dunnebier, G., I. Weirich, and K.U. Klatt, "Computationally Efficient Dynamic Modeling and Simulation of Simulated Moving Bed Chro matographic Processes with Linear Isotherms," Chem. Eng. Sci., 53, 2537 (1998) 8. Centra, J.A., Reflective Faculty Evaluation, Jossey-Bass, San Francisco, (1993) 0 Chemical Engineering Education

PAGE 31

Random Thoughts ... HOW TO TEACH (ALMOST) ANYBODY (ALMOST) ANYTHING RICHARD M. FELDER AND REBECCA BRENT North Carolina State University Raleigh, NC 27695 I t seems it's no longer enough for you to teach about the Navier-Stokes equations and potential flow past sub merged objects. The ABET coordinator says that students in the fluids course have to be taught oral communications too, and the department head got inspired at some workshop and now wants to teach critical thinking in every course, includ ing fluids. You argued at the faculty meeting that it's all you can do to get through fluids in the fluids course but got little sympathy, and it looks like there's no way out of it. You probably have some questions at this point, like, ( a) Exactly what are those skills I'm supposed to teach? (b) Can they be taught (as opposed to you either have them or you don't)? and ( c) How? For an answer to ( a), we invite you to check out an article we wrote on learning objectives, teach ing strategies, and assessment methods that address ABET Outcomes 3a-3k. 1 The answer to (b) is, yes. This column suggests some answers to ( c )-how do you enable students to develop and improve a targeted skill, whether ABET-re lated or not? While we don't guarantee that the techniques we'll recommend will always work for all students, we're confident the results will be good enough to satisfy ABET and your department head, and-as long as your expectations are realistic-you. 1. R.M. Felder and R. Brent, "Designing and Teaching Courses to Sat isfy the ABET Engineering Criteria," J. Engr. Education, 92( 1 ), 7-25 (2003 ), . If you're not an engineering educator or you are one and just got back from the latest Mars expedition, let us explain that Outcomes 3a-3k are specified attributes engineering students in accredited programs should have by the time they graduate. They include the usual technical abilities but also such things as communication skills, the ability to work effectively in multidisciplinary teams, and an understanding of the professional and ethical responsibilities of an engineer. 2. N.E. Gronlund, How to Write and Use Instructional Objectives (6th Ed.), Upper Saddle River, NJ: Prentice-Hall, 2000. See also R.M. Felder and R. Brent, "Objectively Speaking," Chem. Eng. Ed., 31(3), 178-179, 1997, . Write detailed learning objectives and let the students in on them Learning objectives (or instructional objectives) are explicit statements of what students should be able to do to demonstrate that they have learned what you want them to learn. 2 The objectives must specify directly observable actions (list, explain, calculate, derive, model, critique, design ... ). Verbs such as "learn," "know," "understand," and "appreci ate" are unacceptable. You can't see students understanding something; to know whether or not they understand, you have to ask them to do something you can see that demonstrates their understanding. For examples, see , a study guide containing a subset of the learning objectives for the introductory chemi cal engineering course. Even if you don't know the course content, you should be able to convince yourself that if the students can do everything on those two pages, they have Richard M. Felder is Hoechst Celanese Professor Emeritus of Chemical Engineering at North Carolina State University. He is co author of Elementary Principles of Chemical Processes (Wiley, 2005) and numerous articles on chemical process engineering and engineering and science education, and regularly presents workshops on ef fective college teaching at campuses and conferences around the world. Many of his publications can be seen at . Rebecca Brent is an education consultant specializing in faculty development for ef fective university teaching, classroom and computer-based simulations in teacher education, and K-12 staff development in language arts and classroom management. She codirects the ASEE National Effective Teaching Institute and has published articles on a variety of topics including writing in un dergraduate courses, cooperative learning, public school reform, and effective university teaching. Copyright ChE Division of ASEE 2006 Summer 2006 173

PAGE 32

probably learned what the instructor wants them to learn in that part of the course. Our first recommendation is to write detailed learning objectives and give them to the students as study guides for exams and as guidelines for other assignments such as project reports and oral presentations. Make sure your objec tives cover all the skills you would like students to master, especially high-level skills (such as critical and creative thinking) and the ABET-mandated outcomes that have not been traditionally addressed in engineering courses, such as communication and lifelong-learning skills. When you are explicit about your expectations, the likelihood that your students will meet them goes up dramatically, especially if the expectations involve difficult or unfamiliar material. 2 Teach skills before you assess them Take problem formulation as an example-another one of those ABET outcomes few of us ever thought about before they showed up in the Engineering Criteria. Suppose an objective in your fluids course states that the students should be able to make up (and solve) fluid dynamics problems whose solutions call for creative thinking. If you simply assign students to do that, most won't get what you're after and you'll see mainly problems that look like, "Given X and Y, calculate Z." That shouldn't surprise or disappoint you, since nobody ever taught them how to do what you're now asking them to do. If, however, you first explain and illustrate what you 're looking for, then show several good and bad examples and have the students work in small groups in class to critique them, then give and grade two assignments that include the same task and perhaps show some submissions from students who got the idea, you'll start seeing creative problems.3 Doing all that would allow you to check off both problem formula tion and critical thinking on the list of outcomes addressed in your course. You can do the same thing for, say, writing techni cal memos ( communication and critical thinking), analyzing workplace case studies (professional and ethical awareness), or critiquing an engineering-related article in the popular press (professional awareness, communication skills, understanding the societal impact of engineering solutions, knowledge of contemporary issues, and lifelong-learning skills). Use rubrics for grading Problem-solving and multiple-choice tests are relatively easy to grade, and if the grading system is rational, students should have no trouble understanding what they did wrong and why they got the grade they did. The same is not true of written project reports, essays, case study analyses, and oral presentations. When students just get grades and a few writ ten comments as feedback, they may understand why they were marked down but may have little idea about how to do it better next time. There's a better way. Use a rubric to grade anything that involves subjective judgment on the part of the grader. Decide on criteria you will use to evaluate the memo, report, or pre sentation (e.g., technically accurate, complete, appropriately documented, well organized, well written, good visuals, sound theoretical analysis ... ); assign weights to each crite rion; and-for a four-point rating scale-briefly summarize the attributes of a 1, 2, 3, and 4 for each criterion. 4 Then, when you give students illustrative written products or oral presentations to critique in class, have them use the rubric individually and then compare their ratings, and then share your ratings with them. When you hand back their assign ments, give them your completed rubric as well-and watch how they improve on the next assignment. As with learning objectives, when a grading system is clear, the students are more likely to meet expectations. If a skill is important to you, assess it Once you've communicated your learning objectives and assessment method for a particular skill, and you've pro vided examples and practice in applying the skill, then (and only then) is it legitimate to test the students' mastery of the skill-and at that point, you definitely should. Engineering students barely have enough time to keep up with their assign ments; they don't have time to dig deeply into everything in all of their courses. If they are sure that something requiring effort on their part to learn won't count toward their grade, most won't bother to learn it. That's not laziness-that's ra tional behavior. The assessment drives the learning: if a skill is important to you, assess it. Here's our challenge to you: (a) Pick a problem-solving or professional ("soft") skill that your students have always had trouble mastering; (b) write one or more learning objectives that list the things you might ask students to do to demon strate mastery of that skill; (c) share the objectives with the students-if possible, as part of an exam study guide; (d) give your class examples of what you're looking for and several opportunities to practice the skill, in and out of class; and (e) assess the students' mastery of the skill (using a rubric if subjective judgment is involved) by asking them to do some of the things specified in the objectives. If you see better performance than you've ever seen (which we always have when we've done all that), consider making this strategy a permanent addition to your teaching toolkit. 0 3. R.M. Felder, "On Creating Creative Engineers," Eng. Ed., 77(4), 222-22 7 ( 1987), . 4. You can see an illustrative rubric for evaluating student presentations at . All of the Random Thoughts columns are now available on the World Wide Web at http://www.ncsu.edu/effective_teaching and at http://che.ufl.edu/~cee/ 174 Chemical Engineering Education

PAGE 33

~S=i curriculum ) ---11111----------HYPER-TVT: DEVELOPMENT AND IMPLEMENTATION OF AN INTERACTIVE LEARNING ENVIRONMENT MARINA SANTORO AND MARCO MAZZOTTI ETH Swiss Federal Institute of Technology Zurich CH-8092 Zurich, Switzerland R apid advances in information technology and easy access to the Web have motivated students and educa tors to use more and more Internet and multimedia technology for educational purposes. Educators and students are challenged by the great potential of the Internet for de livering and sharing a large amount of information among a greater public, and by the possibility of creating alternative and breakthrough ways of teaching and learning by using advanced software. Today, in almost every field of educa tion, a broad range of e-learning material is available online. "Computer-based learning system" has been the catch phrase for the last few years to indicate a wide area of systems using the Web and multimedia technology for education. These systems can be classified depending on the main functionality and research focus as: computer-aided educa tion systems; multimedia/virtual laboratory; distance-learning systems; or intelligent tutoring systems_[ll This classification identifies different levels of e-education: from the simple integration of computer technologies into the traditional teaching system-where face-to-face meetings and personal relationships are still of primary importance-to the stage where the information transfer rate is no longer managed by a human tutor but it is adaptively controlled by an intelligent computer system_[ll In all cases the objective is always to use the technology not just for technology's sake, but to enhance the quality of higher education. This contribution aims to presentHyper-TVT, a public interactive learning environment on separation process technology ( ). ThenameHyper-TVT derives from the word "Hypertext" and the German acronym of the course name, i.e., "Thermische VerfahrensTechnik," meaning thermal separation processes. MOTIVATION In recent years, the educational concept has been extended toward the integration of new generations of Internetand computer-based courses intended to lower the gap between theoretical knowledge and practical experience in the mod em curriculum for process and chemical engineering. This concept addresses the nowadays frequent request of students for continuous, adequate preparation (i.e., solving practical Marina Santoro is a chemical engineer at the ETH Swiss Federal Institute of Technology Zurich, in Switzerland, specializing in sepa ration process technology, with particular emphasis on the development of the new learning technologies and of e-leaming tools. She has developed the e-leaming system for chemical engineers, Hyper-TVT. Marco Mazzotti is professor of process engineering at ETH Swiss Federal Institute of Technology Zurich, in Switzerland. He received his B.ChE. and his Ph.D. from Politecnico di Milano, in Italy. He teaches unit operations and mathematical methods and has published more than 100 scientific papers on separation processes. Copyright ChE Division of ASEE 2006 Summer 2006 175

PAGE 34

problems and facing real issues), thus allowing them to be come better aware of the physical reality of the processes, the industrial world, and their future profession. Likewise, Lewin, et al., claimed that the "instruction of chemical engineers should reflect the challenges they face in industry."[ 2 J Indeed, teaching and learning in the field of chemical engineering can be enormously enhanced by the use of information technology and Web tools. In comparison to traditional textbooks, these new avenues can offer videos, animations, and interactivity that help students visualize the reality behind equations and diagrams, thus compensating for the lack of practical experi ence in common engineering curricula. In that respect, a very good reference for an e-learning en vironment is, in our opinion, the one developed by Fogler and Gulmer in the field of chemical reaction engineering Pl The Web site is offered by the University of Michigan and is freely ac cessible online. Besides texts, equations, and diagrams, this Web site provides interactive tools for self-assessment, vid eos, and audio descriptions to demonstrate the concepts of chemical engineering, helping users visualize the applications within the industrial world. ~ Synchronous 1 Asynchronous .,t / Textbooks ing databases and links to facilitate searches for material, organizations, communities,joumals, and publications about e-learning technology, such as ChemEnginfo,l1 2 lWorldLecture Hall,l1 3 l EuPaCE.net,l1 4 l and MERLOTY 5 l The class on separation process technology is compulsory in most curricula in process and chemical engineering. It demonstrates the application of chemical engineering prin ciples within an industrial context for effective design of processes, particularly of multistage separation processes. Topics covered are: fundamentals of separation processes; absorption and stripping; flash evaporation; distillation; and liquid-liquid extraction. These subjects are particularly suited for the development of e-learning tools due to the synergy be tween theoretical issues and practical experience. Hyper-TVT Face-to-face M-lecture Hyper-M is a freely accessible e-learning system for students of chemical and process engineering at the In the field of separation process technology as well, all top-ranking technical universi ties provide e-learning systems and online courses in an effort to extend traditional chemical engineering classes. Access to these modules, however, is Figure 1. Illustration of the role of Hyper-TVT with respect to traditional education methods. ETH Swiss Federal Institute of Technology Zurich, and also for all individuals and institutions involved or interested in process and chemical engineering educa tion and practice. Its purpose is to complement the already existing tools, and, at the same time, to compensate for their limitations in access and scope, and to take full advantage of interactive and multimedia technology. Therefore Hyper-TVT is not comparable to commercial pro cess simulators such as HYSYS. plant, CHEMCAD, or Aspen Plus, which have different scopes and purposes. very often restricted to enrolled students of the course or to faculty,[ 4 5 l and a lot of material is available only in closed environment via CD-ROM or password-protected LAN.[ 6 l In other cases, the courses are freely accessible but a lot of the online material provides only course information and assignmentsPl or provides the course syllabus as pdf files or PowerPoint presentations. Although this static material is of high quality and has the benefit of being always available online, a very low degree of interactivity, or none at all, is offered and therefore this approach exhibits only minor advantages as compared to traditional textbooks. A lot of interesting material covers only some specific topic of the separation processes, such as distillation. Com prehensive theory, pictures, schemes, diagrams, and videos are published online by research groups at universities and companies working in the field. [S llJ Finally, other universities and organizations are active in the e-learning field, provid176 Hyper-TVT is a didactic sup port, which is complementary to, but not a replacement for, traditional lectures and textbooks. In fact, combining Hyper TVT with conventional education elements-e.g., lecture, textbook study, and discussion with the teacher-contributes to enhancing teaching effectiveness and flexibility. The Web site offers both educators and students a number of tools such as videos, animations, simulations, and a self-assessment environ ment, thus creating a better balance between synchronous and asynchronous approaches in teaching and learning, as illustrated in Figure 1. The design concepts, the didactic content, and the technical features of Hyper-TVT will be discussed in detail in the following chapters. DESIGN CONCEPT "Textbook" structure The structure of the Web site is organized in chapters and paragraphs, as it would be for a traditional textbook. Figure 2 is a screenshot of the page content of the "Contactors" Chemical Engineering Education

PAGE 35

lesson as an example of page structure. The choice of minimizing any structural complexity has two purposes: to let students and educators focus bet ter on the Web site content, and to make using the tool as easy as possible. Within each paragraph the topic is presented using explanatory text, images, and interactive diagrams and schemes. Every paragraph develops a certain concept independently and completely by integrating the multimedia material within the text and by providing logical links to other pages of the learning environment. Navigability Due to its simple structure and the implementa tion of a navigation system, the Hyper-TVT Web site is easy to browse. In every section of the Web site, menus help answer three fundamental ques tions[16l: Where am I?; Where have I been?; Where can I go?Within every lesson, an interactive table of contents highlights the user's current location within the lesson, helps to find other related content, and suggests the logical learning sequence, as shown in Figure 3. Interactivity Hyper herm l sche V erfahrens echn l k Lessons Links Downloads Ca l culator Contact us '11Contactors: industrial equipment 1. GuLiQuid contac:tors 2, GasLiQuid contactors: T ray columns 4, Liquid-Liquid contactors l llyper T\"T: on line Thenni sc h e ... [;ii[d tJ Prof. Marco Mazzotti Institute of Process Engmeenng ETH Zun Cl Jmpnnt I Last update : ~mber l 4, 2005 Figure 2. Structure of the "Contactors" lesson with chapters and paragraphs. A screenshot of a video on column internals is shown on the bottom right. The added value of a Web-based educational system is the interactivity offered to students and educators. The HyperTVT system stimulates interactivity through pictures, anima tions, a modeling and simulation environment, videos, and self-assessment tools, all of which are very consistent with the learning objectives in the domain of process and chemi1 ~~~ ~-~~ -----------~~--~ '~ H ype rhermische V erfahrens echnik '! 1 c ontactor s 1. Gas Liquid 2 Tra y c o lumn s 3. Pa ck ed c o lumn 3 1 description 3 3 compari s on Summer 2006 Lin k s D ow nload s C alculator C ontact u s Y our f e edbac k I ll -Gas Liquid Contactors: Packed columns 3 2 Packing types The t y pe of packing materials in industrial use c an be distinguished into random pac k ing ( al s o called dumped) and structured pac k ing ( also c a lled ordered s ta c ked o r arra n g e d ) Figure 3. Packing types in the "Contactors lesson. On screen the path is indicated by orange highlighted items in the menus. 177

PAGE 36

cal engmeenng. Two examples of animations about distillation topics are shown in Figure 4. Reusability T Hyper-TVT teaches some of the key separa tion processes technolo gies, i.e., a fundamental subject in the process and chemical engineer ing curriculum that is well established. For this reason, the content of Hyper-TVT is not ex pected to be subjected to revision in the near future, if ever. Specification on contactors, pres ents an overview of the industrial equip ment for gas-liquid and liquid-liquid separation. Lessons four, six, and seven cover the design of three of the most im portant multistage separation processes. The fifth lesson, on Flash evaporation, introduces many concepts useful to understanding distil lation. The didactic approach first pro vides students with all basic concepts and tools needed. Then, students are xo= ~ x,= ~.O H ti d R /~ ~-0 2 --fc ;( --1----++--+-+l o s I. XO Therefore the Hyper TVT Web site is avail able, now and in the fu ture, to students, educaFigure 4. Interactive Flash animations within the "Distillation" lesson. challenged to use tors, and practitioners of different institutions, countries, and educational sectors, e.g., higher education, vocational training, professional organiza tions, industry, and schools. Outreach to other categories of users The target users of Hyper-TVT consist of all individuals and institutions involved in process and chemical engineering education. Its simple structure, links, and interactive modules, however, make Hyper-TVT easy to be followed by others, who, though not chemical or process engineers, still need or want a quick and complete overview of the separation process technologies. DIDACTIC CONTENT AND METHOD Hyper-TVT consists of seven lessons presenting class ma terial using text, images, animations, interactive tools, and simulation environments. These are: 1. Introduction of separation processes 2. Fundamentals of separation processes 3. Contactors 4. Absorption and stripping 5. Flash evaporation 6. Distillation 7. Liquid-liquid extraction The first lesson is a presentation of the separation processes and their role and importance in the industrial context. In the second lesson, basics of thermodynamics and mass transfer are revisited; these are fundamental for further understanding of multistage separation technologies. Lesson number three, 178 the new material to solve problems given as home assignments. Besides the les sons, three additional sections of the Web site contain videos that can be streamed, PowerPoint and pdf files that can be browsed, and tests that can be used for self-assessment. The videos focus on industrial equipment for separation processes (i.e., lesson three) and have been partly created for this use and partly provided by companies, e.g., Sulzer ChemTech, Ktihni, and FRI (Fractionation Research, Inc.). In the videos, an off-screen narrator guides the visitor into a virtual tour of the real equipment to observe directly and in detail phenomena that neither words nor photos alone would be able to clarify. A screenshot of a video about column internals is shown in Figure 2. Videos can have a great didactic value, and not just in the field of chemical and process engineering, because they overcome physical barriers and bring the world into the classroom. The PowerPoint and pdf files are a collection of mathematical derivations of the equations used in the lessons for process design. Their format allows students to use them interactively online or to download them for further reading. The test section provides multiple choice and descriptive questions. Students also have the option to submit completed questionnaires, receive support online, and access and print homework assignments as pdf files. Another important ele mentoftheHyper-TVTWeb site is its database. The database contains five specific search categories, e.g., text, notation, images, glossary, and bibliography. The "text" category provides links to pages where keywords are mentioned. "Notation" contains detailed explanation of all the symbols used in the Hyper-TVT Web site. "Images" is a collection of diagrams and graphics already present in the lessons, but it Chemical Engineering Education

PAGE 37

Feedback from the ETH Zurich students indicates 95% appreciated the learning environment and found it useful, both during lectures and outside the classroom .... also contains additional pictures, schemes, and photos of real equipment provided by various companies. The "glossary" category is a compact dictionary of terms specific to the les sons presented in Hyper-TVT. Finally, the "bibliography" provides a list of suggested textbooks, handbooks, and further readings. All five categories can be accessed by a search tool. This allows visitors and students to make quick searches in a specific area of interest and find direct links to the Web site's relevant section. A further aspect of the Web site is its toolbar. This provides three additional links: to some of the most important chemical engineering companies in the field of the separation process technology; to the required-plug-ins Web page; and to an evaluation section for online feedback and comments. TECHNICAL FEATURES Hyper-TVT is an independent learning environment in which technical tools and software have been chosen not only to suit the pedagogic needs of the class on separation processes but also to facilitate its future management, maintenance, and upgrade. Access is also possible from limited-band con nections, and compatibility has been guaranteed for all main browsers on both PC and Mac machines. The architecture of the Web site is based on the PHP scripting language, which allows for creation of dynamic pages and the use of databases, particularly of MySQL. Javascript has been used to build the navigation menus and the interactive tools, i.e., the table of contents and the self-assessment environment. Animated diagrams, schemes, and interactive presentations have been shown with Flash, Real Player, and QuickTime. Each has also been used for video streaming. Finally, MathML (version 1.0), has been used to edit and display the mathematical equations. At ETH Zurich, Hyper-TVT has been one of the first educa tional projects to use MathML, thus providing a useful and successful experience for other projects as well. DISCUSSION AND CONCLUSION The development of Hyper-TVT started in July 2001 and its realization has taken three-and-a-half years with one per son working on it full time. A prototype of the Web site was released to students in June 2002, and their feedback has been used during revision of the site and completion of the project. The whole environment is continuously revised based onfeedSummer 2006 back collected in different ways. The ETH students attending the class are asked to fill out an evaluation form at the end of the semester, and an online evaluation form allows students and educators from other institutions to provide helpful com ments. On one hand, these have been used for modifications, refinements, and improvements of some parts of the Web site. On the other hand, and more in general, feedback from ETH Zurich students indicates 95% appreciated the learning environment and found it useful, both during lectures and outside the classroom, as support material and in completing homework assignments. The Web site is used very intensively in preparation for the final exam. The major advantage, as in dicated by students, is the interactive and audio-visual content, i.e., the flash animations, videos, diagrams, and pictures. As a further positive comment, students and practitioners underline the easy navigability offered by the Web site. Very positive reactions have been received from educators of other institu tions and technical universities as well. The result, although possibly not yet statistically relevant, is very encouraging about the usefulness of the Web site. Use of the Web site is not a requirement of the separation processes class for students at ETH Zurich. Rather, the Web site is presented at the beginning of class as additional support and its interactive material and videos are used during some of the lectures. In this way, students get a first impression of the real world of separation process technology while gaining familiarity with the Web site, thus making its use easier. The purpose of this approach is to increase student interest about the proposed subject with more motivating material and tools, thus stimulating their self-study skills and responsibility. The most difficult issue we have experienced as instructors has been integration of the Web site during traditional lectures. In fact, this implies an allotment of time in the lecture plan for use of the computer-based didactic material. This is not a simple task since the traditional course has to be reorganized and restructured to implement a new hybrid methodology (tra ditional and computer -based lecture). It requires an additional effort and a lot of motivation by the instructor. For students as well the integration of the Web site into their traditional way oflearning is challenging. Many studies have been con ducted to determine what factors impact the perception and acceptance of new e-learning tools on studentsY 7 l Such tools generally require students to become more independent and 179

PAGE 38

more responsible of their personal learning processes, and to apply new technologies -an additional workload that is not always well received at first. With the time and guided assistance of the tutor, however, students usually recognize and appreciate the benefits of the computer-integrated edu cational system. Typically, their learning efficiency and performance improve, resulting in increased self-confi dence and greater motivation. The Hyper-TVT system is available online () and can be accessed without restrictions through the Web. Hyper-TVT shows how dissimilar pedagogic methods tra ditional and Web-based-can be implemented in parallel to offer a more stimulating and productive learning environ ment. Students, in turn, gain tools for self-paced learning and become more autonomous. Use of Hyper-TVT can foster their skills in analysis, synthesis, and evaluation. Finally, the lecture time with the instructor can be invested more effec tively to discuss advanced issues. Our experience in using Hyper-TVT has been very positive so far, and we encourage students, educators, and chemical and process engineers to explore it as well. ACKNOWLEDGMENTS Hyper-TVT has been developed within the Filep program 180 of ETH Zurich, which the authors thank for funding and support. REFERENCES 1. Shin, D., E.S. Yoon, K.Y. Lee, E.S. Lee, "A Web-Based Interactive Virtual Laboratory System for Unit Operations and Process System Engineering Education: Issues, Design, and Implementation," Comp. & Chem. Eng., 26, 319-330 (2002) 2. Lewin, D.R., WD. Seider, andJ.D. Seader, "Integrated Process Design Instruction," Comp. & Chem. Eng., 26, 295-306 (2002) 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. Nielsen, J., Designing Web Usability, New Riders, USA (2000) 17. Platteaux, H., "How Students Perceive e-Learning Situations: The Case of the SVC Embryology Course, "Proceeding of the 5th International Conference on New Educational Environments, Lucerne (2003) 0 Chemical Engineering Education

PAGE 39

~S=i curriculum ) ---11111----------INTEGRATING BIOLOGICAL SYSTEMS in the Process Dynamics and Control Curriculum ROBERTS. PARKER University of Pittsburgh Pittsburgh, PA 15261 FRANCIS J. DOYLE III University of California at Santa Barbara Santa Barbara, CA 93106 AND MICHAEL A. HENSON University of Massachusetts at Amherst Amherst, MA 01003 T he discipline of chemical engineering is evolving, as evidenced by the recent wave of departmental name changes that reflect both the increasing number of chemical engineering faculty involved in research on biology oriented topics, and the fact that the percentage of chemical engineering undergraduates obtaining initial employment with companies in the biotechnology and biomedical sectors increased from 4.6% in 1998 to 10.3% in 2001-02Yl A series of MIT-organized and NSF-sponsored workshops examined the current state of undergraduate chemical engineering education and recommended a sweeping set of changes.[ 2 l Foremost among the proposed changes were the introduction of biology as a core science, the importance of addressing complexity, and the expanded use of the systems approach. The present discussion focuses on these three elements within the context of the traditional process dynamics and control curriculum. The dynamics and control course, typically taught late in the junior or senior year, is a natural point for including biological systems content along with chemical process material. Due to the focus on general principles rather than specific processes, biological systems can be integrated without detracting from the coverage of more traditional applications. This expanded vision of the system dynamics and control curriculum requires the following difficult issues to be addressed: (1) how can these complex systems be introduced in a meaningful way to undergraduate chemical engineers with little background in biology?; and (2) what changes are required to include biological content without sacrificing the traditional core of process dynamics and control? The objective of this paper is to provide some practical answers to these questions using the experiences of three courses taught at our respective instituCopyright ChE Division of ASEE 2006 Summer 2006 tions. The first two examples illustrate the introduction of biological content into the traditional process control course, while the third example focuses on the development of a new course in which the systems approach is applied to a diverse set of biological problems. Robert S. Parker is an associate profes sor in the Department of Chemical and Petroleum Engineering at the University of Pittsburgh. His educational interests focus on the area of dynamical systems analysis and control. He is currently involved with the implementation of an integrated curriculum and the develop ment of cross-cutting biological prob lems to assist students with integrating material across courses. Francis J. Doyle Ill holds the Duncan and Suzanne Mellichamp Chair in process control in the Department of Chemical Engineering at the University of California at Santa Barbara. He holds joint appointments in Electrical Engi neering as well as in the Biomolecular Science and Engineering Program. His educational interests are in process dynamics and control, systems biol ogy, and the Introduction to Chemical Engineering course. Michael A. Henson is a professor of chemical engineering at the University of Massachusetts, Amherst. His edu cational interests are in the areas of process modeling and control. He is involved in a variety of educational initiatives including development of a cross-disciplinary biological-systems engineering curriculum and participation in a CACHE task force on systems-biol ogy education. 181

PAGE 40

INTEGRATION OF BIOLOGICAL SYSTEMS CONTENT A typical process dynamics and control course covers a broad range of new material at a rather brisk pace. To produce students who can apply traditional dynamic analysis and controller design techniques is a formidable challenge even when the focus is purely on chemical process systems. The addition of biological content along with the requisite model ing and analysis techniques requires a carefully crafted course to avoid leaving students overwhelmed. A possible structure for a semester-long course is illustrated by the syllabus in Table 1, where NL is the number of lectures allotted to the specific topics listed in all caps. Bold entries represent new topics specific to biological systems. Italicized entries are theoretical topics often considered optional in a traditional course but which are viewed as important for a biologically oriented course. The introduction of state-space behavior and analyze system response in the presence of nonlinear phenomena. Without question, this topic could comprise a course unto itself. Some basic tools (e.g., phase planes, limit cycles, bifurcation) are easy enough to teach in a class or two, however. These provide students with an ability to identify nonlinear system characteristics, even if they cannot design a linearizing-state feedback controller to address the underlying nonlinearity. Feedback is a concept that is introduced naturally in the context of biological system examples. The representation of biological control systems using various elements of the traditional block diagram is particularly effective. This approach, however, should be used carefully to avoid concealing the complexity of the underlying biological processes. Throughout the topic sequence in Table 1, a number of examples serve to highlight the breadth of opportunities for application of the theoretical concepts presented in the course. Table 2 provides a list of potential case studies. For each TABLE 1 models and associated analysis tools is essential for the treat ment of biological systems due to their complexity (e.g., high order, multi variable, highly nonlinear), which often precludes simple Laplace domain treatment. A few lectures on matrix algebra and linear state-space systems are necessary to review core material and ensure that students with defi cient backgrounds understand the basic concepts. When combined with the linear systems analysis lecture, this material allows the calculation of eigenvalues to de termine stability and matrix rank for the analysis of controllability and observability. The nonlinear systems theory lecture includes the traditional topic of Jacobian linearization as well as Proposed Syllabus for a Biologically Oriented Dynamics and Control Course introductory coverage of phase plane analy sis, multiplicity, and bifurcations. Biological systems are inherently nonlinear, given the ex istence of saturation phenomena, stable os cillations, etc. As such, a student must have a working knowledge of nonlinear systems to be able to identify such 182 NL=number of lectures all caps=topic area bold=new topics italicized=optional topics NL Topics 4 DYNAMIC MODELING Principles of fundamental modeling; chemical and biological process examples; introduction to empirical modeling 7 LINEAR AND NONLINEAR SYSTEMS ANALYSIS Matrix algebra and linear state-space systems; linear systems theory; introduction to nonlinear systems theory; dynamic simulation; chemical and biological process examples; introduction to the Laplace transform 7 FEEDBACK SYSTEMS Basic principles of feedback; physiological control systems; homeostasis as a setpoint-free feedback system; feedback in biochemical reaction networks; closed-loop response analysis; servo vs. load behavior; feedback control of chemical process systems; closed-loop drug delivery 8 FEEDBACK CONTROL SYNTHESIS Basic principles of model-based controller design; PID controller design and tuning; advanced single-variable control techniques; multivariable control techniques; model predictive control; chemical and biological process examples 4 ADVANCED TOPICS Large-scale systems and plantwide control; parameter estimation and experimental design; state estimation; introduction to systems biology TABLE2 Possible Case Studies for the Process Dynamics and Control Course Chemical Processes Continuous and/or fed-batch polymerization reactor; distillation column; continuous pulp digester; paper machine; simple plantwide example (e.g., reactor and separator); semiconductor process (e.g., lithography); photovoltaic film processing; fuel cell Biotechnological Systems Continuous and/or fed-batch fermentor; yeast energy metabolism; cell stress response (e.g., heat shock); eukaryotic cell cycle; bacterial chemotaxis Biomedical Systems Baroreceptor vagal reflex (blood pressure control system); insulin-dependent diabetic patient (glucose-insulin metabolism/ control); circadian rhythm gene regulatory network; anesthesia control; drug delivery for HIV treatment; drug delivery for cancer treatment Chemical Engineering Education

PAGE 41

topic where examples are listed in the syllabus, two chemical process and two biological system examples could be used to develop lecture materials, in-class exercises, and recitation problems. Ideally the biological problems are divided equally between the biotechnology and biomedical lists. A major conclusion of the MIT-organized education work shops was that multiscale phenomena should be incorporated throughout the undergraduate chemical engineering curricu lumPl A useful connection between the traditional chemical and biological examples listed in Table 2 is the wide range of time and length scales at which these systems can be analyzed. Polymerization reactor models can be developed using input output representations Pl detailed descriptions of the individual polymer particles and their interactions,l 46 l or a variety of scales in between. [ 7 9 l Analogous models can be developed for microbial fermentors where lumped descriptions of cel lular processes are provided by unsegregated models[ 10 13 J and detailed descriptions of the individual cells are provided by cell population models. [l 4 J While the introduction of biological systems content is not necessarily required to illustrate these concepts, we feel that an integrated program of chemical and biological examples will reinforce key concepts and demon strate that these diverse examples are conceptually similar. UMASS CHE 446: INCORPORATING BIOTECHNOLOGY The process dynamics and control course at the University of Massachusetts ( ) has traditionally focused on Laplace transform methods and chemical process applications. This course usually repre sents the only extensive exposure to dynamic modeling and feedback control in the undergraduate curriculum. Biological systems were chosen as an appropriate vehicle for introducing the key elements of biological transformations, multiscale phenomena, and systems-level analysis identified in the MIT sponsored education workshops.[ 2 l Rather than completely change the existing course content, a more conservative approach based on the integration of biological systems and the requisite analysis techniques was pursued. The current syllabus for the UMass course (ChE 446) is shown in Table 3, where new topics introduced in the past two years are italicized. The first few weeks are focused on fun damental modeling because undergraduate students typically have little experience formulating dynamic balance equations. Two biological examples-a continuous yeast fermentor model and a structured yeast cell model-are introduced and revisited throughout the semester. Both time domain and Laplace domain analysis techniques receive extensive cover age. A major focus is the formulation and stability analysis of linear state-space models. Engineered and natural-feedback systems are introduced in parallel to highlight their common features and unique properties. While most of the material on single-loop controller synthesis is traditional, an introduction to time domain controller design and analysis techniques is provided to parallel the Laplace domain methods. The final few weeks are focused on multi variable control systems with an emphasis on linear model predictive control. To accommodate the new material on biological systems and time domain techniques, material previously covered in the course had to be de-emphasized or virtually eliminated. Topics that received reduced coverage included transfer func tion models, Laplace domain analysis and design techniques, advanced single-loop control strategies, and traditional chemi cal process examples. Frequency domain techniques received very limited coverage. While these topics are admittedly important, a broader view of dynamic systems and feedback TABLE3 Syllabus for UMass Course ChE 446: Process Control NL=number of lectures italicized=new topics all caps=topic area NL Topics 5 FUNDAMENTAL MODELING Basic principles; chemical process examples (nonisothermal chemical reactor; binary flash unit; binary distillation column); biochemical system examples (continuousfermentor model; metabolically structured yeast cell model) 7 DYNAMIC SYSTEM ANALYSIS Linear algebra (solution of matrix equations, state-space models; eigenvalues and eigenvectors); time domain analysis (basic stability concepts, linearization of nonlinear models, linear stability analysis, continuousfermentor example); Laplace transforms; transfer function models; empirical models; parameter estimation 6 FEEDBACK SYSTEMS Process control systems; biological feedback systems ( engineered vs. natural feedback systems, yeast sulfate assimilation pathway, baroreceptor vagal reflex); closed-loop transfer functions; closed-loop stability 7 FEEDBACK CONTROL SYNTHESIS PID-controller tuning; internal model control; time domain controller design (state feedback, pole placement, model matching, continuousfermentor example); feedforward control; cascade control 5 MULTIVARIABLE CONTROL Control loop interactions; decentralized control; discrete-time models (discretization of continuous-time models, convolution models, prediction models); model predictive control ( controller design and tuning, constraint handling, real-time optimization, continuousfermentor example) Summer 2006 183

PAGE 42

control was deemed to be more important given current trends in the chemical engineering profession. In fall 2003, each student was asked to evaluate the biological systems content using a score ranging from "5" if they strongly agreed the objective was achieved to "1" if they strongly disagreed the course objective was achieved. Results obtained from the 21 respondents are summarized in Table 4. The average scores are similar to those obtained for the other course objectives, thereby indicating that the biological content was successfully integrated into the course. PITT CHE 0500: INTRODUCING BIOMEDICINE The biology component in the Systems Engineering I: Dynamics Modeling course (ChE 0500, ) at Pittsburgh focuses on the analysis of, and controller synthesis for, biomedical systems at the whole-organism level. By integrating the research activities in modeling and control of diabetic and cancer case studies within the undergraduate class, students are exposed to a novel application area. This format has resulted in a steady flow of undergraduates interested in undergraduate research, and an increased interest in graduate study. Students at Pitt were posed the same questions as those at UMass; responses can be found in Table 4. While confidence in dynamic balance construction is not as high as that shown in the UMass course, the other questions return similar quantitative responses indi cating that the biomedical topics were well received. ChE 0500 is approached from a model-based perspective; approximately half of the course is focused on modeling systems using both fundamental and empirical approaches, in both continuous and sampled-data (i.e., discrete) domains. From the fundamental modeling perspective, the students are taught to distinguish pharmacokinetics (the time profile of a drug) from pharmacodynamics (the disease dynamics, effect of the drug on the disease, and toxicity) in much the same way valve dynamics and process output response are captured by separate blocks in a block diagram. The remainder of the course focuses on the model-based synthesis and analysis of classical and advanced control systems, as in Table 1. As a case study, consider the insulin-dependent diabetic patient depicted in Figure 1. Fundamental model construction introduces students to the key variables of the diabetic-patient problem and demonstrates the utility of skills developed elsewhere in the curriculum (e.g., dynamic mass balance with reaction, transport resistance) in the modeling of biomedical problems. Students then work with this model, or suitable lower-order approximations,l1 5 l throughout the semester on in-class problems, homework, etc. The case study method[ 16 l is commonly employed in teach ing to facilitate in-depth treatment of problems in limited -' L,Insulin Meal Disturbanccf 1 Exercise Disturbance I B~;i~ f Heart/ Lungs Liver I ... IKidne~ :: i-Periphery Glucose Measurement Figure 1. Open-loop schematic of the diabetic patient. Small solid blocks represent the fundamental model, with manipulated input insulin delivery rate, meal distur bance, exercise disturbance, and glucose concentration measurement. TABLE4 Student Responses to Biological Systems Content in the UMass (21 respondents) and Pitt (17 respondents) Process Control Courses Score Question UMass Pitt I can construct a dynamic model of a biological system. 3.83 3.23 I can perform dynamic system analysis and controller design in the time domain. 3.78 3.71 I can apply dynamic system analysis techniques to biological systems to evaluate 3.89 3.76 properties such as stability. I can describe the relevance of feedback control theory to biological systems. 3.83 3.77 184 Chemical Engineering Education

PAGE 43

UCSB CHE 154: A COURSE IN SYSTEMS BIOLOGY classroom time. An added benefit would be to use a unifying application, thereby allowing students to focus their attention on a single problem. The diabetic patient is one such problem, and case studies from the literature have been mapped onto the course outline (Table 1). The map in Table 5 provides a guide to focused literature reading that allows biomedically motivated problems to be quickly brought into the classroom. Case study-specific tables of this form are most useful to faculty who are not dynamics and control experts, but who are responsible for teaching the course, because the dynamics and control class is a challenging course for nonexperts to teach. A collection of these paper-topic maps, for traditional and biological case studies, would provide those teaching the dynamics and control course with a variety of examples tailored to each section of the course. In addition to the required dynamics and control course, de scribed earlier, there is a demand in many chemical engineer ing programs for elective courses that facilitate specialization in either systems engineering or biotechnology/biomedical engineering. At UCSB, a new course was offered in the spring 2004 quarter entitled Engineering Approaches to Sys tems Biology (ChE 154/BMSE 255). The course is taught at a dual level (seniors and new graduate students), and fulfills the track requirement for both systems and biology emphases in the undergraduate chemical engineering program. The current syllabus is listed in Table 6, detailing the topics for a single quarter course (20 lectures of duration 75 minutes). TABLES Integration of Sample Case Study (insulin-dependent diabetic patient) with Course Outline Topics DATA-DRIVEN MODELING Sorensen FOTD,[ 23 l Bolie two-state linear,l1 5 l Bergman "minimal" model 124 l FIRST PRINCIPLES MODELING Physiologically based pharmacokinetic/pharmacodynamicl 18 23 l LINEAR SYSTEMS ANALYSIS Bolie two-state linear ODEs[ 15 l LINEAR SYSTEMS ANALYSIS w/ LINEARIZATION Linearize and analyze Bergman "minimal" 124 l DYNAMIC SIMULATION All models, including AIDA as a different performance classification[ 15 18 23 26 l FEEDBACK SYSTEMS Glucose-insulin interactions[ 15 l; nonlinear feedback response 124 l; healthy pancreas response 123 27 l CLOSED-LOOP ANALYSIS Sorensen healthy patient12 3 l PIDCONTROL Controller design from FOTD, [ 23 J low-order OD Es, [!SJ and linearized systems and/or effects of nonlinearity 124 l ADVANCED CONTROL Feedforward for meal disturbances 128 l and exercise, [ 29 J with simple[ 15 24 l or complex[ 23 J case studies MULTIVARIABLE CONTROL MISO (glucose and insulin inputs; G, I, and exercise inputs)B 0 l or MIMO (glucose and insulin control) for a variety of systems[ 15 23 24 l MODEL PREDICTIVE CONTROL Linear MPC in analytical[lSJ or data-drivenB 1 l forms; MPC with a linearized model[ 23 24 32 33 l; nonlinear MPC if desired 123 24 32 34 l TABLE6 Syllabus for UCSB Course: ChE 154 Engineering Approaches to Systems Biology NL Topics 6 CELLULAR REGULATION Central dogma; genome sequences; genome expression; genomic circuits; protein, metabolic, signaling networks; high throughput biological data; biological databases 6 MATH MODELING AND SYSTEMS ANALYSIS TOOLS Modeling strategies; boolean models; nonlinear ODE models; discrete stochastic models; systems biology modeling packages; network analysis-robustness, identifiability; design of experiment issues 6 BIOSYSTEMS CASE STUDIES Bacterial chemotaxis; lambda phage virus; circadian rhythm gene network; signal transduction in apoptosis; synthetic biological circuits 2 COURSE PROJECTS Midterm progress reports; final presentations Summer 2006 185

PAGE 44

The course focuses on the emerging problems in systems biology and computational biology. There is a substantial level of effort being invested in these areas in both academia and industry, and the demand for training of students has increased in proportion. These advances have been facilitated by developments in both computational modeling and high throughput biology-enabling a systematic approach to ana lyzing complexity in biophysical networks that was previously untenable. These studies provide increasingly detailed insights into the underlying networks, circuits, and pathways respon sible for the basic functionality and robustness of biological systems. They also create new and exciting opportunities for the development of quantitative and predictive modeling and simulation tools. Model development involves translat ing identified biological processes into coupled dynamical equations that are amenable to numerical simulation and analysis. These equations describe the interactions between various constituents and the environment, and involve mul tiple feedback loops responsible for system regulation and noise attenuation and amplification. The discipline of "systems biology" has emerged in re sponse to these challenges,D 7 l and combines approaches and methods from systems engineering, computational biology, statistics, genomics, molecular biology, biophysics, and other fields. The recurring themes include: (i) integrative viewpoints toward unraveling complex dynamical systems, and (ii) tight iterations between experiments, modeling, and hypothesis generation. In response, there have been a number of courses introduced in a variety of departments across the country that address elements of systems biology and computational biology. These have been targeted at both undergraduate and graduate audiences, and in some cases involve continuing education participants from industry. The balance of topics in the syllabus in Table 6 is approximately one-third on ba sic cellular regulation, one-third on applications of systems engineering tools to biological problems, and one-third on detailed case studies to illustrate current methodologies and future challenges. Although the UCSB curriculum is based on quarters, the same general template could be extended to a semester-long course without significant modification. Assignments for this course consist of short homework problems, primarily at the beginning of the course, and a major course project. The project entails a midterm progress report, a final presentation, and a written report. The case study offers a mechanism to tailor the course to a diverse student popula tion-seniors work in teams with a reduced scope, while graduate students work as individuals on a more detailed project. OPEN ISSUES Laplace Domain Methods Traditional process control courses emphasize Laplace transform methods for analyzing and designing feedback systems. While traditional analysis may be facilitated by 186 Laplace domain representations, the applicability of these methods to the complex systems commonly encountered in biological problems is severely limited. Biological systems are inherently nonlinear with phenomena ranging from pro tein interactions in gene regulatory networks to adaptation in systemic reflexes. Furthermore, modeling of biological systems at resolutions below the macroscopic scale often leads to high-state dimensionY 4 18 l As is evident from Table 1, Laplace domain methods have been de-emphasized and frequency domain techniques have been effectively removed from the proposed curriculum. While we do not dispute their potential value, transform-based methods introduce concep tual difficulties that cause many students to lose their physical insights and view the material as applied mathematics. On the other hand, the syllabus in Table 1 is sufficiently flexible that limited coverage of frequency domain methods at the expense of other topics is possible. Time Domain Methods Complex dynamic system models are most effectively for mulated and analyzed in the time domain using conservation equations. Consequently, the syllabus in Table 1 focuses on linear and nonlinear state-space models. Connections with the corresponding Laplace domain concepts can be introduced as appropriate (e.g., stability via eigenvalues vs. poles). On the other hand, the Laplace transform is a particularly useful tool for single-input, single-output (SISO) systems with time delay and/or zero dynamics. We acknowledge that analytical treatment of zeros in the time domain is more involved than the corresponding Laplace methods. Time domain analysis of transportation and measurement delays is most conveniently performed using a discretized framework based on state augmentation. Because this approach can lead to potentially large state dimensions, evaluating student understanding of this material can be challenging. A possible solution is to use a combination of relatively simple exam questions and more detailed homework problems. While control system design issues can be addressed using continuous state-space models, we believe that a discrete-time framework is pre ferred for introducing data-driven model identification and sampled-data systems. Recent results have shown that a properly tuned SISO model predictive controller cannot be outperformed by a conventional proportional-integral-deriva ti ve (PID) controllerY 9 l Because we expect this fact to be reflected in industrial practice, the syllabus in Table 1 offers increased exposure to controller synthesis techniques based on discrete-time representations such as step response models. While a comprehensive treatment is beyond the scope of this course, model predictive control (MPC) should be foremost among the topics covered due to its industrial importance. As outlined in the UMass course syllabus (see Table 3), the introduction of MPC necessitates limited discussion of real time optimization and draws on the discrete-time modeling tools discussed above. Chemical Engineering Education

PAGE 45

Multivariable Control While most traditional courses treat multi variable systems as a straightforward extension of SISO systems, a more com prehensive approach that addresses the unique challenges of multi variable controller design is warranted. A formal intro duction to decentralized control would support the systems viewpoint of multi variable processes -a set of optimal SISO feedback loops generally does not result in overall system optimality. Another advantage of introducing MPC is that multi variable system complexity is handled in a transparent and systematic manner. Students can gain appreciation for the effects of constraints and optimization-based methods for constraint compensation. Robustness A critical topic in the analysis of both process control systems and biological regulation is robustness. While the remarkable levels of robust performance attained in nature are enviable from an engineering perspective, this issue is not widely appreciated in biology. The critical importance of ro bustness in understanding disease states, as well as evolution and development, motivates its incorporation in the system dynamics and control curriculum. While a detailed theoretical treatment[ 20 J is beyond the scope of a typical undergraduate course, key concepts of robustness can be emphasized us ing simple tools such as sensitivity analysis-effectively capturing the gains from uncertain system elements to the controlled output or performance measure. Students would be well positioned to evaluate parametric sensitivities using state-space models in the proposed curriculum. Robustness analysis could also be used to study closed-loop strategies such as redundancy, feedback, filtering, and modular protocols commonly used in nature. Nonlinear Analysis and Control Most biological systems are not adequately described by linear dynamic models since nonlinear effects such as satu ration phenomena are ubiquitous. Consequently, linear and linearization-based analysis techniques are rarely sufficient. Nonlinear analysis techniques, such as phase plane analysis and bifurcation theory (see Table 3), can be introduced explicitly, thereby exposing students to theoretical concepts and analysis tools with wider applicability than Laplace domain methods. Nonlinear phenomena are also common in industrial plants, and linear control methodologies often require specialized tools to handle strong nonlinearities. Linear controllers exhibit poor performance for some nonlinear processes (e.g., high purity distillation columns) and completely fail for particularly dif ficult processes (e.g., those displaying input multiplicity). Given increased exposure to linear MPC in the revised curriculum, a brief introduction to nonlinear MPC is entirely feasible. Teaching Control for Nonexpert Faculty Our experience indicates that the process dynamics and control class is not a popular choice as a teaching assignment Summer 2006 among nonexperts in the field. This lack of interest is due to a variety of issues, including the mathematical complexity of the material and the significant focus on feedback con troller synthesis. An additional concern is that the material is challenging to students, who have had limited exposure to dynamical systems prior to this course. The syllabus in Table 1 represents a significant departure from the traditional controller-synthesis-dominated course to a more balanced presentation of system dynamics and feedback. A notable benefit of the proposed syllabus is the degree of potential customization. While our focus has been on the introduction of biological systems content, the treatment of other application areas such as advanced materials can be ac complished in a similar manner. This flexibility provides an excellent opportunity for instructors to integrate their research interests into the course. In fact, the three courses described here were heavily influenced by the work performed in our re search groups. Possible benefits of such integration include: (i) increasing the diversity of application examples by encourag ing nonexperts to teach the course; and (ii) introducing students to cutting-edge research that influences their perception of the field and may affect their future career directions. SUMMARY Biological processes have assumed an increasingly im portant role in chemical engineering research and practice. Modifications of the existing chemical engineering curriculum are necessary to provide undergraduate students the needed exposure to this emerging field. We believe that the capstone process dynamics and control course provides an excellent opportunity to integrate biological systems content and draw parallels with chemical process applications that have been the traditional focus of this course. This paper provides a sum mary of work on this problem at our respective institutions. The proposed curriculum allows biological content and time domain concepts to be introduced in a synergistic man ner without adversely affecting the coverage of traditional material. As outlined in the proposed syllabus, this requires a decrease in time spent on traditional topics such as PID controller synthesis, Laplace transform techniques, and fre quency response analysis. Advances in feedback controller tuning (e.g., auto tuning and model-based methods) combined with the availability of simulation/analysis tools (e.g., MAT LAB, Lab VIEW) bring into question the need for extensive treatment of pencil-and-paper analytical techniques that are rarely employed, even at the graduate level. While focused time on these topics has been reduced in the name of incorpo rating biology, it should also be noted that the analysis tools introduced in the dynamics and control class are applicable to problems beyond biological systems. Hence, students are no less prepared for "traditional" industrial positions, and they are certainly more equipped for positions in pharmaceuticals and systems biology. 187

PAGE 46

A key hurdle that must be overcome is the lack of instruc tional materials to support the new process dynamics and control curriculum. For the courses outlined above, the authors are using new textbooks (System Modeling in Cell Biology, MIT Press) or have developed supplementary materials to complement existing textbooks. Researchers in process dy namics and control can contribute in a variety of ways. The construction of extended case studies such as Table 5 for various applications would ease the burden on nonexperts teaching the course. Software tools such as the Process Con trol Modules[ 21 i and Java-based Control Modules[ 22 i are well suited for introducing traditional concepts and applications. New software tools are needed to expose chemical engineer ing undergraduates to biological complexity and to allow the application of theoretical concepts to representative biological systems. Ongoing efforts, such as those organized by MIT and the CACHE Corporation, are focused on the development of biologically focused systems courses. A task force headed by the second author of this paper is currently working on course revisions as well as software module development as a means to integrate biological content throughout the chemical engineering curriculum. More details on this effort will be made available at . ACKNOWLEDGMENTS Support for RSP was provided by the National Science Foundation CAREER program (CTS #0134129). REFERENCES 1. AIChE, 2001-2002 initial placement of chemical engineering graduates (2002) 2. Massachusetts Institute of Technology, Frontiers in Chemical Engineer ing Education Initiative (2003) 3. Parker, RS., D. Heemstra, F.J. Doyle III, RK. Pearson, and B.A. Ogun naike, "The Identification of Nonlinear Models for Process Control Using Tailored 'Plant-Friendly' Input Sequences," J. Proc. Control, 11, Sp. Issue SI:237-250 (2001) 4. Immanuel, C.D., C.F. Cordeiro, S.S. Sundaram, E.S. Meadows, T.J. Crowley, and F.J. Doyle III, "Modeling of Particle Size Distribution in Emulsion Co-Polymerization: Comparison with Experimental Data and Parametric Sensitivity Studies," Comp. Chem. Eng., 26, 1133-1152, (2003) 5. Semino, D., and WH. Ray, "Control of Systems Described by Popula tion Balance Equations I. Controllabity Analysis," Chem. Eng. Sci., 50(11), 1805-1824 (1995) 6. Semino, D., and WH. Ray, "Control of Systems Described by Popula tion Balance Equations II. Emulsion Polymerization with Constrained Control Action," Chem. Eng. Sci., 50(11) 1825-1839 (1995) 7. Congalidis, J.P, J.R. Richards, and W.H. Ray, "Feedforward and Feedback Control of a Solution Copolymerization Reactor," AIChE J., 35(6) 891-907 (1989) 8. Uppal, A., W.H. Ray, andA.B. Poore, "On the Dynamic Behavior of Continuous Stirred Tank Reactors," Chem. Eng. Sci., 29, 967-985, (1974) 9. Daoutidis, P, M. Soroush, and C. Kravaris, "Feedforward/Feedback Control of Multivariable Nonlinear Processes," AIChE J., 36(10) 1471-1484 (1990) 10. Henson, M.A., and D.E. Seborg, "Nonlinear Control Strategies for Continuous Fermentors," Chem. Eng. Sci., 47, 821-835 (1992) 188 11. Chang, Y.K., and H.C. Lim, "Experimental and Simulation Studies of MultivariableAdaptive Optimization of Continuous Bioreactors Using Bilevel Forgetting Factors," Biotech. Bioeng., 34, 577-591 (1989) 12. Semones, G.B., and H.C. Lim, "Experimental MultivariableAdaptive Optimization of the Steady-State Cellular Productivity of a Continuous Baker's Yeast Culture," Biotech. Bioeng., 33,16-25 (1989) 13. DiBiasio, D., H.C. Lim, and WA. Weigand, "Experimental Investi gation of Stability and Multiplicity of Steady States in a Biological Reactor," AIChE J., 27, 284-292 (1981) 14. Henson, M.A., "Dynamic Modeling of Microbial Cell Populations," Current Opinion in Biotechnology, 14, 460-467 (2003) 15. Bolie, V.W, "Coefficients of Normal Blood Glucose Regulation," J. Appl. Physiol., 16, 783-788 (1961) 16. Mustoe, L.R, and A.C. Croft, "Motivating Engineering Students by Using Modern Case Studies," Int. J. Eng. Ed., 15, 469-476 (1999) 17. Kitano, H., "Systems Biology: A Brief Overview," Science, 295, 16621664 (2002) 18. Parker, R.S., J.H. Ward, N.A. Peppas, and F.J. Doyle III, "Robust H 00 Glucose Control in Diabetes Using a Physiological Model," AIChE J., 46, 2537-2549 (2000) 19. Pannocchia, G., N. Laachi, and J.B. Rawlings, "A Fast, Easily Tuned, SISO, Model Predictive Controller, Proc. DYCOPS (2004) 20. Skogestad, S., and I. Postlethwaite, Multivariable Feedback Control, John Wiley & Sons, New York (1996) 21. Doyle, F.J. III, RS. Parker, and E.P Gatzke, "Process Control Modules: A Software Laboratory for Control Design," Prentice Hall International Series in the Physical and Chemical Engineering Sciences, PH PTR, Upper Saddle River, NJ (2000) 22. Yang, D. R, and J. H. Lee, "Process Control Education Software Using Java Applet," AIChE Annual Meeting, (2002) 23. Sorensen, J.T., "A Physiologic Model of Glucose Metabolism in Man and its Use to Design and Assess Improved Insulin Therapies for Diabetes," Ph.D. thesis, Department of Chemical Engineering, MIT, (1985) 24. Bergman, RN., L.S. Phillips, and C. Cobelli, "Physiologic Evaluation of Factors Controlling Glucose Tolerance in Man," J. Clin. Invest., 68, 1456-1467 (1981) 25. Lehmann, E.D., T. Deutsch, E.R Carson, and PH. Sonksen, "AIDA: An Interactive Diabetes Advisor," Comp. Meth. Frog. Biomed., 41, 183-203 (1994) 26. Agar, B.U., G. Biro!, and A. Cinar, "Virtual Experiments for Control ling Blood Glucose Level in Type 1 Diabetes," in Proc. Second Joint EMBSIBMES Conj., p. 2609 (2002) 27. Nomura, M., M. Shichiri, R Kawamori, Y. Yamasaki, N. Iwama, and H. Abe, "A Mathematical Insulin-Secretion Model and its Validation in Isolated Rat Pancreatic Islets Perfusion," Comput. Biomed. Res., 17, 570-579 (1984) 28. Lehmann, E.D., and T. Deutsch, "A Physiological Model of Glucose Insulin Interaction in Type 1 Diabetes Mellitus," J. Biomed. Eng., 14, 235-242 (1992) 29. Lenart, PJ., and RS. Parker, "Modeling Exercise Effects in Type 1 Diabetic Patients," Proceedings of the 15th IFAC World Congress on Automatic Control, Barcelona (2002) 30. Parker, R.S., E.P Gatzke, and F.J. Doyle III, "Advanced Model Pre dictive Control (MPC) for Type 1 Diabetic Patient Blood Glucose Control," in Proc. American Control Conj., Volume 5, pp. 3483-3487 (2000) 31. Parker, RS., F.J. Doyle III, and N.A. Peppas, "A Model-Based Algo rithm for Blood Glucose Control in Type 1 Diabetic Patients," IEEE Trans. Biomed. Eng., 46(2) 148-157 (1999) 32. Parker, RS., F.J. Doyle III, and N.A. Peppas, "The Intravenous Route to Blood Glucose Control," IEEE Eng. Med. Biol., 20, 65-73 (2001) 33. Lenart, PJ., and RS. Parker, "Glucose Control During Exercise in Type I Diabetic Patients," Proceedings of the Topical Conference on Bioinformatics and Genomics, AIChEAnnual Meeting (2001) 34. Florian, J.A. Jr., and RS. Parker, "Empirical Modeling for Glucose Control in Diabetes and Critical Care," Eur. J. Control, 11 (2005) 0 Chemical Engineering Education

PAGE 47

5 =i learning in industry ) llil-11111-~----==-------------=--This column provides examples of cases in which students have gained knowledge, insight, and experience in the practice of chemical engineering while in an industrial setting. Summer internships and co-op assignments typify such experiences; however, reports of more unusual cases are also welcome. Description of the analytical tools used and the skills developed during the project should be emphasized. These examples should stimulate innovative approaches to bring real-world tools and experiences back to campus for integration into the curriculum. Please submit manuscripts to Professor W.J. Korns, Chemical Engineering Department, Georgia Institute of Technology, Atlanta, GA, 30332-0100. THE ROLE OF INDUSTRIAL TRAINING IN CHEMICAL ENGINEERING EDUCATION MAMDOUH T. GHANNAM United Arab Emirates University Al-Ain, United Arab Emirates Cooperative Education Program I ndustrial training plays an important role in preparing engineering students to be future professional chemical engineers. The training offers a golden opportunity to acquire numerous technical and nontechnical skills that can not be obtained in a classroom environment. Some of the benefits of industrial training are[l 4 l: observing daily work activities firsthand in a real setting, gaining the ability to apply technical and theoretical knowledge to industrial problems, direct exposure to nontechnical skills such as oral and written communications, understanding the diversity of the chemical engineering industries, applying computer software programs to real industrial situations, teamwork experiences, time man agement and deadline objectives, getting familiar with the industrial environment to set and achieve future career goals, working effectively in a multidisciplinary environment, and boosting the student's self-esteem and confidence by gaining today's industrial skills. One such program is the cooperative education program. [l, 3 l Co-op education is based on rotation between schooling and full-time work periods. It connects undergraduate students directly with industry to gain strong fundamentals and in valuable insight into the chemical engineering profession, acquire technical knowledge, earn academic credits, and receive wages. Some universities offer co-op programs on an optional basis while others are mandatory. Among the colleges of engineering offering an optional program are the Univer sity of South Alabama,(Sl the University of Minnesota,l 6 l and the University of PittsburghYl Mandatory programs can be UNDERGRADUATE INDUSTRIAL PROGRAMS There is no classroom course that could simulate or replace the industrial experience gained from working with several operators in areal industrial environment. Numerous universi ties allow their students to gain industrial experience through a variety of programs. Mamdouh Ghannam received his Ph.D. from the University of Saskatchewan, Can ada, in 1991. He was an assistant professor at Cairo University, Egypt, from 1992-1995, after which he joined the University of Sas katchewan as a research engineer. Since 1998, he has been with the Department of Chemical and Petroleum Engineering at the United Arab Emirates University as an as sociate professor. His research interests are coating of Newtonian and non-Newtonian solutions, and interfacial and rheological properties of crude oil emulsions. Copyright ChE Division of ASEE 2006 Summer 2006 189

PAGE 48

found in universities such as Drexel,[ 8 l Ryerson,[ 9 l Toledo,l1l and Cincinnati. [l lJ Cooperative education programs have been well established in these universities for a long time. For ex ample, cooperative education was founded at the University of Cincinnati[l 1 lin 1906 and atDrexel Universityl 8 l in 1919. Most cooperative education programs off er financial benefits to students. The financial reward for students is usually based on location and type of task. Students' wages from their co-op jobs can even help finance their educations. At the University of Cincinnati,llll average salaries for co-op jobs are almost twice the tuition fees. Therefore, students do not have the burden of having a part-time job, giving them more time to devote to academics and other activities. At the University of Pittsburgh's[7l Department of Chemical and Petroleum Engineering, approximately 40-50% of the students take advantage of the available co-op program. According to a study completed by the co-op office in Cincinnati,llll 96% of graduating students acknowledged that the college program including co-op experience provided a better education than the traditional program without co-op. Most engineering schools that offer cooperative education have certain requirements for a student to participate in the program. For example, at the University of Minnesota,l 6 l the undergraduate student needs to be in good standing, have completed all program course requirements including fall semester of the third year, have completed at least five out of seven elective courses, and maintain a minimum GPA of 2.8. The co-op student will work full time for one year con tinuously in the industry. After successfully completing their co-op programs, students earn two credit hours per semester, which count toward technical elective courses. M.micipality 7% Central Lab H::>spital Glass A[urrinum Cerrent 7%. Fertilizer 17% Water Industrial Training Program Around 10 years ago, the College of Engineering at the United Arab Emirates University (UAEU) recognized the importance of industrial training and its crucial role in pre paring students for professional engineering. In 1995, a very committed program was established by the Unit of Industrial Training and Graduation Projects at the college. This pro gram is mandatory-i.e., part of the engineering education curriculum for all engineering students in the disciplines of chemical, civil, electrical, mechanical, and petroleum engineering. Students can be granted 15 credit hours after suc cessful completion of their industrial training. The following discussion focuses only on the chemical engineering industrial training program (ITP). The industrial training program at UAEU is selected in this article as an example to address its benefits and highlight areas that can be improved. PROGRAM LOGISTIC OF ITP To ensure that students have enough theoretical background to comprehend industrial training tasks, the program requires students have a minimum GPA of 2.5 and to have completed 114 credit hours. Any student with a lower GPA will be required to complete 126 credit hours (the total credit-hour requirement to earn a B.S. degree in engineering is 168). Each eligible student should prepare and submit a file to the industrial training unit that contains academic records, a one page resume, and a completed application form. The industrial training unit works with each candidate to find a training position with the available participating in dustrial partners. Students will be required to interview with 1% Cerarrics Oil 48% Figure 1. Participating industries. 190 Chemical Engineering Education

PAGE 49

potential companies to be placed. If a student is not accepted by one company, he/ she will have an opportunity with another company since the number of industrial partners exceeds the number of training candidates. The industrial training unit ensures that each candidate will receive placement in a relevant industry within the country or abroad. At the beginning of the training period, each student will re ceive a training-program schedule outlining the whole 16-week period. This schedule includes a weekly job description with tasks, academic advisor visits, and deadlines for reports and presentations. The schedule is established by the industrial supervisor in agreement with the student's academic advisor. The student is also advised of the evaluation process used to determine his or her grade. Student evaluation during this program is based on weekly progress reports submitted to the academic advisor (15% of total score), academic advi sor visits to the industrial site (15%), industrial supervisor's evaluation report (20% ), and final report and oral presentation (50%). The presentation and final report are assessed by an examination committee consisting of a college represen tative, a departmental faculty member, and two or three professional engineers. Upon successful completion of the ITP, student participants are eligible to register into the final academic year in chemical engineering. Failed participants must repeat the ITPprogram at another industrial site. With this in mind, most students are committed very seriously to the ITP. From the above ITP description, it is notable that the objec tives are similar to cooperative education in allowing students to gain industrial experience and learn the basics and fundaUmn Al Quw ain 1 % Fujairah Qata 1% Sharja 7% 11% mentals of the industry. The ITP, however, strictly considers students as trainees without financial compensation, not as employees with wages as in co-op programs. COMPANIES INVOLVED IN THE INDUSTRIAL TRAINING PROGRAM The number of participating companies in the ITP for the whole college increased tremendously from around 10 compa nies during the initial year of 1995-1996 to approximately 140 companies in the academic year of 2004-2005. This increase in participating companies, both locally and internationally, reflects industrial appreciation of the important role of the ITP. The total number of chemical engineering students who participated in the program from 1999-2005 is 198. The number of academic advisors involved in these activities changed from one semester to another based on the number of students. For example, the number of faculty members in volved in the first semester of 2004-2005 was four, and in the second semester of 2005-2006 it was seven. For the chemical engineering discipline, the specialties of the involved indus tries vary widely, as can be seen in Figure 1. The reported percentage represents the number of students completing the program for each industry during the period of 1999-2005. In addition, Figure 2 displays the various regions within the United Arab Emirates, with exception of France and Qatar, in which students of the UAEU carried out their training. These percentages also cover the period of 1999-2005. The following is a brief description of the ITP for a chemi cal engineering student who completed his training program at the oil service company Dowell Schlumberger, in Abu Dhabi, Al 1 Ain 6% 2% France Figure 2. Industrial locations by region. Summer 2006 191

PAGE 50

UAE. His training period was February to May 2005. The first eight weeks of his training included attending orientation, learning company safety standards and injury protection, performing laboratory tests, utilizing software, and learning equipment maintenance. During the second half of his training program, the student attended a company course on equipment and machine safety. He also had the opportunity to learn different aspects of the business such as the transfer of product. In the laboratory, the student was able to analyze materials and perform standard API tests including rheology and stability tests. The student was also able to use a Fann35 instrument to study the slurry flow behavior and measure the slurry viscosity at various RPM. One sample used consisted of water, antifoam, disper sant, retarder, and cement, with composition of 18.65, 1.0, 0.15, 0.2, and 80 wt%, respectively. The were involved in this study. The study showed that most of the employers' evaluations exceeded the 70% limit (the ac ceptable limit established by the UAEU training office) for all the criteria of a to k with a few exceptions. These excep tions occurred during the second semester of 2003-2004 and first semester of 2005-2006. The employers' evaluation was slightly less than the 70% limit for criteria c and j (the ability to design a system, component, or process, and knowledge of current engineering trends). Improvement of ITP If the faculty members are not significantly involved in industrial-site supervision, project choice, follow-up, and academic evaluation of student performance during the train ing period, the chances of an undergraduate student achieving industrial experience are not very good. [ll As student measured shear stress data versus shear rate and concluded the yield stress and viscosity for this sample was 9.83 lbr I 100 ft2 and 40.03 cp, respectively. Based on an academic advisor participating for the last seven years in the ITP, I feel improvements can be made to the current program. Implementing these ideas will enhance the overall perfor mance and outcomes of the ITP. Suggested I visited the student at the end of the third and 13th week to check on his train ing performance and diary, and discuss the feedback and recommendations of his industrial supervisor. The student's gradua tion project proposal was also discussed. a study performed by the Industrial improvements are: (a) Site Selection IMPACTS AND POTENTIAL IMPROVEMENT OF ITP Training Unit ... 86% of the students who completed the ITP To ensure the success of the training program for students, an academic advisor should be involved from the very beginning. The advisor needs to work closely with each student in selecting the appropriate training site since there can be several choices. Doing so guarantees matching the student with an industrial site that meets the program objecImpact By becoming a part of the ITP, students gain practical experience, technical knowl edge, confidence, time management skills, gained strong technical skills. tives, and avoids having students select an industry based on convenience rather than relevancy. teamwork capability, and a better understanding of what they've learned in class. Based on a study performed by the Industrial Training Unit at UAEU in the first semester of the 2003-2004 school year, 86% of students who completed the ITP gained strong technical skills. A total of 69 students hav ing completed their industrial training in 42 industrial sites participated in this study. This training mechanism enables students to define their career goals and provides an oppor tunity to find a permanent employment position. Additionally, this program is an excellent recruiting tool for participating companies. Employers will have the chance to train, evaluate, and select candidates for future job opportuni ties. A study was done by the industrial training unit at UAEU to survey participating industrial employers from the first semester of 2003-2004 to the first semester of 2005-2006. This study investigated the responses of the industrial employers on the trainees' performance with regard to the ABET2000 criteria (a to k criteria)Y 3 l A total of 75 industrial employers 192 (b) Industrial Field Visits The department should encourage and enforce a one-day field visit once a semester for all students, especially freshmen, to provide early exposure to different daily industrial activi ties_[3l Doing so allows students to start forming an opinion as to what type of industry they want to pursue. (c) Industrial Short Residence Due to the short period of the ITP (i.e., four months) in comparison with that usually spent by a trainee at other schools such as Pittsburgh,[7l Drexel,[ 8 l Cincinnati,[1 1 l and Min nesota,l6l participating departments should develop a yearly industrial event in which sophomore and junior students spend two to four weeks during the summer at a local industry in the residence city. Of course, each student should achieve certain limited objectives during this period and be required to provide a written report and oral presentation detailing his/her activities and experiences gained. To make this idea Chemical Engineering Education

PAGE 51

even more practical, one industrial site can be selected for each student. This requires the student to spend each summer break in different departments of an industrial site. (d) Nontechnical Skills Course To improve the nontechnical skills of undergraduate students in general and the ITP in particular, the college of engineering should develop a special course equivalent to two credit hours. The course would focus on enhancement of nontechnical skills such as communication, effective presen tation, technical writing, accessing information, judgment, software applications in industry, job interviews, resume building, and explaining technical information to nontechni cal customers. [ 4 12 J This course would be offered to all students during the second academic year, thus allowing enough time before the ITP stage. In this way, students will be well prepared for the ITP. They will be able to maximize the benefits and achieve all the expected objectives of the ITP. (e) Graduation Project Proposal One requirement for the successful completion of the ITP is the submission of a graduation project proposal within the final report. This proposal should reflect a valuable idea that attracted the trainee's attention during the training period. The proposal may then be selected as a graduation project by a faculty member within the chemical engineering department. Based on my own experience over the period of 1998-2005, I noticed problems associated with the graduation project proposals such as unclear or incomplete proposals and the lack of technical description and industrial data. It seemed students completed their proposals just to meet the require ment criteria, or did not complete them at all. The requirement to submit a graduation project proposal itself is a great idea. For greater benefit, however, it needs more involvement and commitment from each side. The fol lowing are some suggestions to strengthen the process of the graduation project proposal: 1. All graduation projects should be based on students' industrial proposals. 2. Clear, limited, and well-defined criteria for graduation projects should be established by the department at the beginning of each academic year. 3. Each group of three students will be assigned an academic advisor relevant to their industrial training site. 4. Forming a solid project proposal should be one of the main duties of trainees in coordination with the industrial supervisor and the academic advisor. 5. Both the industrial supervisor and academic advisor Summer 2006 should be involved with students to achieve this task by the end of the training period. 6. The proposal should cover the details of problem definition, industrial importance, negative impacts, alternative solutions, suggested solution and reasoning, and positive impacts. 7. A significant grade should be assigned to the project proposal. 8. The academic advisor should involve the industrial supervisor throughout the graduation project stages. 9. The industrial supervisor will be entitled to attend the final project exam and to receive a copy of the final report with all results and recommendations. (I) Academic/Industrial Interaction One of the main benefits of the suggested mechanism of the ITP and graduation project proposal is the strong link and interaction made possible between the industrial supervisor and the academic advisor. This relationship can help both parties advance their mutual interests, including: 1. Faculty member can make a strong connection with different industries leading to research and technical cooperation. 2. Industrial supervisor will have better access to the university environment for assistance such as techni cal recommendations, hiring new graduates, acquiring samples and data analysis, and participating in scien tific activities sponsored by the university. 3. Research cooperation between the two parties resulting in scientific publication reflects well on the image of the industrial partner. Due to the large number of activities requiring faculty involvement, these modifications will be better suited for departments with a small number of students. A larger number of students would require too much faculty time investment, negatively affecting research. SUMMARY It is clear that industrial experience in any program format can be beneficial to students. Some programs may have more benefits than others, but each provides valuable skills and training for our future engineers. It is important that we con tinue to study and evaluate these programs to make improve ments and adopt new ideas. What may work in one university may not work in another. By sharing the fundamentals of the programs, however, engineering colleges may find useful information to improve their current programs. After all, the main objective of any engineering education program is to produce the best possible engineers and to develop, enhance, and advance our society. 193

PAGE 52

REFERENCES 1. Huvard, G., "Make Summer Internship a Learning Experience," Chem. Eng. Ed., 32(1), 48 (1998) 2. Fricke, A.C., "From the Classroom to the Workplace: Motivating Students to Learn in Industry," Chem. Eng. Ed., 33(1), 84 (1999) 3. Bradburn, T., "Cooperative Education: A Key Link Between Industry and Engineers in the Making," Chem. Eng. Ed., 35(1), 58 (2001) 4. Bendrich, G., "Just a Communication Course? Or Training for Life After the University," Chem. Eng. Ed., 32(1), 84 (1998) 5. 6. 194 7. 8. 9. 10. 11. 12. Newell, J., D. Ludlow, and S. Sternberg. "Development of Oral and Written Communication Skills," Chem. Eng. Ed., 31(2), 116 (1997) 13. ABET Criteria for Accrediting Programs in Engineering in the United States, Engineering Accreditation Commission, Accreditation Board for Engineering and Technology (ABET) Inc., New York (1994) 0 Chemical Engineering Education

PAGE 53

5 =i laboratory ) ---11111---------=--------Validating THE EQUILIBRIUM STAGE MODEL for an Azeotropic System in a Laboratorial Distillation Column B.P.M. DUARTE, M.N. COELHO PINHEIRO, D.C.M. DA SILVA, AND M.J. MOURA lnstituto Superior de Engenharia de Coimbra 3030-290 Coimbra, Portugal D uring their fourth year of undergraduate studies, chemical engineering students at the Instituto Su perior de Engenharia de Coimbra (ISEC), Portugal, take a full laboratory course in unit operations and process control. The topics covered include evaporation, distillation, absorption in packed columns, solid-liquid extraction, dry ing, and the control of variables often found in industrial units (e.g., pressure, flow, level, and temperature) employing laboratory units or bench-scale kits. The course's basic aim is the practical demonstration of theoretical concepts taught in courses on process separation, chemical thermodynamics, and process dynamics at laboratory scale. It also provides students with experience in operating and controlling complex units. Regarding the work on distillation, the students are asked to validate the steady-state behavior of a laboratory unit used for separating an azeotropic mixture of aniline and water. The interest in this binary system arose from an intensive research program carried out in the chemical engineering department of ISEC in collaboration with a Portuguese company that produces aniline-with the aim of optimizing the aniline production section. In addition, this work also aims to vali date knowledge relating to the distillation of heterogeneous azeotropic mixtures, of which the aniline-water system is a simple and easily handled example. The conceptual basis used to describe the phenomena involved in a distillation column is the equilibrium stage model, which assumes thermodynamic equilibrium between Summer 2006 the species and perfect mixing in each phase and tray.[ 1 2 i But it is quite common for equilibrium not to be achieved in vapor-liquid-liquid dispersions arising from heterogeneous azeotropic mixtures. This is because of the occurrence of kinetically controlled phenomena, such as mass transfer, coalescence, and nucleation in liquid phases. Regardless of Belmiro Duarte is auxiliary professor in the chemical engineering department of lnstituto Superior de Engenharia de Coimbra, Portugal. He graduated in chemical engineering from the University of Coimbra, Portugal, in 1991. He received his Ph.D. in chemical engineering from the same university in 1995, and the M.Sc. degree in business management from Coimbra in 1998. His research interests include simulation, process optimization, and modeling. Maria Nazare Coelho Pinheiro is coordinator professor in the chemical engineering department of lnstituto Superior de Engenharia de Coimbra, Portugal. She graduated in chemical engineering from the University of Porto (FEUP), Portugal, in 1986. She received her Ph.D. in chemical engineering from the same university in 1994. Her research interests include mass transfer and hydrodynamic characterization in bubbling columns. Dulce Cristina Martins da Silva is researcher at CIEPQPF. She gradu ated in chemical engineering in 1996, and received her Ph.D. in chemi cal engineering in 2006, both from the University of Coimbra (FCTUC), Portugal. Her research interests include simulation, process optimization, and control. Maria Jose Moura is assistant lecturer in the chemical engineering department of lnstituto Superior de Engenharia de Coimbra, Portugal. She graduated in chemical engineering from the University of Coim bra (FCTUC), Portugal, in 1993 and received her M.Sc. in chemical processes from the same university in 1999. Her research interest is in biomaterials. Copyright ChE Division of ASEE 2006 195

PAGE 54

the mismatch of theoretical assumptions and real behavior, the equilibrium stage model is still used to represent the operation of distillation columns phenomenologically, and several research groups have presented data for three-phase distillation experiments to validate it. [ 36 l Thus, this work aims simultaneously to enable students to gain experience in operat ing and analyzing a distillation procedure, to use vapor-liquid equilibrium (VLE) prediction methods, and to contribute to the research community's efforts to validate the equilibrium stage model for the aniline-water system. Three four-hour sessions are required to complete the work. The first is devoted to a review of the basic theory behind mixture thermodynamics and methods for predicting activ ity coefficients for VLE, analysis of the column layout, and understanding its operation and control. Between the first and second sessions, the students develop an Excel workbook to obtain the VLE prediction for the aniline-water system at atmospheric pressure. Each group of three or four students is asked to use a different prediction method for the activ ity coefficients among UNIFAC, UNIQUAC, two-constant Margules, and van Laar,Pl then to compare the results with experimental data published in the literature for the aniline water system at atmospheric pres sure_ [s,Arren
PAGE 55

(Figure 1). When the valve is open, a glass stem is pulled magnetically from the seat and the liquid from the condenser is collected in the distillate receiver. When the valve is closed, the stem is pushed back, the seat is closed, and, when it is partially filled, the liquid overflows through a side tube and returns to the column (Figure 3). The reflux ratio value is set by employing two time preselectors on the control device, thus establishing the valve dead band. For the aniline-water binary system, however, the control of the distillation column based on the reflux ratio causes severe problems due to the formaTABLE 1 Density of Aniline-Water System for Different Temperatures!~ Appendix GI density (kg/m 3 ) temperature (C) aqueous phase organic phase 20 999 1023 30 997 1014 40 995 1006 50 991 998 60 987 989 70 982 982 2 4 5 6 5 Figure 2. Schematic representation of column trays. 1. Cavity for liquid inlet. 2. Cavity for liquid outlet. 3. Wall. 4. Reflux zone. 5. Downcomer to the lower tray. 6. Vapor flow trajectory. valve open from condenser glass stem ~u_,/ vmveseat/ Ltodistillate receiver from condenser to column valve closed Figure 3. Schematic detail of the solenoid valve operation: open and closed positions. Summer 2006 tion of two immiscible liquid phases in the condenser-the organic being denser than the aqueous at the temperature at which it leaves the condenser. The difference in density means that, when the valve opens after being closed for a while, a small volume of liquid retained in the valve seat (below the side tube level) is poured into the distillate receiver. This produces significant changes in the distillate composition since it is rich in the heavier phase (the organic phase). Table 1 indicates the density values for the aqueous and organic phases formed at different temperatures for the aniline-water system. [8, Appendix G] To overcome this drawback, instead of reflux ratio adjust ment, an on-off control scheme based on the boiling point limit in the upper tray of the column is employed, with the head temperature measured through a resistance thermometer (Fig ure 1). Whenever it exceeds the value selected at the control unit (98.7 C-the azeotropic temperature), the distillate flow rate is automatically interrupted, and the condensate starts to flow back to the column. Consequently, the temperature of the upper tray starts to fall, and when it reaches the preselected value the distillate starts to be collected again. This control scheme leads to small changes in the position of the solenoid valve, which is open most of the time. The feed stream is preheated to 99.3 C before entering the column in liquid state at atmospheric pressure. The feed stream is sampled for the quantitative determination of aniline concentration in the mixture. When the steady state is reached, the students collect the distillate and bottom products in graduated receivers for a period of 90 minutes. At the end, the distillate is trans ferred to a separatory funnel, and after the separation of the two phases, the volume of each layer is measured in graduated cylinders. The aniline concentrations in the aqueous phase of distillate, in the bottom product, and in the feed sample are quantitatively determined by spectrophotometry, after appropriate dilution. The absorbance of the solutions is measured at 279.5 nm in a UV-Vis spectrophotometer unit after setting the calibration line. The quantitative analysis of water concentration in the organic phase of the distillate is determined by titration with Karl Fisher reagent. For the range of flows used in the experiments, the column manufacturer indicates that the Murphree efficiency is 92%. The experimental confirmation of this value is not performed exactly because of the column configuration, which does not allow the extraction of liquid samples in consecutive trays. MODEL VALIDATION Aspen Plus v. 11.1 is used to validate the equilibrium stage model; the steady-state flows, compositions, and temperatures of the upper tray and bottom stream result197

PAGE 56

ing from the model solution are compared with laboratory data. The Radfrac module is chosen to describe the column unit since it is based on the rigorous solution of the equilib rium stage model for multistage vapor-liquid fractionation, steady-state operations_[ 9 10 J The model consists of a set of nonlinear algebraic equations, comprising the material bal ance (M) and thermodynamic equilibrium relation (E) for each component and tray, and the summation of mole fractions (S) and enthalpy balance (H) for each tray, generally called MESH equationsY 1 l This system can be augmented with the trays' hydraulic relations and pressure-drop profiles across the column when the unit geometry is known. The broader generality of the Newton-Raphson algorithm led this to be chosen to solve Radfrac module in rating mode. The Newton-Raphson al gorithm implemented is based on the classic Naphtali and Sandholm algorithmY 2 l First, the number of equations and variables resulting from unit modeling is reduced through the condensation of mole fractions, liquid, and vapor flows into new variables representing component molar flows. Next, the complete set of variables is ordered, and the resulting algebraic equation system solved iteratively by employing a Newton-type algorithm. The convergence is checked after each iteration by comparing the sum of squares of all vari ables, conveniently weighted by scale factors, with a tolerance defined as a function of the number of degrees of freedom the system involves and inlet flows. An azeotropic convergence algorithm is chosen to handle the current binary system that forms disregarded due to the small flows involved in the operation. The characteristics of the feed stream, including its tempera ture, pressure, and molar composition, are also entered into Aspen Plus. It is considered that there is no sub-cooling in the condenser and the Murphree stage efficiency is set to 92%. The Appendix at the end of this paper presents the Aspen Plus Input Summary file for a successfully converged model. RESULTS The first step in analyzing the results is to compare the VLE data calculated by the Aspen Properties module with the prediction obtained by students. Next, both must be validated with experimental data. The diagram in Figure 4 shows the agreement between the VLE data predicted using the UNIFAC method and the experimental data published in the literature. [S, Appe nd ix GJ The system presents a heterogeneous azeotrope at 98. 7 C and 0. 044 of aniline mole fraction, where three phases are in equilibrium: a vapor phase and two liquid phases [an organic phase with30.3% (mole/mole) water, and an aqueous phase with 98.6% water]. This happens because the vapor-liquid envelope overlaps the liquid-liquid envelope, as illustrated in Figure 4. This task allows students to understand that for heterogeneous azeotropes the vapor formed during boiling has the same composition as the overall liquid, but the three phases in equilibrium have distinct compositions, contrary to what "" .. a minimum-boiling azeotrope in the region of low aniline concentrations. The operating 200~-~-~--~-~--~,,conditions, including the molar flow of distillate stream, the heat consumed in reboiler, the stages at which streams enter/leave the unit, the pressure-drop profile across the column, additional information regarding the condenser operation, and the characterization of the second liquid phase-formed essentially by anilineare introduced into Aspen Plus. Molar flows are set equal to the experimental steady-state values, and the pressure drop is Figure 4. Comparison of experimental data from literature 18 Appe nd ix GJ vs. VLE data determined through UNIFAC for aniline-water system. 198 180 -160 140 8 120 Q) D.. 100 E Q) I80 60 40 20 0 0.1 + + 0.2 ........ 0.3 -+o siss o.'516 o 9&5 o. : 9? o 975 o se o. o 99 o.m 1 Mol~l'r.-c:lionQt~O T-x (P=1 atm) UNIF AC prediction T-y (P=1 atm) UNIF AC prediction O T-x (P=1 atm) literature data (L-\/) + T-y (P=1 atm) literature data + T-x (P=1 atm) literature data (L-L) 0.4 0.5 0 6 0.7 0.8 0.9 Molar fraction of H 2 O Chemical Engineering Education

PAGE 57

happens for homogeneous azeotropes, where the liquid and vapor formed have the same composition. The VLE prediction methods using UNIQUAC, two-con stant Margules, and van Laar activity-coefficient models require binary parameters for the aniline-water system, which students estimate by employing the following procedure and theoretical basis. For a binary system containing two liquid phases and one vapor phase in equilibrium, the fugacities of each compound in each of the phases are equal. That is: vapor pressure of the pure component i at temperature T; P is the pressure; and yi is the mole fraction of component i in the vapor phase. The subscript s stands for saturated liquid phases. According to the phase rule, a binary system with three phases in equilibrium has just one degree of freedom, which means that by fixing the pressure (atmospheric pressure) the system becomes determined. Setting the activity coefficient model, the functional forms of ya and y w can be explicitly written, and the preceding equations lead to 'Yorgp* org as a xas aqp* aq y as a xas P Ya "'org (A A xorg) xorg = aq (A A xaq) xaq (1) I as a,w' w,a' as as Yas a,w' w,a' as as '-v-' fugacity of fugacity of fugacity of aniline (a) in the aniline (a) in the aniline (a) in the organic phase aqueous phase vapor phase and 01-gp* org y ws wxws aq p* aq y ws wxws P Yw (2) fugacity of fugacity of fugacity of water (w) in the water (w) in the water (w) in the organic phase aqueous phase vapor phase when the vapor phase behaves like an ideal gas mixture. The expressions rrg and y~ 4 represent the activity coefficients of component i in the organic and aqueous phases, respectively; xorg and Xaq are the molar fractions of component i in the 1 1 organic phase and aqueous phases, respectively; Pi is the TABLE2 Values of Binary Interaction Parameters for UNIQUAC, Two Constant Margules, and van Laar VLE Prediction Models binary interaction parameters model A,w (J moJ-1) A (J moJ-1) w,a UNIQUAC 1524 575 two-constant Margules 6247 5836 van Laar 13164 4798 (3) and "'org (A A xorg) xorg I ws a,w' w,a' ws ws (4) aq (A A xa 4 )xa 4 "fws a,w' w,a' ws ws where A and A are the binary interaction parameters of the model ch;sen fo;"the aniline-water system. When the liquid liquid equilibrium data is available, the fractions x~;g, x:;, x';;, and x:~ can be used to evaluate the two parameters A and A Solubility data of aniline in water and of water in"~line 7or a temperature range of 20 C to 100 C can be found in the reference/ 8 Appe nd ix GJ thus allowing students to estimate mutual solubility values at the azeotrope temperature (98.7 C). The parameters obtained are used to calculate activity coefficients for subsequent vapor-liquid equilibrium calculations in the regions of 0
PAGE 58

I!) ... 0 Block B1 (RadFrac) Profiles Compositions 0 ... 0 0 I!) "' 0 d "' 0 0 I!) N 0 0 I!) a 0 a 0 ---------\ "" "' I I I I I I I I I -=Liquid (mole frac) ANILINE Vapor (mole frac) ANILINE ,c ::_ C lo nd e fn se --lr -+----1--~_j__j__j__j____j_ _j__--+-----'l == + '--= <;:'-..=:>-........__J ,.__----6--_ ? __ == __ iT \ 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 Stage Figure 5. Composition profiles of aniline across the column. between the model's prediction and experimental data, thus validating the equilibrium stage model for the aniline-water system. Once the heat duty in the reboiler is introduced into the model, in rating mode, it is able to determine the reflux ratio, which just represents an average value since, with the control strategy implemented, it varies discretely between 0 and oo, depending on the valve position. Indeed, such a value represents a possible set-point value, if the control scheme was based on the reflux ratio. Using the Profiles form, it is possible to view the results from Radfrac as compositions, temperatures, and flow rates for each column tray. Figure 5 shows the profiles of aniline composition in both liquid and vapor phases across the col umn. As expected, the aniline concentration in both liquid and vapor streams increases from the bottom to the top of the col umn, with the aniline mole fraction in the vapor always being greater than that in the liquid since the feed stream is located to the right of the azeotropic point of the VLE diagram (see Figure 4). The composition profiles provide strong evidence that the number of trays is over-projected for the experimental conditions tested. Indeed, in some of the stages the enrich ment of vapor phase in aniline is quite small, thus leading to the conclusion that the column can successfully separate a higher flow of feed stream (with a similar composition) with the same efficiency, provided the heat supplied to the reboil er increases and flooding does not occur. Figure 6 shows that the liquid and vapor flows across the column are approximately constant in the enrichment and 200 0 L!) ..... Block COL 1 (RadFrac) Profiles TPFQ Liquid flow Vapor flow Stage Figure 6. Liquid and vapor flows across the column. stripping zones. Since stage 11 (12 in Figure 6) is the feed stage and the state of the stream is saturated liquid, the liquid flow rate in the stripping zone is increased by an amount equal to the feed flow rate. These conditions enable the McCabe Thiele graphical construction[ 13 J to be used to estimate the number of theoretical trays required to perform the separation. Typical values achieved by students are about 14 trays plus the condenser and the reboiler. Chemical Engineering Education

PAGE 59

CONCLUSIONS Operating a laboratory distillation column is a good experi ment for demonstrating the application of some concepts of unit operations, vapor-liquid equilibrium prediction, and process simulation. The experiment described in this paper embraces a wide range of topics including process control, chemical analysis, and numerical methods for handling rigorous distillation models. It also enables students to gain experience in operating and controlling a distillation unit. Moreover, the results provide the research community with sufficient evidence to support the validation of the equilibrium stage model for the heterogeneous azeotropic system formed by aniline and water. The results show that the column is too large for the experimental conditions tested, and additional knowledge regarding its behavior can be acquired if the inlet flow is increased and digital temperature meters are installed in each tray to validate the temperature profile. But the ex periment might be successfully applied to other azeotropic systems, such as a benzene-mononitrobenze-water mixture. REFERENCES 1. Seader, J.D., and E.J. Henley, Separation Process Principles, John Wiley & Sons, New York (1998) 2. Perry, R.H., and D. W Green, Perry's Chemical Engineer's Handbook, 7th Ed., McGraw-Hill Companies, New York (1997) 3. Kovach, J.W., and WD. Seider, "Heterogeneous Azeotropic Distilla tion: Experimental and Simulation Results," AIChE Journal, 33, 130 (1987) 4. Cairns, B.P, and I.A. Furzer, "MulticomponentThree-PhaseAzeotropic Distillation 1. Extensive Experimental Data and Simulation Results," Ind. Eng. Chem. Res., 29, 1349 (1990) 5. Herron, C.C., B.K. Kruelski, and J.R. Fair, "Hydrodynamics and Mass Transfer on Three-Phase Distillation Trays," AIChE Journal, 34, 1267 (1988) 6. Mueller, D., W Marquardt, T. Hauschild, G. Ronge, and H. Steude, "Experimental Validation of an Equilibrium Stage Model for Three Phase Distillation," in Distillation and Absorption '97, Vol. 1, 149 (1997) 7. Poling, B.E., J.M. Prausnitz, and J.P O'Connell, The Properties of Gases and Liquids, 5th Ed., McGraw-Hill, New York (2001) 8. Coulson, J.M., R.K Sinnott, and J.F. Richardson, Coulson and Richardson's Chemical Engineering, Vol. 6: Chemical Engineering Design, 3rd Ed., Butterworth-Heinemann, Oxford (1999) 9. Aspen Tech, Aspen Plus 11.1. Unit Operation Models, Aspen Technol ogy, Inc., Cambridge, MA (2001) 10. Seider, W.D., J.D. Seader, and D.R. Lewin, Process Design Principles. Synthesis, Analysis, and Evaluation, John Wiley & Sons, New York (1999) 11. Biegler, L.T., I.E. Grossmann, andA.W Westerberg, Systematic Meth ods of Chemical Process Design, Prentice Hall PTR, Upper Saddle River, NJ (1997) 12. Naphtali, L.M., and D.P Sandholm, "Multicomponent Separation Calculations by Linearization," AIChE Journal, 17, 148 (1971) 13. McCabe, W, J.C. Smith, and P Harriott, Unit Operations of Chemical Engineering, 6th Ed., McGraw-Hill, New York (2001) APPENDIX AspenPlus Input Summary file Input Summary created by AspenPlus Rel. 11. 1 at 18: 15:48 Summer 2006 Sat., Jan. 28, 2006. DYNAPLUS DPLUS RESULTS=ON TITLE 'Coluna do Isec' IN-UNITS SI DEF-STREAMS CONVEN ALL SIM-OPTIONS IN-UNITS ENG SIM-OPTIONS NPHASE=3 ATM-PRES=!. PARADIGM=EO ACCOUNT-INFO USER-NAME= "BELMIRO DUARTE" DATABANKS PURE! I / AQUEOUS / SOLIDS / INORGANIC I & NOASPENPCD PROP-SOURCES PUREll / AQUEOUS / SOLIDS / INORGANIC COMPONENTS WATERH2O/ ANILINE C6H7N-1 FLOWSHEET BLOCK Cl IN=FEED OUT=DESTIL RESID PROPERTIES UNIFAC PROPERTIES NRTL / UNIQUAC PROP-DATA NRTL-1 IN-UNITS SI PROP-LIST NRTL BPVAL WATER ANILINE 2.238300000 362.5433000 .3000000000 0.0 & 0. 0 0. 0 372.1500000 441.1500000 BPVALANILINE WATER-.8969000000 509.3646000 .3000000000 & 0.0 0.0 0.0 372.1500000 441.1500000 PROP-DATA UNIQ-1 IN-UNITS SI PROP-LIST UNIQ BPVAL WATER ANILINE .6554000000-168.0642000 0.0 0.0 & 372.1500000 441.1500000 BPVALANILINE WATER-.4676000000-172.2809000 0.0 0.0 & 372.1500000 441.1500000 STREAM FEED SUB STREAM MIXED TEMP=99.3 PRES=! & MASS-FLOW=2.427 MAXIT=l00 MASS-FRAC WATER 0.977 / ANILINE 0.023 BLOCK Cl RADFRAC PARAM NSTAGE=32ALGORITHM=STANDARD EFF=MURPHREE & INIT-OPTION=STANDARD MAXOL=l00 TOLOL=0.0001 JMETH=INIT & LL-METH=GIBBS NPHASE=2 DAMPING=NONE COL-CONFIG CONDENSER=TOTAL REBOILER=KETTLE FEEDS FEED 12 ON-STAGE PRODUCTS RESID 32 L / DESTIL 1 L P-SPEC 1 1. / 2 1. COL-SPECS QN=335. MOLE-D=17.72 14 SC-REFLUX OPTION=0 STAGE-EFF 10.92/20.92 / 3 0.92 / 4 0.92 / 5 & 0.9216 0.92 / 7 0.92 / 8 0.92 / 9 0.92 / 10 & 201

PAGE 60

202 0.92 / 11 0.92 / 12 0.92 / 13 0.92 / 14 0.92 / & 15 0.92 / 16 0.92 / 17 0.92 / 18 0.92 / 19 0.92 / & 20 0.92 I 21 0.92 / 22 0.92 / 23 0.92 / 24 0.92 / & ~Q92/~Q92 / TIQ92 /~Q92/~Q92/ & 30 0.92 / 31 0.92 / 32 0.92 T-EST 1 371.1 / 2 372.3 / 3 372.6 / 4 372.6 / 5 & 372.6 I 6 372.6 / 7 372.6 / 8 3 72. 6 / 9 372.6 / & 10 372.6 / 11 372.6 / 12 372.6 / 13 372.8 / 14 & 372.9 I 15 373. / 16 373. / 17 373. l / 18 373.1 / & 19373.l / 20373.l I 21373.l / 22373.2 / 23 & 373.2 I 24 373.2 / 25 373.2 / 26 373.2 / 27 & 373.2 I 28 373.2 / 29 373.2 / 30 373.2 / 31 & 373.2 L-EST 13.938E-006 / 23.98E-006 / 3 3.984E-006 / 4 & 3.985E-006 / 5 3.985E-00 6 / 6 3.98 5 E-006 / 7 & 3.985E-006 / 8 3.985E-006 / 9 3.985E-006 / 10 & 3.985E-006 / 11 3.985E-00 6 / 12 4.3 64E005 / 13 & 4.366E-005 I 14 4.367E-005 / 15 4.368E-005 / 16 & 4.369E-005 / 174.37E-005 / 184.37E-005 / 19 & 4.37E-005 I 20 4.37lE-005 / 21 4.371E-005 / 22 & 4.371E-005 / 23 4.371E-005 / 24 4.371E-005 I 25 & 4.371E-005 I 26 4.371E-005 / 27 4.371E-005 / 28 & 4.371E-005 I 29 4.371E-005 / 30 4.371E-005 / 31 & 4.371E-005 V EST 1 0. / 2 8.86E-00 6 / 3 8.902E-006 / 4 & 8.907E-006 / 5 8.907E-006 / 6 8.907E-006 / 7 & 8.907E-006 / 8 8.907E-00 6 / 9 8.907E-006 / 10 & 8. 907E-006 / 11 8. 907E-00 6 / 12 8. 90 7 E006 / 13 & 8.927E-006 / 14 8.945E-00 6 I 15 8.959E-006 / 16 & 8.971E-006 / 17 8.98E-00 6 / 18 8.986E-006 I 19 & 8.99E-006 I 20 8.993E-00 6 / 21 8.9%E-006 / 22 & 8.997E-006 / 23 8.998E-006 / 24 8.999E-006 I 25 & 8.999E-006 I 26 8.999E-006 I 27 9E-006 I 28 9E-006 / & 29 9E-006 I 30 9E-006 / 31 9E-006 BLOCK-OPTION FREE-WATER=NO EO-C ONV-OPTI SENSITIVITY S-1 DEFINE Zl BLOCK-VAR BLOCK=Cl VARIABLE=RR SENTENC E=RESULTS DEFINEZ2BLOCK-VAR BLOCK=Cl VARIABLE=VRATE & SENTENCE=PROFILE ID1=2 TABULATE 1 "Z l TABULATE 2 "Z 2" VARYBLOCK-VARBLOCK=Cl VARIAB LE=Q NSENTENCE--COL SPECS RANGE LOWER= 250 UPPER="350" INCR="l0" SENSITIVITY S-2 DEFINE Zl BLOCK-VAR BLOC K=Cl VARIABL E =RR SENTENC E=RESULTS DEFINE Z2 BLO C K-V AR BLOCK=Cl VARIABLE=VRATE & SEN TE NC E=PROFIL E ID1=2 TABULATE l Zl TABULATE 2 Z2 VARY MASS-FLOW STREAM = FEE D SUBSTREAM=MIXED COMPONEN T=A N ILI NE RANGE LOWER= 5.0E-6 UPPER ="4.0E-5" INCR="2.5 E-6 CON Y-OPTIONS PARAM TEAR-METHOD=NEWTON SPEC-LOOP=INSIDE STREAM-REPOR MOLEFLOW MASSFLOW PROPERTY-REP PCES NOPARAM -PL US 0 C hemical Engineering Education

PAGE 61

5 =i classroom ) ---11111----------A TIRE GASIFICATION SENIOR DESIGN PROJECT That Integrates Laboratory Experiments and Computer Simulation BRIAN WEISS AND MARCO J. CASTALDI Columbia University New York 10027 T he Accreditation Board for Engineering and Tech nology (ABET) requires that students in accredited engineering programs complete certain requisites to graduate. One constituent is engineering design, "the process of devising a system, component, or process to meet desired needs,"[ll specifically Criterion 4--Professional Component, which is particularly focused on a major design project in corporating appropriate engineering standards and multiple realistic constraints. To fulfill the program, the Department of Earth and Environmental Engineering at Columbia Uni versity allows undergraduate seniors the opportunity to work independently under the supervision of a faculty advisor. The faculty advisor's purpose is to guide the student's activities and ensure progress. The student typically decides on a topic area, such as waste to energy, in consultation with a faculty member. Once the overall project area is identified, a rigorous task plan and schedule are given to the student to begin the design effort. Clearly, this plan must be consistent with any guidelines outlined by the department (e.g., midsemester report, final presentation). In this case, Figure 1 (next page) details the efforts to be undertaken by the student and that are aligned with the department's requirements of a fall and spring term presentation (not shown) and the final report. One recent student's project involved the conversion of waste tires by thermal treatment to either energy generation or chemical synthesis. The field of study was selected based on the expertise of the mentoring professor and capabilities of Summer 2006 the laboratory. Weekly meetings between the student and pro fessor were arranged and a schedule of activities was drafted to facilitate progress toward the design. The first semester (fall) concentrated on researching the topic, creating a design, and justifying the initial feasibility of the design with ecoMarco J. Castaldi is an assistant pro fessor in the Earth and Environmental Engineering Department at Columbia University. He received his B.S. ChE from Manhattan College and M.S. and Ph.D. ChE from the University of California, Los Angeles. Prior to joining Columbia Uni versity, he worked in industry for seven years researching and developing novel catalytic reactors. His teaching interests lie in thermodynamics, combustion phe nomena, and reaction engineering. His research is focused on beneficial uses of CO 2 in catalytic and combustion environments, waste-to-energy processes, and novel extraction techniques for methane hydrates. Brian Weiss received his B.S. from the School of Engineering and Applied Sciences at Columbia University in spring 2005. The work with tire gasifi cation culminated his education in the Department of Earth and Environmental Engineering. Currently, he looks forward to pursuing similar projects in efficient chemical conversion as a chemical engineering graduate student at the University of California, Berkeley. Copyright ChE Division of ASEE 2006 203

PAGE 62

Plan for Fall Semester Plan for Spring semester (Due dates in bold): 0) Motivation (September 22) a. Market demand b. Environmental/Economic/ Global impacts/benfits 0) Review of tire pyrolysis and combustion characteristics. (Jan 14) a. Literature search c. Type of waste 1) Previous work (October 15) 1) Materials list for Prototype (Jan 21) a. Patent office (Next Week) (www.USPTO.gov) 2) Lab Work (Feb 25) b. Textbooks c. General literature (journals) d. Company info a. Combustion TGA, micro GC (Jan 28) 1. Test Plan (Jan 14) 11. Kinetics 2) Identify most promising processes 111. volatile fraction; CO 2 /CO; optimum atmosphere (November 15) a. Product yields b. Feedstock accessed c. Prototype built d. Economics e. Energy balance 3) Understand (December 10) a. Flow diagram b. Gasification Same apparatus (Feb 25) 1. Test Plan (Jan 28) 11. w/ H 2 0 (g + l); CO 2 CO, arr b. Chemistry and reactions 111. GC/MS product analysis 1v. Effect of Temperature; optimum atmosphere; Figure 1. The student and professor drafted a schedule of activities at the beginning of each semester. The schedule followed the guidelines given by the academic department. c. Thermodynamics d. Pitfalls e. Lab work 4) Improve (Ongoing) a. Imagination b. Lab work nomic and thermodynamic calculations. The second semester (spring) focused on executing laboratory work necessary to provide data for input to a theoretical modeling of an overall system. Typically an industry-accepted modeling package is used to conduct simulation and analysis of an overall process. The package used during this project was Aspen Plus v. 12.1 by Aspen Technology, Inc. Other design projects in which students combined ex perimental work and computer simulation have demonstrated positive results. In such projects, students can gain a profound understanding of industry in a more stimulating setting than 204 3) Computer simulation (Apr 8) a. Aspen Plus b. Use TGA data c. Team with Kimberly? 4) Prototype build (Apr 29) a. Wire and Cellophane b. Smoke experiments 5) Final Report (May 2) Notes to plan: The prototype build might run overtime depending on the complexity of the project. The priorities will be physically observing the design, i.e. the lab work and the prototype build. The main benefit of a computer simulation could be a comparison with the models from last semester so it need not be the focus. a lecture class. In one project, students were instructed to use theory and experimentation to execute a boric acid dehydra tion process. The result won recognition from education competitions and Borax Europe, Ltd., an industry leader.[ 2 l In another case, students were prompted to design, fabricate, and test single-component, mechanical productsYl In both projects, an emphasis was made to complete the curriculum within one or two semesters. In the latter, computer aided design was necessary for compressing the work into the desired time frame. The success of each example project lay in the ability to organize a broad scope of activities into Chemical Engineering Education

PAGE 63

manageable blocks. The first step of the present design project was for the student and professor to propose a schedule of activities ensuring the student gained exposure to all aspects of the design experience. Shown in Figure 1, this type of work plan allows the student to understand the usefulness of preliminary calculations in guiding subsequent work. All segments of the plan were executed. The prototype build was attempted but not completed, however, because unanticipated opportunities to present the work arose and demanded more effort be devoted to refining what was already accomplished. The remaining sections of this paper are taken from work done by the student during the two-semester design project. This paper is an example of how to integrate environmental issues such as pollution prevention, reuse, and recovery into the design experience. While this particular project did not use life-cycle assessment methods and tools, it could be a very worthwhile effort to employ them to explore the outcomes for waste tire and other waste-to-energy processes. BACKGROUND WORK The project began with a literature survey of academic papers, government programs, and business efforts. Currently in the United States, 290 million waste tires are generated an nually. Additionally, nearly 300 million tires(~ 6 million tons) reside in environmentally unsound stock piles. Accordingly, government impetus has created a market for waste tires that currently awards tipping fees to users between $50-100 per ton_[ 4 l Discarded tires have applications as a co-combustion fuel and ground rubber fills in construction materials. Only 9% of the tires currently generated in the United States go to landfills. Current applications for tires are both economi cally practical and regulated by national standards, but initial research showed opportunities for improvement. [SJ Most tires are used as supplementary feed in processes for which they were not explicitly designed. Since many advantages can be realized from a method specifically adapted to the feed, the purpose of the project became to design a system specific to scrap tires. Since tires possess several distinct qualities as a fuel, ther mal processing may offer a broad range of opportunities to improve existing practices. In the United States, three facili ties have been established to convert exclusively scrap tires to electricity: Exeter Energy in Connecticut, Modesto Energy in California, and the Ford Heights facility in Illinois. Each facility was built in the early 1990s with a capacity to handle 8-10 million scrap tires per year on conventional equipment to produce 25-30 MW of electricity. Only one remains in business, however, indicating that their profitability is, in the current environment, only marginal.[ 6 l It is anticipated that, with rising energy and landfill prices, an innovative approach may enable a more successful business. Summer 2006 PRELIMINARY DESIGN AND ECONOMICS Before drafting a reactor design, the economics and thermo dynamics of the process were investigated using information found in the literature. These steps were intended to provide a "first-cut" analysis and set the boundaries of feasibility and profitability. A process was proposed with unit operations including: a standard thermal reactor; an electricity-produc ing system with boiler, turbine, and generator; a gas clean-up system consisting of an electrostatic precipitator, a scrubber and a stack; and auxiliary capital (pipes, pumps, etc.). Cost ing estimates were obtained from a chemical engineering textbook; size, efficiency, and inflation were incorporated into the model. [ 7 l Tires consist of a mixture of rubber polymer (C 5 H 8 )n and carbon black with added fillers such as light oils, fibers, trace metals (zinc), and a steel wire belt. A chemical and elemental analysis of tires was taken from a published source and showed a combustion enthalpy of 35 kJ/ g. [SJ The economic model was created for a process scaled to 10 million tires per year (164 thousand tons). The results showed that 16.7% of the tire's enthalpy of combustion could be converted to an output of 28.6 MW of electricity. The revenue generated by the system included tipping fees from the tires ($100 per ton) and electricity sales ($0.05 per kWh). At an interest rate of 5%, the revenue after a 40% tax totaled to $21.5 million per year. The net present value was $201.5 million, which annualized to $8.4 million per year. The profits were $13 .1 million per year indicating an internal rate of return of 19%. These results include all relevant parts such as working capital, debt servicing, permitting, and cit ing. The experiences of the other facilities combusting tires support the estimate. Although the process proves profitable, the return is lower than most investors would prefer for new technology, indicat ing that a technological breakthrough is necessary to secure the business. Several reactors for converting waste-to-energy were researched including fixed beds, moving beds, fluidized beds, and rotary kilns_[ 9 10 J A fixed-bed type reactor was select ed for its cost effectiveness, ease of use, and appropriateness to the feed. Additionally, it was proposed that the end product would be syngas (primarily CO and H 2 ), which was thought to enable more controllable conditions. Syngas can be created by reforming the tires with CO 2 and Hp. Because these reac tions are endothermic, a heat source is required. Thus it was proposed that the combustion of tires with air could provide that heat to drive the reactions. Using sewage sludge as the water source could minimize costs, extending the scope of the design to a truly novel integrated waste converter. The ideas behind the design evolved by employing the principles of process intensification to existing technologies. Process intensification basically is the reduction of process volumes by combining and consolidating multiple unit operations into one physical unitY 1 l 205

PAGE 64

DESIGNING THE REACTOR The reactor was sketched in StudioTools (Alias). Antici pated flow patterns are shown in Figure 2. Scrap tires and stoi chiometric air are fed separately into the combustor through an annular pipe. Tires fall and combust on a ceramic plate similar to a grate-type combustion system. The wall of the combustor has two layers: an outer impermeable steel barrier and an inner screen within which primary air emerges. This concept was adapted from the gas turbine industry to maintain moderate metal temperatures of the combustor. The purpose of the double wall is to maintain a unidirectional flow pattern that minimizes the amount of hot gases impinging on the wall. Secondary air actively cools the struts supporting the ceramic plate and assists the combustion. The combustion product leaves the combustor from below and enters the gasifier. The material to be gasified falls from the top in a counter-current flow to the hot gases. Heat transfers to the gasifier across the dividing wall of the combustor and via the enthalpy contained in the combustion product stream. The syngas produced is extracted through a pipe from the top or side. The ash falls to the bottom of the reactor where it is collected and removed using standard equipment. The reactor has similar attributes to a previously disclosed design for a combustion-gasification system of wood chipsY 2 l Nonetheless, because the system proposed for this project was conceived independently, there are considerable differences to make this design unique. The design possesses capabilities beyond existing technology because it is prepared to handle fuel with a large heating Figure 2. A rendering of the proposed reactor. Tires and air flow in an annular pipe and combust on a ceramic plate in the center. The combustion product exits the inner chamber from below and mixes with additional water, sludge, and tires. The syngas is extracted from a pipe in the side or on top and the ash falls to the bottom. Baffles ensure the flow patterns run along the planned routes. 1 Air (combus on & liner cooling) value and high ash content, such as tires, while beneficially using sludge material. MATERIAL AND ENERGY BALANCES Material and energy balances were performed for both the combustor and the gasifier. For a reactor that consumes 10 million tires per year (164 thousand tons), combusting 30% in stoichiometric air and gasifying the remainder with 87,600 m 3 of water per year, the results of the calculations showed the syngas to consist of 18.9% H 2 16.6% CO, 6.0% Hp, 8.4% CO 2 and 49.9% Nr The total energy output is 37 MW of sensible heat and 103 MW of chemical energy. The tempera tures of the combustor wall and gasifier are 1,040 C, and 614 C, respectively. A flame temperature for the combustion of the tires was calculated to be 1,469 C. Based on a material residence time of one hour, a combustor size of 27 m 3 was calculated with a 2.8 m base diameter. The gasifier was sized to allow a moderate flow rate of the syngas produced, while enabling most of the large ash particles to settle. This led to a unit of 77 m 3 that is 5.0 m diameter and 3.8 min height. The success of the process will be determined by the ability to control the material and heat flow. Because com bustion temperatures can run higher than the limitations of most metals, the wall must be kept cool, which will be ac complished by air flow augmented by the fresh tire/sludge mixture entering the gasifier. The double wall will create a unidirectional air-flow pattern and minimize the impingement of hot combustion gases onto the wall. Both the temperature Water/Tire Double Walled Combustion Unit (Tire+ 0 2 CO 2 + H 2 0 + heat) (C + H 2 0 + CO 2 + heat CO+ H 2 ) 2 Air (combustion & strut cooling) 206 Chemical Engineering Education

PAGE 65

and the syngas quality will be regulated by the composition of the feeds: adding more tires produces more energy and higher temperature whereas more water yields more hydrogen in the product stream. LABORATORY WORK While most, and probably all, parameters could be ob tained from the literature, one of the advantages of having individual design projects is the ability for students to get hands-on experience working in a laboratory. This enables them to generate pertinent data needed for their design. It also forces students to think of the experimental outcomes before doing the work, thus preparing them to develop practical test methods for efficiently generating data. To enable realistic engineering, basic thermodynamic and kinetic parameters of tires were required. Some of these parameters were obtained from equipment readily available in the faculty advisor's combustion laboratory. An oxygen bomb calorimeter (Parr) yielded the heat of combustion of tires. The bomb calorimeter adiabatically combusted the tire sample at constant pressure. The temperature rise of a water bath correlated to the enthalpy of combustion of the tires. The enthalpy of combustion of tire was determined to range from -33.37 to -36.33 kJ/g which was consistent with established data.[ 13 14 l Kinetic informa tion was obtained from thermogravimetric analysis (TGA). A Netzsch TG 409 PC instrument was used to record the mass loss of a sample as the temperature was increased at a constant rate. Constant flow rates (100 mL/min) of air (20% 0 2 ; 80% Nz) and inert purge gas (20% CO 2 ; 80% Nz) flowed over the sample. The air enabled evaluation of combustion parameters while the inert atmosphere was selected to re semble the gasification zone. A plot of the fractional weight loss, a, versus temperature, T, showed that the sample mass decreased as temperature increased. A derivative plot of a showed reaction rates as peaks. At a constant temperature ramp, the reaction rate can be described by da/dT, which follows an Arrenhius rate law. da / dT= A~eE,IRT (1-a)" (1) where A, Ea, and n are the Arrenhius frequency, activation energy, and reaction order, respectively; R is the universal gas constant, and f3 is the heating rate. A representative TGA analysis is shown in Figure 2 in which the student had to convert that raw data to usable data for input into a model for design simulations. Combustion of tires in air revealed five peaks in the derivative plot implying an equal number of reactions, which have been proposed to correspond to light oils, natural rubber, synthetic rubber, and tarsY 5 l Under inert atmospheres only the first three peaks were observed, implying that the tar only combusts in the presence of oxygen. At higher heating rates the peaks overlap but maintain recognizable reactions. Because Eq. (1) cannot be solved explicitly, quantitative parameters were derived from a method from the literature. [l 6 l Plotting 1/T versus -log[1-(1-atn]/T 2 (1-n) forn;tl (2a) log [ log ( 1a) ]/T 2 for n= 1 (2b) reveals a linear plot for the appropriately chosen reaction order. The Arrenhius rate parameters can be related to the slope, m, and the intercept, b, by Ea= 2.3 m R A= lOb~E, /R( 12 R TI Ea) (3a) (3b) By this method, rate parameters for air and 20/80 CO/N 2 atmospheres could be obtained as shown in Table 1. TABLE 1 Parameters Derived from TGA Experiments Ascertained by the Derivative Method of Fristky, et al., (1994) The "Weight" column refers to the amount of mass change that can be accounted for by the reaction. Under air, there was a 5% residual, whereas under an inert atmosphere there was a 38% residual. Reaction under Air Reaction 1 2 3 4 5 Weight 15% 13% 23% 20% 24% Ea (kj/mol) 120 180 187 325 258 A(hz) 5.5 X 10 8 1.3 X 10 13 2.3 X 10 11 4.0 X 10 19 2.7 X 10 13 Reaction under 20% CO 2 Reaction 1 2 3 Weight 18% 12% 32% Ea (kj/mol) 95 203 176 A(hz) 1.4 X 10 6 2.5 X 10 14 3.0 X 10 10 Summer 2006 207

PAGE 66

The outflow of the TGA was connected to an Agilent 3000 micro-GC gas chromatograph (GC). Figures 3 and 4 show a typical species analysis of the product gases by the GC, and allowed the student to conduct a material balance ensuring the integrity of the data generated. Combustion under air revealed that CO 2 was the main constituent of the exhaust. Pyrolysis of tires under CO 2 indicated that the amount of CO 2 in the product gas increased slightly during the reaction and no CO was detected. The GC proved useful for determining the product species and can be used in future work to identify important constituents. SIMULATION Following the laboratory work, the student then reduced the data and calculated the parameters needed to input into a simulation. There were two sets of simulations done, one was thermodynamic and the other used kinetics obtained from the laboratory experiments to more accurately simu late the combined combustor-gasifier reactor. This allowed the student to better understand the type of information thermodynamics can provide versus an actual operating system where the kinetics play an important role. Results are shown in Table 2. Thermodynamic Simulation The thermodynamic data obtained from literature and the student's experiments (bomb calorimeter) were input into Aspen. Tires were defined as a mixture of the rubber mono mer (C 5 H 8 ), graphite, iron, and zinc so that the final assay equaled the literature value_[sJ Figure 5 shows the process flow diagram of the equilibrium simulation. The combustor and gasifier were represented as two Gibbs reactors con nected by material and heat streams. The ideal "separators" surrounding the gasifier selectively remove solids from the streams. A calorimeter module was programmed to com bust the syngas with stoichiometric amounts of oxygen only, to enable the student to conduct an energy balance and compare that with initial calculations. The model output of 170 MW total power from 18,700 kg per hour (5.2 kg per second) indicates an energy input for tires of -33 MJ/kg of tire, which is consistent with bomb calorimeter measurements. The product gas is composed of 24.0% H 2 and 10.8% CO. The temperature of the gasifier is 786 C, which suggests that the temperature of the inner wall may be maintained at an acceptable level. The Aspen equilibrium simulation reflects the material and energy balance calculations from preliminary assessments. Kinetic Simulations Upon completion of the thermodynamic simulations, programming of the kinetic parameters began. For this task, the student used his own data generated from the TGA and compared them to literature values to ensure the integrity of the data. The TGA data from Table 1 was used for the Aspen 208 kinetic simulations. The kinetic simulations first attempted to model the TGA experiments to ensure the results were consis tent. The simulated TGA was a semi-batch reactor with a charge of the tire sample and a constant flow rate. Due to the kinetic parameters' sensitivity to the phase of the reactants, the tires were modeled as a mixture of coal and graphite-both present in the Aspen database. The Aspen-defined coal had many similar properties to tires. The mixture was modified so that the final analysis of the material would have an enthalpy of -33 MJ/kg Combustion Reactions 10 C/min 0 0 -0.02 0 vi vi Cl CIJ 2: > 0 -0.04 ;: CP C -0.06 200 300 400 500 600 700 800 Temp (C) Figure 3. The results of the TGA with derivative plots show the combustion reactions under 0 2 and N 2 atmospheres (Air20% 0 2 ; enhanced 0 2 28% 0 2 ; deficient 0 2 3% 0) heated at 10 C/min. TABLE2 The results of one Aspen equilibrium simulation show a systern that handles 164,000 tons of tires (~10 million tires) per year. The power output is 170 MW for a syngas that consists of 34.8% by volume of useful material. The temperatures of each reactor have reasonable values. Reactant Flow Rates Product Flow Rates ton/hr Nm 3 /hr X 10 3 % Tires (total) 18.7 H2 25_8 24.0 Combusted 5 6 co 11 6 10 8 Gasified 13 1 CO 2 11 5 10.6 Water 18.0 H 2 0 11 2 10.4 Air 97,490 Nm 3 /hr N2 47.2 44.0 Temperature ( 0 C) Energy (MW) Combustor 1626 Chemical 130 (eq) Gasifier 786 Sensible 40 Chemical Engineering Education

PAGE 67

165 125 '2 100 0 E 6 75 2 e 50 3: 0 a:: (/) 25 "' CJ 0 15 TGA and GC under Air I N2 / 30 ~02110 ~ co 7irCO2 -Tire (TGA) I Temp ( 0 C) 265 365 465 565 665 25 35 45 55 Time (min) 65 300 O "' g:, 200 g :, ro 100 3 0 = Figure 4. GC flow rates (left axis) plotted with TGA mass loss (right axis) for flow under air. The 0 2 and N 2 curves are scaled by 1110 and 1/40, respectively. CO 2 is produced throughout combustion; CO increases in the tar region. and a total chemical assay of the literature. [SJ For the combus tion reactions, the tire was separated into six components based on the reactions determined by TGA in Table 1. The gasification simulation used four components. The simula tion results of Figure 6 show that the simulation closely ap proximates the measured data. The correspondence suggest that the kinetic model can be developed further to simulate the combustion-gasification reactor. Although the ultimate plan was to simulate the design of the integrated gasifier and combustor, time did not permit this step. The exercise of conducting a thermodynamic simulation of the entire design and programming the kinetics to simulate the TGA experiments provided the student with sufficient experience to complete the project had time permitted. Upon comple tion of this task, the student was keenly aware of the many ways to arrive at designing a new technology or modifying an existing technology. Moreover, the student was now well prepared to understand the importance of experimental data generation and how to attack such a process in the future. OUTCOMES The undergraduate design project taught the student how to address engineering challenges. The literature research aided in the design of the combined combustion-gasification system, which was devised solely by the student. The professor's role was to provide insights into the merits and limitations of such a device. Using a more experienced know ledge base, the prof es sor was able to recommend calculations and experiments that would prove the design. Figure 5. Flow diagram of the primary reactor. All feeds began at room temperature (20 C). The calorimeter was added to measure the heating value (HV) of the final product gas (FINPROD). The independence of the student allowed greater opportunities for learning about the busi ness environment surrounding waste manageCombustion in Air 6 5 4 Cl E .._,, (/) 3 (/) (ti 2 1 0 200 300 400 500 600 Temp (C) Summer 2006 700 DJ 4 E .._, rJ) 3 (/) m 2 2 1 Gasification in 20% CO 2 200 300 400 Temp (C) 500 600 Figure 6. The measured data (solid line) and the Aspen simulation (dashed line); the model shows potential for developing further investigations. 209

PAGE 68

ment, industrial thermal processes, reactor designs, and the engineering process. Performing the laboratory experiments helped the student comprehend the operation of standard analytical tools. Further, the student began to understand how to devise a test plan and allocate adequate time to attain a sufficient amount of data. The Aspen simulations were in structive in demonstrating the next level of design execution and evaluation. Finally, to expose the student to the experience of communicating the work, poster or lecture opportunities were pursued with presentations at two academic departments in Columbia University and Barnard College, at a university wide undergraduate research symposium, and at an American Chemical Society meetingY 7 l The individual responsibility for the project encouraged a greater commitment from the student and allowed a wider platform for innovation. The development and successful implementation of the project, however, may have benefited from a larger undergraduate or graduate team. Nonetheless, the student was able to maintain the directives assigned in the predetermined schedule and provide a report at the end of each term. The time frame of the project allowed ample time to understand basic aspects of designing a reactor and carry out the preliminary steps. The student was able to identify future directions for the project in the final report. It is antici pated that the student will be better prepared for future work as a professional engineer and that the project described herein may be continued under other circumstances. IMPLEMENTATION A suggested implementation strategy is briefly presented for those who wish to augment a traditional chemical or environ mental engineering capstone experience with a similar effort. Provided the scope of the project is contained and a schedule is put forth at the beginning of the semester, a project such as this becomes very manageable. As evidenced in Figure 1, all elements of engineering design are covered. An important aspect of this type of project is to have the course span two semesters to give students time to assimilate material, develop and design processes, and possibly build devices or conduct limited experiments. 210 So as not to risk skipping or eliminating any of the critical areas of the design process, the real task, for the student and faculty advisor, is not to spend extensive time on any one area. All components should be developed to the extent that the student can see a clear path to the outcome of each task. Most engineering departments maintain a software license to the common programs used in industry and typically have some level of laboratory capabilities. This leaves only the task plan and timing to be formulated and strictly followed. For reference, the hours spent by the student and faculty advisor on this project were no more than a typical design course and the cost of everything except for the software license was under $200. Typically, software licenses are heavily discounted for educational institutions. REFERENCES 1. E.A. Commission, Criteria for Accrediting Engineering Programs, ABET, Inc., Baltimore, Nov. 1, 2004, pp. 1-19 2. Shaw, A., H.N. Yow, M.J. Pitt, A.D. Salman, and L. Hayati, Trans actions of the Institution of Chemical Engineers, Part A, 82, 1467 (2004) 3. Moller, J.C., and D. Lee, Journal of Engineering Design, IO (1999) 4. EPA in Management of Scrap Tires (2005) 5. ASTM in Standard Practice for Use of Scrap Tire-Derived Fuel, Vol. D6700-01 ASTM International (2005) 6. U.S. Scrap Tire Markets: 2003 Edition, Rubber Manufacturer's As sociation (2004) 7. Ulrich, G., A Guide to Chemical Engineering Process Design and Economics, John Wiley and Sons (1984) 8. Reisman, J.I., and PM. Lemieux, Air Emissions from Scrap Tire Com bustion, (Ed.: C.A.T. Center), EPA (1997) 9. Belgiorno, V., G.D. Feo, C.D. Rocca, and R.M.A. Napoli, Waste Management, 23, 1-15 (2003) 10. Bridgwater, A.V., Chem. Eng. J., 91, 87 (2003) 11. Tsouris, C., and J.V. Porcelli, Chem. Eng. Progress, 99, 50 (2003) 12. Susanto, H., andA.C.M. Beenackers, Fuel, 75, 1339 (1996) 13. Gonzalez, J.F., J.M. Encinar, J.L. Canito, and J. J. Rodriguez, J. of Analytical and Applied Pyrolysis, 58-59, 667 (2001) 14. Jones, RM., J.M. Kennedy, and N.L. Heberer, TAPP! J. (1990) 15. Conesa, J.A., R. Font, A. Fullana, and J.A. Caballero, Fuel, 77, 1469 (1998) 16. Fritsky, K.J., D.L. Miller, and N.P Cernansky, J. of Air and Waste Management Association, 44, 1116 (1994) 17. Weiss, B., and M.J. Castaldi, Novel, Integrated Process for Beneficial Use of Waste Tires, Poster Presentation, ACS, Washington, D.C. (2005) 0 Chemical Engineering Education

PAGE 69

5 =i outreach ) ---11111----------Demonstration and Assessment of A SIMPLE VISCOSITY EXPERIMENT FOR HIGH SCHOOL SCIENCE CLASSES T.M. FLOYD-SMITH, K.C. KwoN, J.A. BURMESTER+, F.F. DALE+, N. VAHDAT, AND P. JoNEst Tuskegee University Tuskegee, AL 36088 T he objective of this demonstration and assessment was to develop an instructional model to inform and enthuse students about chemical engineering. Figure 1 shows the number of B.S. degrees granted nationally in chemical engineeringY 2 i Rhinehart observed a 13-year-cycle period for the production of B. S. degrees in chemical engineering at Oklahoma State University, dating back to the 193Qs_[ll It is not clear at this time if the 13-year-cycle period for chemical engineering degrees awarded will hold,Dl butitis clear that the peak has dropped from approximately 7,500 degrees awarded to approximately 6,500 degrees awarded, representing a 13% decline. Rhinehart attributes the cycling to B.S. chemi cal engineering supply/demand being out of phase but does not discuss the magnitude of the peaks. Halford,[3l however, suggests the decline is due to a rising attraction of potential chemical engineers to the environmental engineering and bioengineering fields. The cause of the decline in chemical engineering enrollment has not been determined conclu sively, but-regardless of the cause-the effect is that when enrollment is low, administrators may question the benefit of maintaining an expensive chemical engineering program_[ll B.S. chemical engineers are indirectly supplied by the nation's high schools. Therefore, one potential approach to positively impact enrollment in chemical engineering under graduate programs is to conduct outreach programs for high schools. Ross and Bayles[ 4 l describe a method for incorporat ing high school outreach into chemical engineering courses. Their goal is to provide role models for high school students by assigning chemical engineering students enrolled in their courses to participate in an outreach project. In contrast, this work describes an outreach program administered and conducted by professors for the purpose of informing high school students about chemical engineering and attracting them to the profession. Copyright ChE Division of ASEE 2006 Summer 2006 8000 ~----------------~ 7000 +------~~~------------, 6000 +-----+----------------.F---------j 5000 +------+------e------------J------+--F--"' >------i Cl iii 4000 ,----,------------a.~------..-----;c===== =;-1 U) :.;:::; ID~ 3000+-------"'....__ ________ -----! 0 ..c E ::::s z 2000 +-----------------, -a-ACS 1000 0 1970 1980 1990 Academic Year 2000 Figure 1. Annual national B.S. chemical engineering degrees awarded. Tamara M. Floyd-Smith is an assistant professor in the Chemical Engineer ing Department at Tuskegee University. She received her B.S. in chemical engineering from Tuskegee University and her M.S .and Ph.D. in chemical engineering from MIT. K.C. Kwon is a professor in the Chemical Engineering Department at Tuskegee University. He received his Ph.D. in chemical engineering from the Colorado School of Mines. Jeffrey. A. Burmester is a physics teacher at Martin Luther King, Jr., High School in Lithonia, Georgia. Frances F. Dale is the science department head at Martin Luther King, Jr., High School in Lithonia, Georgia. Nader Vahdat is the head of the Chemical Engineering Department at Tuskegee University. He received his B.S. in chemical engineering from the Abadan Institute of Technology, his M.S. in chemical engineering from the University of Califomia, Berkeley, and his Ph.D. in chemical engineering from the University of Manchester, England. Paul Jones recently retired as director of product supply for the Snacks and Beverages Division at the Procter & Gamble Corporation in Cincinnati. +: Martin Luther King, Jr., High School, Lithonia, GA t: Procter & Gamble Corporation, Cincinnati, OH 211

PAGE 70

The overall objectives of this demonstration were twofold. First, the authors wanted to develop a presentation giving an overview of the field of engineering with emphasis on chemi cal engineering. Second, the authors wanted to conduct a simple experiment with the high school students so that they have an opportunity to learn a chemical engineering concept and be exposed to principles and problems that practicing chemical engineers will expect to encounter. PRESENTATION DESCRIPTION The demonstration was conducted at Martin Luther King, Jr., High School in Dekalb County, Georgia, in November of 2004. A junior/ senior-level physics course ( a chemistry course may also be appropriate) was chosen for an introductory presentation followed by hands-on viscosity experimentation. Twenty-six students participated in the demonstration during a class period of 90 minutes. General engineering, chemical engineering, and the concept of viscosity were discussed first. In the general discussion of engineering, the major engineering disciplines were described in basic terms (e.g., civil engineering was described as the branch of engineering responsible for designing municipal structures such as bridges and roads). After a general discussion on engineering, the presenta tion was focused on chemical engineering. The facilitator discussed the kinds of jobs that chemical engineers are re sponsible for and the types of engineering fundamentals that chemical engineers study. The job areas described included petrochemicals, intermediate chemicals, food processing, cleaning products, plastics, and pharmaceuticals. When de scribing what chemical engineers study, several core examples were included. The list of what chemical engineers study included accounting for material flows (material and energy balances), how fluids move (fluid mechanics), how heat is transferred, and how materials react to create new things (reaction engineering). The students were informed that the viscosity experiment for the day was related to fluid mechan ics. During the discussion on heat transfer, the example of an egg cooling was introduced. As expected, the students had a good idea about how long it would take for an egg to cool un der different conditions (free vs. forced convection, in air vs. in cool water) but overall were surprised that it is something that chemical engineers expect to predict theoretically and/or empirically. During the discussion of reaction engineering, the example of how an antacid helps indigestion was intro duced. The students were aware of acid/base reactions from their chemistry class, but again didn't realize that chemical engineers are involved in producing the antacids (bases) that are administered to neutralize excess stomach acid. The presentation ended with a discussion on viscosity. Viscosity was described as a fundamental physical property in the study of how fluids move or how "thick" and "slip pery" a fluid is. Several examples including paste, pancake 212 syrup, water, and motor oil were discussed. Viscosity was not mathematically defined during the presentation, and a discussion on Newtonian vs. non-Newtonian fluids was not included because the facilitators thought that it was beyond the scope of what was appropriate for a high school science class. APPARATUS AND THEORY The viscometer used for the demonstration has been de scribed previously_l5l Briefly, the viscometer is a tank-tube viscometer as illustrated in Figure 2. It consists of a tank and a vertical drain tube attached at the bottom of the tank. In addition, a balance, a thermometer, a stopwatch, and a bottle of water at room temperature are required for the experi ment. The viscosity of a fluid is inferred from the drain rate of the fluid through the drain tube of the viscometer tank. The drain rate is dependent on the viscosity of the fluid and follows the behavior described in Eqs. (1) through (4). The detailed derivations of these equations have been described previously. [ 5 l In (H+L .L( gR!p i(t) (1) lh+L) lsR 2 L) h=H-~ (2) nR 2 p -In( 1(H+WnR2p j= l 8~~t j(t) (3) m*=-ln(l-(H+WnR 2 p j (4) where H: initial height of the fluid in the tank (9.3 cm, illustrated in Figure 3) h: height of the fluid in the tank L: length of the drain tube (56.4 cm) g: acceleration due to gravity R 0 : equivalent radius of the tank p: density of the fluid : viscosity of the fluid t: drain duration R: radius of the drain tube (0.0509 cm) m: accumulated amount of a fluid drained from tank 5 m': left-side value of the viscosity equation, as shown in Eq. (3) During the experiment a tank with a rectangular cross sec tion, illustrated in Figure 3, was used instead of a tank with a radial cross section. This modification was made because the tank with the rectangular cross section is easier to fabri cate. Thus, the equivalent radius R 0 was computed with the Chemical Engineering Education

PAGE 71

following equation (5) where W: width of the rectangle (25.4 cm) D: depth of the rectangle (3.81 cm) The experimental procedure for determining the viscosity of water using the tank-tube viscometer is as follows: Fill the reservoir with water. Set up the balance with automatic data acquisition so that the data from the balance are input directly into Microsoft Excel in real time. Use a sampling rate of 1/s. Remove the end cap on the drain tube and allow the water to collect on the balance. After ~ 90 s, stop the data acquisition. Plot m* [left-hand side value of the viscosity equation as shown in Eq. (3)] vs. t (time) and obtain the slope of the line. Extract from the expression of the slope as illus trated in Eq. (6). Measure the temperature of the water used in the experiment. Compare the experimental to the literature value. (6) Calculate the measurement error based on a percent dif.!erence. In addition to the experimental procedure outlined, brief explanations on linear regression, Microsoft Excel features, and standard deviation ( a) were provided to the class. Units were not discussed and, due to time limitations, only one experimental run was performed. RESULTS AND DISCUSSION The viscosity experiment was demonstrated using water at room temperature. The experiment was successful with a measurement error of ~3% which is the typical result obtained in a simple viscosity experimental setting with the tube-tank viscometer in the absence of a temperature regulat ing circulator. Facilitator's Perception Overall, the students were enthusiastic and attentive, sug gesting that the activity was structured appropriately to main tain the interest of a high school student. The students were also willing to interact with and participate in the presentation and the hands-on viscosity experimentation. The experiment Summer 2006 could be improved by structuring it for more student partici pation. Ideally, there should be one station per four students so that the students can perform the experiment themselves. Excluding the computer and the balance, the fabrication cost is ~$100 so the concept is economically feasible. Also, if time permits, it would be illustrative to measure the viscosity of more than one fluid. For example, in addition to measuring the viscosity of water, one could measure the viscosity of an alcohol and its aqueous solutions or water with the viscosity modified by adding a second component such as sugar. Student Survey Students were asked to rate the following five questions on a scale of one to 10 before and after the demonstration, where Rectangular Reservoir Tank Vertical Drain Tube Balance Computer Figure 2. Set-up of a viscosity experiment. : w : r---------------------1 --:u----t Hi ----",r -------" _..._ ____ __. ___ 4 ____ ' r---------w----------1 2R-------t ~----L ----llaal _______________ y ____ Figure 3. Tank-tube viscometer. 213

PAGE 72

one is "no knowledge" and 10 is "very knowledgeable": 1. How much do you know about engineering? 2. How much do you know about chemical engineering? 3. How much do you know about viscosity? 4. How interested are you in engineering? 5. How interested are you in chemical engineering? The students were also asked to comment on what they liked most about the presentation/experiment and what could have been improved. Survey Results Table 1 shows the results from the survey given to the students. The results summarize the students' knowledge and interest before and after the presentation followed by the hands-on viscosity experimentation. The table also shows the difference between the two values and the statistical sig nificance of the results. Using a paired-sample t test, it was concluded that the students gained by at least 36% in the knowledge of and interest in general engineering, chemical engineering, and the viscosity topic. In the future, however, a short test may be more informative than the student self-as sessment for determining how much the students learned during the demonstration. Overall, the survey shows that the students are more interested in general engineering, but their interest in chemical engineering increased between 95% and 230%. Survey Comments Approximately half of the students indicated that the most interesting part of the demonstration was the experiment. The other half indicated that they enjoyed learning about different types of engineering and/or learning about chemical engineering. Most students didn't comment on potential im provements, but of those who did, the majority indicated that more audience (i.e., student) participation was preferred. SUMMARY AND OUTLOOK An experiment/presentation appropriate for high school stu dents was developed and demonstrated. Based on the survey results, the students gained by at least 36% in the knowledge of and interest in engineering, chemical engineering, and fluid viscosity. Furthermore, interest in chemical engineering increased between 95% and 230%. Based on the survey results and the facilitator's perception, for any high school experimental demonstration, a significant portion of the time allotted should be devoted to talking to the students about engineering and chemical engineering. In the future, the facilitators would like to contact high schools and offer to send them simple tank-tube viscometer kits so that a viscosity experiment can be incorporated into their exist ing curriculum. Also, the facilitators would like to develop a program so that undergraduates can participate in the viscosity experiment at local high schools as one of the department's outreach efforts_[ 4 l ACKNOWLEDGMENTS The authors would like to thank the Procter & Gamble Cor poration for funding this work under a curriculum grant. REFERENCES 1. Rhinehart, R.R., "An Analysis of Enrollment Cycling in ChE," Chem. Eng. Ed., 35(1) 50 (2001) 2. The ACS Committee on Professional Training, "ACS Committee on Professional Training Annual Report Tables: 1995-96 through 20032004," (2004) 3. Halford, B., "Pursuing New Paths," ASEE Prism Online, Vol. 13, No. 3 (2003) 4. Ross, J.M., and TM. Bayles, "Incorporating High School Outreach into Chemical Engineering Courses," Chem. Eng. Ed., 37(3) 184 (2003) 5. Kwon, K.C., S. Pallerla, and R. Sanjeev, "Experiments on Viscosity of Aqueous Glycerol Solutions Using a Tank-Tube Viscometer," Chem. Eng. Ed., 33(3) 232 (1999) 0 TABLE 1 Student Survey Results Knowledge Interest Description General Chemical Viscosity General Chemical Engineering Engineering Concept Engineering Engineering Before Averruze 3.7 2.6 1.6 4.2 2.1 Standard Demonstration Deviation 1.9 1.9 1.3 3.0 1.5 After Averruze 8.3 8.5 8.0 7.3 5.5 Demonstration Standard 0.9 1.1 1.7 2.2 2.5 Deviation Difference Average 4.6 5.8 6.3 3.1 3.4 Between Before and After Standard 2.1 2.3 2.0 2.8 2.5 Demonstration Deviation t (ta.= 0.005 = 2.807) 10.7 12.4 15.4 5.4 6.7 99% confidence interval 3.4-5.8 4.5-7.1 5.2-7.4 1.5-4.7 2.0-4.8 214 Chemical Engineering Education

PAGE 73

5 =i laboratory ) ---11111---------=--------PLANT DESIGN PROJECT: Biodiesel Production Using Acid-Catalyzed Transesterification of Yellow Grease RAFAEL HERNANDEZ, TRENT JEFFREYS*, ANIRUDHA MARWAHA, AND MATHEW THOMAS Mississippi State University Mississippi State, MS 39762 0 ver the last 10 years, the chemical industry, federal and state agencies, and the chemistry and chemi cal engineering profession have been increasingly investing intellectual, technical, and financial resources on the research, development, and application of chemicals and fuels generated from renewable raw materials and sustain able processes. The main goal of the involved parties is to develop energy-efficient and cost-effective processes that prevent pollution and decrease our dependency on foreign oil. The number of papers describing sustainable processes and renewable fuels that have appeared in the publications and conferences of the American Chemical Society (ACS) and American Institute of Chemical Engineers (AIChE) have increased significantly over the last five years. The 2005 57th AIChE Institute Lecture was titled "Energy Sup ply Challenges and Opportunities." The ACS dedicated one issue of Environmental Science and Technology, the society's main publication on environmental research, to sustainable processesYl Presently, that journal includes a section on sus tainable technologies in every issue. Additionally, numerous papers were presented at the 2005 Annual AIChE Meeting (Cincinnati) on biorefineries, sustainable technologies, and renewable fuels. Albermarle Corporation in Orangeburg, SC Summer 2006 Rafael Hernandez has a B.S. (1993) and M.S. (1995) in chemical engineering from the University of Puerto Rico, Mayaguez, and a Ph.D. (2002) in chemical engineering from Mississippi State University (MSU), Mississippi State, MS. He worked forthe U.S. Army Corp of Engineers' Engineering Research and Development Center (1994-1997) on the development, design, and implementation of groundwater treatment technologies. Presently, he works as an assistant professor in the Dave C. Swaim School of Chemical Engineering at MSU. His research interests are the development of technologies forthe remediation of contaminated media and the use of biomass for producing value-added chemicals. Trent Jeffreys has a B.S. (2004) in chemical engineering from Missis sippi State University. Currently, he works for Albemarle Corporation in Orangeburg, SC as a process technology engineer. He has worked on projects for cost reductions, automation and control system develop ment, and sample campaigns for new products in the Fine Chemicals Division. Matt Thomas received his B.S. in chemical engineering from Mississippi State University in 2004. He is currently seeking his master's degree in chemical engineering at Mississippi State University, studying the remediation of nitroaromatic contaminated groundwater. After receiving his master's, he plans on continuing his education at Mississippi State University by working toward a Ph.D. in chemical engineering and focus ing on producing oils from glycerol for biodiesel production. Anirudha Marwaha has his B.S. degree in chemical engineering (2003) from Mississippi State University. He is currently pursuing a M.S. degree in chemical engineering at Mississippi State. Since 2001, he has been working for the E-Tech Lab in the chemical engineering department. His thesis work is on the solidification and stabilization of high-level radioactive waste. Copyright ChE Division of ASEE 2006 215

PAGE 74

The main contributing factor to the chemical engineering and chemistry professions' focus on efforts to promote the development of sustainable technologies and the production of renewable alternative fuels such as ethanol, biodiesel, and hydrogen has been the commitment of resources by the U.S. Environmental ProtectionAgency (USEPA), the U.S. Depart ment of Agriculture (USDA), the National Science Founda tion (NSF), and the U.S. Department of Energy (USDOE). One way to develop creative new production processes for renewable chemicals is to educate future chemists and chemi cal engineers on the design, advantages, disadvantages, and economics of current production techniques. knowledge, most chemical engineering capstone design projects focus on the use of petroleum-based raw materi als for producing specialty and commodity chemicals. To broaden the students' perspective on the potential contribu tions of chemical engineering to areas such as new energy sources, global warming, and environmental sustainability, they should be introduced to the conversion of plants, natural oils, microorganisms, and other types of biomass into alterna tive energy sources and value-added products. The capstone course represents an excellent opportunity to assign projects in which students synthesize and analyze renewable-chemicals production facilities. The objective of this paper is to describe a project entitled "Design of a Biodiesel Production Facility Using Acid-Catalyzed Transesterification of Yellow Grease," assigned to the capstone design course at Mississippi State University (MSU). Research on biodiesel is conducted by the class instructor's research group. Thus, the design problem The last core course in the chemical engineering curriculum at most universities in the United States is capstone design. In this course, students have the opportunity to practice, for the last time in an academic environment, the design and economic evaluation of industrial chemical plants. To our 216 Biodiesel is an alternative renewable fuel derived from vegetable oils or animal fats, which conforms to ASTM D6751 specifications for use in diesel engines.[2J Biodiesel utilization has increased significantly over the last 10 years, mainly due to environmental benefits and government efforts to reduce dependence on for eign oil. The use of biodiesel reduces emissions of CO 2 CO, s0 2 and particu lates from operating diesel engines. Under a newly established sustainable energy policy by the U.S. Department of the Interior, over 20 national parks operate boats, trucks, heating systems, electricity generators, and other fuel related systems on 100 percent biodiesel and/or biodiesel/petroleum diesel blends. Blends of 20 percent biodiesel with petroleum diesel require no engine modifications. Furthermore, biodiesel/petroleum diesel blends have demonstrated lubricity en hancements over the newly required low sulfur petroleum diesel. Numerous school districts, transit authorities, public utility companies, and recycling companies have also successfully used biodiesel. Recently, the U.S. military has begun to procure biodiesel for use in on-base vehicles. These numerous experiences with the use of biodiesel have clearly shown the environmental and high performance characteristics of this alternative fuel. In spite of the fact that Mississippi ranks 4th and 16th nationally in yellow grease generation and soybean production (main biodiesel feedstocks), there are no biodiesel production facilities in the state. The Swalm Engineering Design Group at MSU was asked by the Alternative Energy Company to perform a prelimi nary design of a 2,240 lb/hr biodiesel production facility using acid-catalyzed transesterification of yellow grease, and to evaluate the process economics. In order to perform sensitivity analysis of process variables, the company requires a simulation of the whole process using ChemCad. The company has acquired land adjacent to a fertilizer manufacturing company in Yazoo City, Miss., at the cost of $1,000,000 as the plant site. The design is to be based on a project life of 20 years. The major equipment, however, is to be depreciated in accordance with applicable IRS regulations. In your final design report, you are requested to provide estimations of the annual return on investment as well as the rate of discounted cash flow taking into account the most recent laws and regulations on corporate taxes. Design basis and specifications, available utilities, and other information will be provided in further communications. Figure 1. Project description. Chemical Engineering Education

PAGE 75

Some of the comments in the students' course evaluations were ... "What I like most about this course was the fact that the project was broken into separate portions over the whole semester," and "I loved the layout of the class .... also represented an excellent opportunity to integrate re search and education. As part of the course, invited speakers and the course instructor presented seminars on ethics,job interview preparation, entrepreneurship, and the social and environmental implications of reducing our dependency on petroleum. A workshop on ChemCad (chemical process simulation software) was offered to students and faculty by the software creators. PROJECT DESCRIPTION The statement of the problem submitted to the students on the first day of class is presented in Figure 1. The class was divided into five groups and each group had four members selected by the instructor. The same project was assigned to all groups. The open-ended nature of the problem statement led to five different design configurations. The design project was divided into three progress reports (memorandums) and one final report. Several activities and rules were established to maximize participation of all students: (A) Progress reports and the final report were accompa nied by an oral defense. The student in charge of present ing the oral defense was selected at the time of the pre sentation. Each member of a design group was questioned extensively during each progress report presentation. (B) Written peer evaluations were required after each progress and final report. The evaluation forms were similar to those suggested by FoglerP 1 C) A panel of industry and academic members judged and selected the best final presentation. The presence of indus try representatives was additional encouragement for all the students to prepare for the presentation. The instructor selected the best report. The group or groups with the best presentation and final report received plaques and cash awards. PROGRESS REPORTS Division of the design project into progress reports had two objectives. The first objective was to evaluate an induc tive approach to the teaching of plant design. This approach consists of the presentation of a general problem or concept, followed by closer focus on details and the solution of component small problems. This method is applied by the Summer 2006 chemical industry and during academic and industrial research and development activities and it is an approach suggested by chemical engineering educators. [ 4 5 l The second objective was to facilitate the organization of the project and enhance students' time-management skills. Some of the comments in the students' course evaluations were related to the second objective. For example, "What I like most about this course was the fact that the project was broken into separate portions over the whole semester," and "I loved the layout of the class-progress reports and the final presentation." The tasks conducted for each progress report were as follows: Progress Report 1: Literature survey, calculation of gross profits, block diagram preparation, overall mass balance calculations, and input of yellow grease components into the ChemCad database. Progress Report 2: Preparation of process-flow diagram and simulation of the transesterification reactor and methanol recovery system. Progress Report 3: Simulation of all the biodiesel purifica tion steps: neutralization, solids removal, glycerol recovery, and biodiesel and glycerol purification. PROJECT SOLUTION Yellow grease is the fat generated during animal rendering activities. It is mainly composed of oleic, palmitic, and stearic fatty acids attached to glycerol,l 6 l and contains a relatively high percentage of free fatty acids ( 15% ). Zhang (2000) used triolein (triacylglycerol) as a test compound to represent yellow grease during a Hysys simulation of a biodiesel production facility.[7l The acid-catalyzed transesterification of this compound using methanol as the alcohol results in methyl oleate and glycerol. Zhang (2000) assumed that biodiesel could be represented by methyl oleateYl To generate a mixture of transesterification products with similar biodiesel chemical and physical proper ties, the students were encouraged to use several triacylglycerols and oleic acid (free fatty acid) as representative of yellow grease for the Chem Cad simulation of the biodiesel production facility. Figure 2 presents the reactions of acid-catalyzed transesterifica tion of the selected triacylglycerols and oleic acid. The acidc57Hl0406 + 3CH30H HzS04 3Cl9H3602 + C3H803 Triolein Methyl Oleate Glycerol c51H9806 + 3CH30H HzS04 3Cl7H3402 + C3H803 Tripalmitin Methyl Palmitate Glycerol C57H3402 + 3CH30H HzS04 3Cl9H3802 +C3Hg03 Tristearin MethylStearate Glycerol cl8H3402 +CH30H-H~z~SO~-c19H3602 + H2 0 Oleic acid Methyl Oleate Figure 2. Acid-catalyzed transesteriftcation reaction for producing fatty acid methyl esters (FAME). 217

PAGE 76

218 WnhW_, @) P-I02A/B YGRacyclo Pumo Vallow G,_. Feed P-30 1 A/B Wash Waler FHCIPump Figure 3a. 14 T-300 LLE P 103AIB YGFoed Pump -Yellow o-. Racycle 11 P-210AIB MoOH Recycle Pump 23 29 Figure 3b. c.lcfum Sluny Figures 3a and 3b. Process flowsheet . Chemical Engineering Education

PAGE 77

catalyzed transesterification of the proposed components of yellow grease results in a mixture of methyl esters of oleic, palmitic, and stearic fatty acids. This mixture contains more than 90% of the methyl esters found in commercial biodie sel from yellow grease. Phase behavior of triglycerideand alcohol-rich phases was ignored for simplicity. Students recognized that they were making this simplification. The complete process flow diagram (PFD) and stream table are presented in Figure 3 (a and b) as well as in Table 1. Both were prepared using the ChemCad process simulation software licensed by Chemstation in Houston. Some of the physical and chemical properties of the yellow grease-assumed components were determined using the UNI FAC Group Contribution method in ChemCad. Other basic properties, such as boiling point and melting point, were input manually into the simulator. The first main unit operation of the PFD is the transesterifi cation reaction system (R200). To determine reactor volumes, it was assumed that the reactors were half full and the reac tions followed first-order kinetics. Reactor volume meeting TABLE 1 Stream Properties Corresponding to the Process Flowsheet Presented in Figure 3 Stream No. 2 7 8 9 10 11 12 14 Name H 2 s0. MeOH YG RxOut MeOHRcy Rx Feed MeOHBot LLEBot Molar Flow, lbmol/h 3.93 7.85 3.35 178.07 162.57 178.07 15.5 32.32 Mass Flow, lb/h 385.12 251.52 2206.75 8200.12 5207.87 8200.09 2992.34 1057.53 Temperature, c 25.12 25.16 25.19 80 64.41 59.02 140.35 139.57 Pressure, kPa 340 340 340 400 101 400 110 111 Vapor mole fraction 0 0 0 0 0.0032 0 0 0.7484 Enthalpy, MMBtu/h -1.3256 -0.80637 -2.7577 -21.14 -16.459 -21.501 -4.6005 -4.6009 Average mo!. weight 98.08 32.04 659.29 46.05 32.04 46.05 193.03 32.72 Actual dens. lb/ft 3 114.42 49.28 55.17 49.07 15.53 5031 53.42 0.09 Std liq. ft3/hr 3.38 5.03 39.63 154.84 104.21 155.02 50.63 13.52 Flow rates in lbmol/h Triolein 0 0 1.17 0.04 0 1.2 0.04 0 Tripalmitin 0 0 0.67 0.02 0 0.69 0.02 0 Tristearin 0 0 0.34 0.01 0 0.35 0.01 0 OleicAcid 0 0 1.17 0 0 1.17 0 0 Sulfuric Acid 3.93 0 0 3.93 0 3.93 3.93 3.9 Methanol 0 7.85 0 162.64 162.49 170.34 0.16 0.15 Methyl Oleate 0 0 0 4.67 0 0 4.67 0 Methyl Palmitate 0 0 0 2.06 0 0.04 2.06 0 Methyl Stearate 0 0 0 1.28 0 0.26 1.28 0 Glycerol 0 0 0 2.18 0 0 2.18 2.18 Water 0 0 0 1.25 0.08 0.08 1.17 26.09 Calcuim Oxide 0 0 0 0 0 0 0 0 Calcium Sulfate 0 0 0 0 0 0 0 0 Summer 2006 219

PAGE 78

TABLE 1 CONTINUED Stream Properties Corresponding to the Process Flowsheet Presented in Figure 3 Stream No. 16 17 18 19 Name LLEOH Biodiesel YGRcy NeutRxOut Molar flow, lbmol/h 8.44 8.07 0.38 36.21 Mass flow, lb/h 2389.81 2240.98 148.83 1276.07 Temperature, C 101.17 41.47 330 100 Pressure, kPa 101 8 20 110 Vapor mole fraction 0.001973 0 0 0 Enthalpy, MMBtu/h -2.6945 -2.6333 0.14603 -6.6235 Average mo!. weight 283.1 277.84 396.16 35.24 Actual dens. 1 b/ ft 3 42.82 52.63 42.82 86.96 Std. liq. ft3/hr 44.4 41.64 2.76 14.16 Flow rates in lbmol/h Triolein 0.04 0 0.04 0 Tripalmitin 0.02 0 0.02 0 Tristearin 0.01 0 0.01 0 OleicAcid 0 0 0 0 Sulfuric Acid 0.03 0.03 0 0 Methanol 0 0 0 0.15 Methyl Oleate 4.67 4.67 0 0 Methyl Palmitate 2.06 2.01 0.04 0 Methyl Stearate 1.28 1.02 0.26 0 Glycerol 0 0 0 2.18 Water 0.34 0.34 0 29.99 Calcium Oxide 0 0 0 0 Calcium Sulfate 0 0 0 3.9 the conversion requirement (97% the initial triglycerides) was minimized by including two equal-size reactors in series. The first and second reactors achieve an overall 83% and 97% conversion, respectively. The volume of each reactor was 200 ft 3 and the material of construction selected was 316 stainless steel. The reactions were performed at 80 C and 400 kPa. The reactor influents were 3.35 lbmol/hr, 3.93 lbmol/hr, 170.42 lbmol/hr yellow grease, sulfuric acid, and methanol. The reactors were simulated in Chem Cad using the equilibrium reactor. This reactor gives the user the capability to simulate multiple reactions. The purpose of the methanol recovery system (T210) is to 220 20 21 22 23 25 27 GlycFeed Waste H 2 0 Glycerol CaSO 4 CaO WashH 2 O 29.76 27.71 2.05 6.45 3.9 25.26 686.56 501.17 185.39 589.52 218.54 455 100.08 98.81 250 100.08 25 25.08 340 101 110 340 101 340 0 0 0 0 0 0 -3.8985 -3.3341 -0.52741 -2.7249 2.1632 -3.1021 23.07 18.09 90.27 91.36 56.08 18.01 63.16 59.66 68.13 154.95 155.46 62.22 10.4 8.05 2.35 3.76 1.4 7.29 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.14 0.14 0 0.01 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2 0 2 0.17 0 0 27.62 27.57 0.05 2.37 0 25.26 0 0 0 0 0 3.9 0 0 0 3.9 0 0 return excess, umeacted methanol to the reactor to save raw material costs. The major challenge in simulating the metha nol recovery tower is to return as much methanol as possible to the reactor, thus minimizing water in the recycle stream. Conversion of triglycerides into biodiesel drops dramatically if the reactants contain between 0.5% and 5% water. The simulation was conducted using the Chem Cad tower module. The column was operated using atmospheric pressure for the overhead stream and a bottom pump-out pressure of 110 kPa. The smallest number of theoretical stages that could be obtained while keeping the water-weight percent of the reactor feed below 0.10% was 13 with a feed stage of seven. This resulted in a water concentration 0.064% by weight in Chemical Engineering Education

PAGE 79

the feed to the reactor. The distillation column recycles 99 .9% by weight of the methanol in the reactor effluent. The remaining portion of the process after the methanol re covery system consists of separation and purification steps to obtain purified biodiesel, glycerol, and yellow grease recycle streams. The effluent of the surge tank (V-220) is pumped into the bottom stage of the liquid-liquid extractor (T-300). Water cascades down the column after being fed into the top stage. The wash water extracts the entrained glycerol while the biodiesel and umeacted yellow grease exit the top of the column. The ChemCad simulation of T-300 resulted in four theoretical stages for complete separation of biodiesel from glycerol. The students conducted a sensitivity analysis using ChemCad to determine the effect of reboiler operating tem perature of the biodiesel distillation column and wash-water flow into the liquid-liquid extractor on cost and biodiesel purity, respectively. Figure 4 shows the effect of wash-water flow on biodiesel purity. It can be observed that water flows in excess of 500 lb/hr have a negligible effect on biodiesel purity. This type of analysis is essential to determine optimum plant operating conditions to meet biodiesel quality. The crude glycerol stream from the bottom of the extrac tor flows to the reactor (R-500) for the neutralization of the sulfuric acid catalyst by the following reaction: H 2 SO 4 + CaO CaSO 4 (gypsum)+ Hp Calcium oxide (CaO) was the base choice due to low cost, 0.980 '2 0 0.960 ::E Cl "iii 0.940 ::::s D.. ai 1/) .!!1 0.920 "C 0 iii 0.900 0 100 200 300 400 limited complications in regard to materials of construction, and low solubility of its salts formed by neutralization. A CSTR was selected to perform the transesterification reaction. The reactor was simulated in Chem Cad using the stoichiomet ric reactor module and assuming 100% conversion of sulfuric acid. The reactor was maintained at 80 C. Effective mixing of solids is easily maintained by physical agitation in a CSTR. Additionally, the CSTR should prevent any excess collection of calcium sulfate in the neutralization reactor. A centrifuge (CN-510) was used to separate the gypsum from the glycerol and water. A solids effluent moisture fraction of 10% was defined for the centrifuge. It was assumed that the gypsum recovered was sold to a cement company at $56/ton. 1 The liquid stream from the centrifuge flowed into the glyc erol purification tower (T-600) to achieve a bottoms product of 99.5% by weight purity glycerol. Four theoretical stages were required to achieve the desired purity. To use high-pressure steam, the column must be operated at a reduced pressure so that the reboiler temperature is 250 C. The final reflux ratio of 1.8 results in a reasonable reboiler duty while maintaining the desired purity of 99.5%. The biodiesel purification column (T-400) must produce 99.6% by weight biodiesel by separating the methyl esters from the umeacted yellow grease. This column presented several challenges in simulating its operation due to the lack of experimental vapor-liquid equilibrium data for biodiesel 1 Communication with cement company. 500 600 700 800 900 1000 Wash Water (lb/hr) Summer 2006 Figure 4. Effect of wash-water flow into the liquid-liquid extractor on the biodiesel purity exiting the distillation tower. 221

PAGE 80

and yellow grease. Operation of the column at atmospheric pressure required extremely high reboiler temperatures of up to 600 C to achieve a sufficient biodiesel purity. There fore, the column was simulated under severe vacuum with top and bottom operating pressures of 8 kPa and 20 kPa, respectively. The necessity oflow vacuum for the separation of triglycerides from biodiesel also has been observed by other investigators.[ 6 7 l Under these conditions, the bottom product temperature was 330 C. The vacuum necessary for this separation can be achieved using multistage steam injec tors_ [SJ Dowtherm G at 357 C was selected as the heating medium for the reboiler. Students prepared the final report following the format suggested by Peters, et a[.[ 9 l Capital and operating costs were determined using the Cap Cost software included in the textbook by Turton, et al.,l 10 J Web sites, and communications with vendors. Several scenarios were evaluated to determine plant economics. For example, students evaluated return on investment (ROI), taking into consideration the biodiesel tax incentive ($1.00/gal) included in the current version of the Energy Bill. ROI also was determined after increasing the op erating capacity of the plant. These scenarios helped students understand the economics of scale and the current situation of the biodiesel industry in the United States, which requires government incentives to be economically feasibleY 1 l The results of the estimation of capital and total product costs are presented in Tables 2 and 3. The information in these tables was essential to determine net present value and the ROI. The prices used for raw materials costs were: yellow grease, $0.117 5/lb; methanol, $0. 6/ gal; sulfuric acid, $67 /ton; TABLE2 Estimation of Capital Investment for the Proposed Biodiesel Production Facility Estimation of Capital Investment Cost components Direct Costs Equipment (including service, installation, and instrumentation) Distillation columns $206,100 Jacketed reactors $252,450 Liquid-liquid extractor $42,000 Heat exchangers $448,900 Pumps $93,892 Centrifuge $88,349 Tanks $153,263 Total equipment costs $1,291,642 Land (buildings and service facilities included) $1,000,000 Indirect costs (20% fixed-capital investment[ 91 ) $572,910 Fixed-capital investment $2,864,552 Working capital (15% of fixed-capital investment) $429,683 Total capital investment $3,294,235 222 and calcium sulfate, $56/ton. Except for the price of calcium sulfate, all the other prices were obtained from the September 2003 issue of the Chemical Market ReporterY 2 lThe students assumed that the solid recovered during the neutralization step (calcium sulfate) was sold to a local cement company. They contacted a local cement company to obtain a calcium sulfate purchasing price. Utilities costs were the following: low-pres sure steam, $2.50/1000 lb; high-pressure steam, $5.50/1000 lb; natural gas, $2.70/1000 SCF; electricity, $0.04/kWh; cool ing water, $0.05/1000 gal; wastewater treatment, $56/1000 m 3 ; and process water, $0.5/1000 gal.[ 1013 l Total income was calculated by adding biodiesel, glycerol, and calcium sulfate sales. The prices used for biodiesel, glycerol, and calcium sulfate were $2.40/gal (the price of petroleum diesel at the time was $1.40/gal and the $1.00/gal tax incentive was added), $0.72/lb,l1 2 l and $15/ton,l1 4 l respec tively. The mass and volume rates are presented in Table 1. The total annual income was $7,083,700. Subtracting the total product costs shown in Table 3 results in annual gross earnings of $888,100. Assuming a35% tax, the after-tax profit (ATP) was $577,200. The after-tax cash flow (ATCF) is the sum of the ATP and depreciation. ATCF is calculated for every year of plant op eration. The depreciation was calculated using straight-line depreciation with 9.5 years recovery period. Thus, deprecia tion and ATCF were given by: d = original investment 9.5years d= $2,291,642 = $241,226 9.5years year ATCF=ATP+d ATCF = $577,277 +241,226 = $818,503 Only half of the depreciation was added in year 10 of operation and no depreciation was added in the final years of operationY 0 l ROI is a profitability measure defined as the ratio of profit to investment. Average profit over the 20 years of plant operation and fixed capital investment were used to calculate ROI for the biodiesel production facility. This value resulted in: 20 L,ATP ROI= 1 xlOO 20 ROI= $ 577 277 x 100 = 20% $2,864,552 This value of ROI is considered acceptable for a new product entering into an established market.[ 9 l The pay back period is the length of time necessary for the total 1 Communication with cement company. Chemical Engineering Education

PAGE 81

return to equal the capital investment. It was calculated using the following equation: FCI PBP = ~ 2 ~ 0 --L, ATCF 20 PBP = $2,864,552 =4.14 ears $13,837,187 / 20 y To calculateATCF, full depreciation was only added the first nine years of operation and only half of the depreciation was added for year 10. As mentioned above, no depreciation was added the final years of operation. The value of PBP obtained is also acceptable for a new product entering an established market. [ 9 l The students concluded that a biodiesel production facility is not economically feasible without government tax incentives. This result gave the students an understanding of the need for state and federal support for developing new industries associated with renewable energy. Some of the results presented above were taken from the class-best final report. Student course evaluations and senior exit interviews indicated that the application of research and teaching was an excit ing and motivating experience for the class. Some of the students' TABLE3 Estimation of Total Product Cost for the Proposed Biodiesel Production Facility Estimation of Total Product Cost Cost Components Cost/Year Manufacturing Cost Raw Materials $2,407,199 Utilities $367,178 Labor (based on plant capacity kg/ day19l) $1,252,912 Maintenance (7% of fixed capital investment $90,415 minus land and indirect costs[ 9 l) Operating (15% of maintenance costs) $13,562 Depreciation (straight line depreciation) $241,226 Local Taxes (1 % of fixed capial investment 19 l) $28,645 Insurance (1 % of fixed capital investment 19 l) $28,645 Overhead (56% of labor and maintenance 19 l) $749,202 Total Manufacturing Cost $5,178,984 General Expenses Administrative (20% of operating labor and $273,186 maintenancel9l) Distribution and Marketing (7% of the total $433,695 product costs[ 9 l) Research and Development (5% of the total $309,782 product costs[ 9 l) Total General Expenses $1,016,663 Total Product Cost (Total Manufacturing + $6,195,647 General Expenses) Summer 2006 comments about the project included: "I liked the fact that the project was a real-life application." "I became more competent with ChemCad." "This class helped with my teamwork skills." Additionally the class benefited by: Access to the instructors extensive literature collection on biodiesel production technology. Excitement of working on the production of renewable Juel with clear environmental, health, and safety benefits. Discussing contemporary issues associated with the economic feasibility of a renewable Juel. Visualizing the importance of lifelong learning on the application of chemical engineering prin ciples to contribute solutions to society s dwin dling energy resources. Determining the capital and operating cost driv ers of the acid-catalyzed transesterification biodiesel production process. CONCLUSIONS The design project offered students the opportunity to apply chemical engineering to the transformation of a nontraditional raw material into a fuel. The students gained a new perspective on the potential contributions of chemical engineering to areas such as new energy sources, sustainability, and policy. The approach of presenting a general problem or concept, followed by a closer focus on details and the solution of component small problems using chemical process simulation, was key to the successful completion of the design problem. ACKNOWLEDGMENTS The course instructor is grateful for the excellent presentations on ethics, ChemCad, and career options provided by Robert Green, Stuart Schwab, and the Mississippi State University Career Services. Funds for the best report and presentation awards were pro vided by Tom Austin, a Mississippi State chemical engineering alumnus. REFERENCES 1. Environmental Science and Technology, 27(23) (2003) 2. US EPA Office of Air and Radiation, A Comprehensive Analy sis of Biodiesel Impacts on Emission Exhaust, EPA Report Number: 420-P-02-001 October (2002) 3. Fogler, H.S., 4. Wankat, PC., The Effective, Efficient Professor: Teaching, 223

PAGE 82

Scholarship, and Service, Allyn and Bacon, Boston (2002) 5. Dahm, K.D., RP Hesketh, and M.J. Savelski, Chem. Eng. Ed., 36 (3), 192, (2002) 6. Canakci, M., and J. Van Gerpen, Trans. ASAE, 44, 1429 (2001) 7. Zhang, Y., M.A. Dube, D.D. McLean, and M. Kates, Bioresource Technology, 89(1)(2003) 8. Perry, R.H., D.W. Green, and J.O. Maloney (Eds.), Perry's Chemi cal Engineering Handbook, 6th Ed., McGraw-Hill, Inc., New York (1984) 9. Peters, M.S., K.D. Timmerhaus, and R.E. West, Plant Design and Economics for Chemical Engineers, 5th Ed., McGraw-Hill, Inc., New York (2003) 224 10. Turton, R., R.C. Bailie, WB. Whitting, and J.A. Shaeiwitz, Analysis, Synthesis, and Design of Chemical Processes, 2nd Ed., Prentice Hall, NJ (2003) 11. Tyson, K.S., J. Bozell, R. Wallace, E. Petersen, andL. Moens, Biomass Oil Analysis: Research Needs and Recommendations, NREL Report No. TP-510-34796 (2004) 12. Schnell Publishing Company, Chem. Marketing Reporter, 264(6) (2003) 13. Seider, WD., J.D. Seader, and D.R. Lewin, Product and Process Design Principles: Synthesis, Analysis, and Evaluation, 2nd Ed., John Wiley & Sons, Inc., New York (2003) 0 Chemical Engineering Education

PAGE 83

5 =i laboratory ) ---11111---------=--------EXPERIMENTAL INVESTIGATION AND PROCESS DESIGN in a Senior Laboratory Experiment KENNETH R. MUSKE Villanova University Villanova, PA 19085 A drying experiment for the senior unit operations labo ratory course at Villanova University is described in this article. This experiment involves the determina tion of the drying rate of a solid material in a forced convection drying apparatus and the scale-up design of this process. The experimental drying rate data is used to determine the appro priate transport coefficients to mathematically describe the drying process. The students then use this mathematical model for the design of a large-scale dryer for a specified production rate of the material under study. Various solid materials such as sand, gravel, clay, sawdust, natural and synthetic fibers, and agricultural products have been used in this experiment. This variety of materials is intended to provide each student group with a different experience that can be compared and contrasted during student group oral presentations at the end of the semester. The main goal of the laboratory exercise documented in this article is to provide the students with hands-on experience in the analysis and design of drying processes. Drying is an essential unit operation in the chemical process industries with applications ranging from forest products[ll and mineral processingl 2 l to food products[ 3 l and pharmaceuticals_[ 4 l Al though this technology has been a key component of chemical engineering since its inception as an academic discipline, the science of drying continues to remain an active area of research and development. [ 5 l Despite its widespread industrial importance, however, it is not emphasized in the heat and mass transfer courses due to time constraints. This laboratory experiment provides an opportunity for students to apply transport phenomena concepts presented in the classroom to the process of drying, while becoming familiar with this common unit operation. A second goal of this experience is to provide students with an opportunity to apply the results of their experimental study to a process design. A similar approach to the unit operations laboratory course is advocated in Reference 6. The emphasis of this experiment is not simply to obtain data to determine transport coefficients. The students must also use their results in the scale-up design of the drying process. This addition of a design element to the laboratory provides a more practical objective for students and a more realistic application of their experimental investigation. This experience also provides additional learning objectives in the laboratory course, such as the development of engineering awareness, mathematical modeling, scale-up, and economic evaluationYl There have been a number of chemical engineering labo ratory drying experiments reported in the literature such as microwave drying of sand[ 8 l and convection drying of a towel.[ 9 l A bench-scale experimental drying apparatus[lOJ and the statistical treatment of drying data[lll have also been re ported. The unique aspect to the experiment described in this article is both the incorporation of a design element and the study of a wide variety of materials with drastically different drying properties for each group. Kenneth Muske is an associate professor of chemical en gineering at Villanova University where he has taught since 1997. He received his B.S.ChE and M.S. from Northwestern (1980) and his Ph.D. from The University of Texas (1990), all in chemical engineering. Prior to teaching at Villanova, he was a technical staff member at Los Alamos National Laboratory and worked as a process control consultant for Setpoint, Inc. His research and teaching interests are in the areas of process modeling, control, and optimization. Copyright ChE Division of ASEE 2006 Summer 2006 225

PAGE 84

LABORATORY EQUIPMENT The experiment is carried out using a batch cross-circula tion cabinet dryer. A schematic of the dryer system is shown in Figure 1. Air is supplied at a rate of 0-440 ft 3 /min by a centrifugal blower with a gate-valve arrangement to adjust the air flow into the cabinet. The air is heated by two steam coil heaters located in the bottom half of the cabinet. The inlet air is passed through the heaters in the bottom half of the cabinet before being redirected to the drying section in the top half of the cabinet. A baffle arrangement (not shown in Figure 1) provides uniform air flow in the drying section. The volume of the drying section is 5. 6 ft 3 and the cross-sectional area for the air flow is 1.9 ft2. The material to be dried is contained within a shallow tray or wire basket that is suspended in the air stream. Dryer air temperature is controlled by a valve on the inlet steam line to the heaters with manual valves to isolate either heater. Steam is supplied from the high-pressure building main. Source pressure to the dryer is maintained at 30 psig by a regulator on the supply line. This reduction in the heater steam supply pressure is for both safety considerations and improved temperature control by increasing the normal operating range of the steam valve. The maximum dryer temperature is restricted to 200 F for all experiments to prevent the possibility of thermal decomposition or ignition of the solid material. Dryer process measurements include: the mass of the tray and material using a load cell (Transducer Techniques EBB-5 load cell [0-5 kg] and DPM-3 digital panel meter with analog output); the air temperature and relative humidity in the dryer using a humidity probe (Omega Engineering HX94C relative humidity/temperature probe); the surface temperature of the material in the tray and a second dryer air-temperature mea surement using thermocouples (Analog Devices 2B52A type T thermocouple transmitters); and the inlet air flow rate inf erred from the differential pressure across an orifice in the inlet air header to the blower (Setra Systems C239D differential Vent + Gate Valves ----! Air Inlet --+c___,~-L.._.----' Orifice Meter Tray Thermocouple Condensate Load Cell Figure 1. Experimental cabinet dryer system. 226 pressure transmitter). The approximate cost of this instrumen tation at the time of installation was $400. In addition to the electronic measurements, there is a thermometer inside the dryer cabinet, a wet/dry bulb thermometer to determine the conditions of the inlet air, and a water manometer connected to the orifice in the inlet air line. The data acquisition and control computer system displays the process measurements in real time and also records these values in a data file for later analysis by the students. A data sampling period in the range of 0.25 to 1 min is suggested to the student groups for data collection by the computer system. LABORATORY EXERCISE There are two three-hour laboratory sessions each week for the experiments in the senior laboratory course. These sessions are composed of two planning, experimental, and analysis sequences. The students only have access to the drying apparatus during the two experimental sessions. In the first planning session, the group is introduced to the material that they will be studying, the range of moisture content that they must consider, and the specifications on the final dried product. Because the students have essentially no practical experience with drying processes, they are expected to re search the drying process and plan their experimental study during this session. Presentations on drying can be found in Perry's chemical engineering handbook[ 12 i and other process engineering handbooks such as Cooper, et al.,l 13 l along with unit operation and mass transfer texts such as McCabe, Smith and Harriot,l1 4 l Geankoplis,l1 5 l and TreybalY 6 l The initial drying experiments are carried out during the first experimental session where the students determine the character of the drying curve for their material. An initial estimate for the tray loading is based on rules of thumb for the design of drying processes such as presented in Reference 13. The drying rate and moisture range for the constant rate period and the transition to the falling rate period are determined from the students' initial experimental results during the first analysis session. An initial determination of the corresponding mass Air Temperature I Humidity Probe Steam Heaters Steam Valve transfer coefficients is also carried out dur ing this session. In the second planning session, the group develops its experimental plan for the sec ond experimental session. Depending on the drying behavior of their material and the results of their initial experimental ses sion, during this session the student groups usually either concentrate on the falling rate period or consider the effect of different tray loadings. The incorporation of a second experimental session allows for the unstruc tured experimental approach adopted in this laboratory course. Benefits to the students of an unstructured approach include exposure to a more realistic experimental study, since Chemical Engineering Education

PAGE 85

in reality the precise details of a procedure are rarely known in advance and the experimental plan often evolves as more information is discovered. The final analysis session is used to analyze the data ob tained from the second experimental session, and to design the scale-up drying process. Each student group is informed of the required production rate and initial moisture content of their source material after completion of the second experimental session. This practice is implemented in order to prevent the students from specifically targeting their experimental study to their scale-up process design requirements. EXPERIMENTAL ANALYSIS The process of drying can be described using an energy balance determined by the heat transfer rate between the hot gas and the moist solid, and a material balance determined by the mass transfer rate between the moist solid and the hot gas. The corresponding liquid component evaporation rate can be calculated using the relationships[ 14 l m=M hA Tg-1; =M Ak (yi-yg) w 6.Hi w y (1-y)LM (1) where riJ. is the evaporation rate of the liquid component from the moist solid, Mw is the molecular weight of the liquid component, A is the contact area between the moist solid and hot gas, his the heat transfer coefficient, T is the bulk hot g gas temperature, Tis the solid-gas interface temperature, Af-1. is the latent heat ~f vaporization of the liquid component at the interface temperature, k is the mass transfer coefficient y expressed on a mole fraction basis, y is the bulk liquid g component mole fraction in the hot gas, yi is the liquid component gas-phase mole fraction at the solid-gas interface, and (1 y)LM is the log mean value of (1 y) and (1 yg). Forreasons of convenience and safety, the liquid compo nent is water and the gas is air for all of the experimental studies in this laboratory. Due to the time constraints in the laboratory sessions and the relatively slow evaporation rates for the materials under study, the dryer system is typi cally operated at near maximum air flow rate to maximize the evaporation rate. Under these conditions, constant air temperature and humidity can be safely assumed. Because the thermal changes to the system occur at a faster time scale than saturation changes, it is also appropriate to assume that the drying rate can be expressed in terms of the mass transfer relationship for drying process design calcula tionsY7l This assumption is applied in both the experimental data analysis and scale-up design. It should be noted that a similar mass transfer expression to Eq. (1) in terms of partial pressure and humidity driving forces can also be employed. Although partial pressure has been the most popular, the student groups have tended to be rather evenly divided in their choice of driving force for their mass transfer coefficient. Summer 2006 The determination of the mass transfer coefficient depends on the behavior of the drying curve for the material under study. All of the materials considered in this experiment exhibit a constant-rate drying period and some exhibit a falling-rate drying period for the moisture range of interest. The constant-rate period is characterized by a constant rate of drying that is independent of the moisture content. Dur ing this period, a continuous film of water exists on the solid surface that is constantly replenished as the surface water evaporates. The falling-rate drying period occurs when the moisture content of the solid falls below some critical point. After this point, there is insufficient moisture present to maintain a continuous liquid film on the solid surface and the liquid mass transfer in the solid phase becomes limiting as opposed to interfacial mass transfer during the constant-rate period. This critical point is a function of the material, the material thickness, and the driving force for mass transfer that usually must be determined experimentallyY 4 l The drying rate typically decreases as the moisture content in the solid decreases during this period. Under typical experimental conditions during the con stant-rate drying period, all of the parameters in Eq. (1) are relatively constant. The constant-drying-rate mass transfer coefficient can then be determined from the slope of the total tray mass vs. time, total material mass vs. time, or the free moisture mass vs. time drying curves as follows k = -a(l-y)LM y MWA(yi -yg) (2) where a is the slope of the drying curve during the constant rate drying period, y is the mole fraction of water in the g inlet air determined from the humidity probe in the cabinet and checked with the wet/dry bulb thermometer, and yi is the mole fraction of water at the solid-gas interface. The interface mole fraction is assumed to be the equilibrium saturation value at the interface temperature, which can be determined using steam tables or the Antoine equation. As discussed in Reference 15, it is possible to apply the dilute gas-phase mole fraction approximation (1 y)LM"" 1 in the analysis of the constant-rate-period mass transfer coefficient. The determination of the mass transfer coefficient for the falling-rate period is more problematic due to the changing conditions of the material and the interface. Although many approaches exist to describe the falling-rate period,l1 415 l one of the simplest is to assume that the drying rate is proportional to the difference between the free moisture in the solid and the equilibrium free moisture, where kx is the solid-phase mass transfer coefficient, X is the bulk solid free moisture, X* is the equilibrium free moisture, m, is the dry mass of the solid material, and the 227

PAGE 86

other parameters are as defined previously. Eq. (3) can be integrated yielding ill t= s ln(XX*)+ b = aln(XX*)+ b (4) MwAkx where a is the multiplicative constant and bis the constant of integration. The solid-phase mass transfer coefficient can be 600 550 determined from the constant, a, obtained by a logarithmic fit of the free moisture vs. time experimental data. k =~ X aM A w (5) This fit is easily accomplished using Excel or any curvefitting numerical package. The use of the free moisture vs. time curve, as opposed to the drying rate computed by central differencing the data as discussed Measured Mass o Linear Regression -in Reference 8, provides a more ac curate estimate of the mass transfer coefficients for both the constant 500 450 (J) (J) co cil 400 -~ cil 350 300 250 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 Time (minutes) Figure 2. Total material mass drying curve. 150 149 148 Material Interface Temperature --&147 146 145 :::, cil ai 144 C. E Q) 143 f142 141 140 139 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 Time (minutes) Figure 3. Material interface temperature. 228 80 85 90 95 100 G> (!) ,jJ (j) 6 i (j) (j) 1 al I rb J (/) 80 85 90 95 100 and falling rate periods because of the noise inherent in the load cell measurements. EXAMPLE EXPERIMEN TAL RESULTS Figure 2 presents the total mate rial mass (load cell reading minus the empty-tray mass) drying curve for a coarse sawdust material ini tially composed of a 2: 1 volumetric mixture of sawdust (900 ml-156 g) and water (450 ml-450 g) with an average material thickness in the tray of 1 cm. Figure 3 presents the material interface temperature for this system determined by a thermocouple embedded in the surface of the sawdust. Linear regression on the material mass for the constant-rate drying time period between 5 and 80 minutes resulted in a slope of -3.89 g/min with a correlation coefficient of 0.992. Assuming an equilibrium interface temperature of 140 F for the constant drying period, the resulting constant-rate drying period mass transfer coefficient is k = 11.2 gmol/m 2 -min. Although y not used in this calculation, the log mean value was (1 y)LM= 0.895, which is close to the dilute-gas phase mole fraction approxima tion. The specification on the final dried sawdust product was that it must be free flowing without any lumps. The student group concluded that the dried sawdust material met this criterion after Chemical Engineering Education

PAGE 87

85 minutes at the end of the constant-rate drying period with a moisture content of 0.8 g Hp!g dry solid. Figure 4 presents the freemoisture drying curve for a clay absorbent material initially composed of a 3: 1 mixture by mass of absorbent (474 g) and water (158 g). The average material thickness in the tray was 0.375 in. A logarithmic regression on the falling-rate data after 31 min resulted in a value of the multiplicative constant of -61.6 min with a correlation coef ficient of 0.953. The corresponding solid-phase mass transfer coefficient is kx = 0.427 gmol/ft2-min. The specification on the dried absorbent product was that it must be dry enough to package. The student group concluded that the dried absorbent met this criterion after 60 minutes with a moisture content of 0.075 g Hp!g dry solid. The justification of this decision was that the drying rate essentially goes to zero after this time result ing in little further drying being possible without a very long additional exposure. SCALE-UP PROCESS DESIGN The scale-up process design is based on manufacturing a speci fied production rate of some product from the wet solid material with a specified initial moisture content. The final moisture content of the material must be determined by the student group based on a subjective performance criterion as indicated in the initial laboratory handout. Examples of this criterion are that the solid must be dried to the point that it is free flowing or dry to the touch. This subjective criterion requires both experimental data and engineering judgment to determine the final moisture content. The intent is to demonstrate that product specifications are often not directly measured physical quantities. Because the mass transfer coefficient is a function of the operating conditions of the dryer, the scale-up 0.325 process operation must not deviate 0.3 significantly from the experimental 0.275 operating conditions if the experimental mass transfer coefficient 0.25 is to be used in the design. This 0.225 restriction does impose limitations -12' on the scale-up design depending 9 0.2 on the experimental conditions Q) 0.175 ::i that were considered. Specifically, u5 a 0.15 the material depth in the tray and Q) Q) 0.125 the air velocity in the scale-up deu::: sign must be representative of the 0.1 experimental conditions used to 0.075 determine the mass transfer coefficient. For example, the air velocity 0.05 is determined by the air flow rate 0.025 and cross-sectional flow area. To change the driving force for mass 0 transfer by changing the air flow 0 5 10 also be changed to maintain a similar air velocity. A benefit of the design aspect in this experiment is the exposure to this relationship between experimental investigation and scale-up design through hands-on experience. Such expo sure is not available in the process design course because of the lack of an experimental component. A further objective of the design aspect is the develop ment of a physically realistic process design. Although most student groups have little difficulty in determining the required surface area and air rate for the scaled-up process, the actual physical design of the dryer can often be unrealistic. A common initial approach is to scale up the experimental apparatus to handle the required production rate. The result is a design with trays that are often too large and heavy to be physically managed. More realistic process designs evolve as the student groups are prompted to consider the size and weight of the material that must be handled. An example initial approach for the scale-up design of a wood-chip dryer with a production rate of 1 ton/day consisted of a batch cabinet dryer using a single 50-ft-long tray containing almost one-half ton of wet wood chips. The final design was also a batch cabinet dryer, but instead consisted of ten 5-ft-long trays stacked on top of each other where each tray initially contained approxi mately 120 lbs of wet wood chips. The consideration of the practical aspects of a process design is an additional benefit of this experience. PRESENTATION OF RESULTS The experimental and scale-up process design results for each student group are reported in a formal written Measured Free Moisture o Constant Rate Linear Regression Falling Rate Logrithmic Regression -------15 20 25 30 35 40 45 50 55 60 65 Time (min) rate, the cross-sectional area must Figure 4. Free moisture drying curve. Summer 2006 229

PAGE 88

report and an oral presentation. The formal written report is due 10 days after the last analysis session for the experi ment. The oral presentations are scheduled over a series of presentation days at the end of the semester. There are also short memo reports required after each laboratory session to document the planning, experimental, and/or analysis results and conclusions. The different materials and scale-up require ments incorporated into this experiment provide each group a slightly different experience that can be shared with the class during the presentation sessions. Each student group is composed of three students and there are three experiments in the senior laboratory course. Therefore, each student in the group takes on the responsibility of group leader for one experiment in the sequence. Each group leader is responsible for the formal written report, oral presentation, and short memo reports on their experiment. The group leader responsibility also includes coordinating the activities of the other group members. When the class is not evenly divisible by three, there will either be one two-member or four-member student group. A two-member student group will not prepare a formal written report and oral presentation for one of the experiments although they will carry out this experiment and prepare the short memo reports. A four-member group will be given two objectives with a separate group leader for one of the experiments. Because of the length of the drying experi ments, they have not been considered for two objectives in a four-member group. STUDENT RESPONSE There are no formal course evaluations for laboratory courses in the chemical engineering department at Villanova University. Student response data for the senior laboratory course is ob tained from departmental surveys administered at the end of the semester. Qualitative assessment of the students' experiences is also based on their comments during and after the experiment. This assessment indicates that the experience has been generally well received by the students. Student comments concerning the drying experiment reveal that the group spent more time on this experiment because of the design aspect, which required more use of the second planning and analysis sessions. CONCLUSIONS The laboratory experiment documented in this article has been developed and implemented over the past two years in the chemical engineering senior laboratory course at Villanova University. Based on the results of informal course surveys, the students have found the experience both challenging and worthwhile, in addition to providing an applied mass transfer and process design experience. The experiment has also pro vided valuable documentation of students' ability to design, conduct, analyze, and interpret experiments for ABET[ 18 l Criterion 3b and their ability to perform as part of a team for ABET[lSJ Criterion 3d. 230 ACKNOWLEDGMENTS The collaboration with Prof. John Myers and Dr. Miriam Wattenbarger in the development of the experimental exercise, the contribution of Elena DeCandia of the class of 2002 in the design and construction of the data-acquisition and control system as part of her undergraduate thesis project, and a cur riculum revision grant to the Villanova University Department of Chemical Engineering from Air Products and Chemical Co. supporting the development of the senior laboratory course are all gratefully acknowledged. I would also like to thank Gloria Benson, Brett Gerasim, and Sebastian Houle of the class of 2005 for supplying their experimental results. REFERENCES 1. Denig, J., E. Wenger, and W Simpson. "Drying Hardwood Lumber," Technical Report FPLGTR-118, USDA Forest Products Laboratory (2000) 2. U.S. Department of Energy, Mineral Processing Technology Roadmap, Office of Industrial Technologies (2000) 3. Clark, J., "Drying Still Being Actively Researched," Food Technology 56(9), 97-99 (2002) 4. Guerrero, M., C. Albert, A. Palmer, and A. Guglietta, "Drying in the Pharmaceutical and Biotechnological Industries," Food Science and Technology International, 9(3), 237-243 (2003) 5. Mujumdar, A., "Research and Development in Drying: Recent Trends and Future Prospects," Drying Technology, 22(1/2),1-26 (2004) 6. McCall um, C., and L. Estevez, "Introducing Process-Design Elements in the Unit Operations Lab," Chem. Eng. Ed., 33(1), 66-70 (1999) 7. Abu-Khalaf, A., "Getting the Most Out of a Laboratory Course," Chem. Eng. Ed., 32(3), 66-70 (1998) 8. Steidle, C., and K. Myers, "Demonstrating Simultaneous Heat and Mass Transfer with Microwave Drying," Chem. Eng. Ed., 33(1), 46-49 (1999) 9. N ollert, M., "An Easy Heat and Mass Transfer Experiment for Transport Phenomena," Chem. Eng. Ed., 36(1), 56-59 (2002) 10. Moor, A., "A Benchscale Drying Laboratory Illustrating Combined Heat and Mass Transfer," ASEE Mid-Atlantic Section Conference, Rowan University, April (2001) 11. Prudich, M., D. Ridgway, and V. Young, "Integration of Statistics throughout the Undergraduate Curriculum: Use of the Senior Chemical Engineering Unit Operations Laboratory as an End-of-Program Statis tics Assessment Course," 2003 ASEE Annual Conference, Nashville, June (2003) 12. Moyers, C., and C. Baldwin, "Psychrometry, Evaporative Cooling, and Solids Drying," Perry's Chemical Engineers Handbook, McGraw-Hill, New York, 7th Ed. (1997) 13. Cooper, J., W Penney, J. Fair, and S. Walas, Chemical Process Equip ment Selection and Design, Gulf Professional Publishing, Burlington, MA, 2nd Ed. (2004) 14. McCabe, W., J. Smith, and P Harriott, Unit Operations of Chemical Engineering, 5th Ed., McGraw-Hill, New York, (1993) 15. Geankoplis, C., Transport Processes and Unit Operations, 3rd Ed., Prentice Hall, Englewood Cliffs, NJ (1993) 16. Treybal, R., Mass Transfer Operations, 3rd Ed., McGraw-Hill, New York (1980) 17. Keey, R., "Process Design of Continuous Drying Equipment," Water Removal Processes, C. King and J. Clark, editors, AIChE Symposium Series, 73(163), 1-11. AIChE, New York (1977) 18. ABET, "Criteria for Accrediting Engineering Programs," Engineering Accreditation Commission, (2004) 0 Chemical Engineering Education

PAGE 89

~S=i curriculum ) ---11111----------ENHANCING THE UNDERGRADUATE COMPUTING EXPERIENCE THOMAS F. EDGAR corresponding author, CACHE Corporation M any chemical engineering departments are wrestling with the following questions: When should computing be introduced to the chemical engineering student? How should computer programming in chemi cal engineering be taught, and how much formal programming instruction on languages such as C should be provided (vs. usage of computing tools such as MATLAB, Mathcad, spreadsheets, etc.)? Is a numerical methods course required and where is it in the course sequence? How many credit hours are needed? Which department teaches it? Should every chemical engineering course include some computing? Since the mid-'80s two approaches have been taken to ward introductory computing for engineers: "CS 101" and the engineering tools approach. The "CS 101" approach was catalyzed by the growth of computer science programs, which provided instruction in computer languages. Over the years the "CS 101" courses have migrated through several programming languages: Pascal, C, C++, and Java. In the engineering branch, software vehicles such as spreadsheets (first Lotus 123, then Quattro Pro, and now Excel), TK Solver, Mathcad, and MATLAB have gradually pushed out program ming languages (primarily Fortran). Programming languages are becoming endangered species in these courses. The "CS 101" branch would claim a number of reasons for existence[ll: engineers should learn fundamental concepts of programming and computer science computing should be taught by computer scien tists, not engineers engineering faculty are not interested in teaching computing languages to their students these courses provide a significant number of stu dent credit hours ( SCH) and budgetary resources There are concrete benefits to an engineering education that incorporates the ability to write computer programs: Students learn what assumptions go into the program, i.e., what the right answer should be what is the input, what is the output clear organization of thought, logic, and calcula tions is required that errors can exist in a program that programming is unforgiving for ambiguities and errors Thomas F. Edgar is the Abell chair in the Department of Chemi cal Engineering at the University of Texas, where he has been a faculty member for 35 years. He is also executive officer of the CACHE Corporation, a nonprofit educational organization that promotes development and distribution of technology-based educational aids for the chemical engineering profession. Copyright ChE Division of ASEE 2006 Summer 2006 231

PAGE 90

232 When it comes to assess1ng computing needs, faculty often confuse what is important for their students with what is important for themselves. The "Engineering Tools Approach" branch believes: engineering students need a solid grounding in problem solving with mod ern computing tools engineering students need the knowledge and tools required in their profes sions engineering computing and problem solving are best taught by engineers there is no room in the curriculum for a separate threeor four-SCH course in programming While the two branches are complementary and most engineering students could benefit from both courses, most chemical engineering program curricula are too con gested to make room for both Many departments no longer require a course in a computer programming language such as Fortran, C, or C++. It has been suggested that teaching computer programming is analogous to teaching plane geometry. It is a way of thinking but you may not have to use it. On the other hand, without some programming ability, engineers are limited by the built-in capability of commercial software without any way to extend it. This has led a number of departments to switch from teaching C++ to using MATLAB as the programming tool. MATLAB is a structured programming language that incorpo rates many elements of Fortran, C, and C++. It allows for modularity, flow control, and input/output control and has the following programming features. 1. Loops: like DO and WHILE in Fortran, MATLAB has for and while 2. Conditional statements: like IF in Fortran, C, and C++; MATLAB has if for testing relational operations 3. Relational operations: like C and C++, MATLAB has the expected suite of <, >, <=, >=, = =,~=.And like C and C++, the result of the operation is I or 0, and can be used outside of a conditional statement 4. Logical operations: like Fortran, C, and C++, relational operations can be strung together with AND(&), OR (I), and NOT(~) 5. Matching: like C, MATLAB has a switch/case syntax for matching string variables, integers, or logicals 6.1/0: not only can a user be prompted for input and then have results output to screen in formatted form (using fprintf, as in C), but MATLAB can read (load) and write (save) to files, in binary or ascii format 7. Modularity: like SUBROUTINE in Fortran or function in C and C+ +, MATLAB allows the user to create user-de.fined functions to be called by a main program; any number of inputs and outputs can be associated with a user-de.fined Junction 8. Error processing: using the try/catch syntax a user can attempt calculation( s) and then gracefully continue execution if an error occurs 9. Array math: like Fortran90, MATLAB transparently accommodates scalar and array math ( i.e., implied FOR loops) Reasons given by faculty for switching from C++ to MATLAB include ease of use and widespread availability due to an inexpensive student version. Because MATLAB is an interpreted rather than compiled language, the user can create (write), debug, and run code in the same environment. The built-in editor can pass code directly to the MATLAB application for execution. Also, MATLAB has a solid graphical interface for creating 2-D and 3-D plots; and plots can be created using appropriate MATLAB code from within a user's program. Based on informal surveys many chemical engineering departments now introduce programming and engineering problem solving in the freshman year. The view is that Chemical Engineering Education

PAGE 91

these subjects are best taught by an engineering department in the context of an application. A typical introductory course has the following outline. problem-solving: engineering method, units, preci sion in calculations symbolic computing: algebra, calculus spreadsheet techniques: solutions to engineering problems, Visual Basic for Application (VEA) in Excel programming fundamentals: data types, program flow, modularity, object-oriented features elementary numerical methods: linear and nonlin ear equation solving, linear regression software tools: Mathcad, MATLAB, Excel When it comes to assessing computing needs, faculty often confuse what is important for their students with what is im portant for themselves. Faculty needs, more often than not, align with their research interests and activities, and these may be disconnected from the needs of their undergraduate students. Also, faculty may have an incorrect impression of the computing needs of professionals by either being out of date or out of touch. Discussions on computing needs do not always proceed on the basis of evidence from alumni and employer surveys. Finally, computing is not part of the daily professional existence of most faculty and is not expected to be. Their computing skills can be oxidized, and most of their computing is carried out by their students. In the area of computing software, there is a noticeable dis connect between industry and academia. The appendix sum marizes a survey of computing practices of recent graduates in chemical engineering, most of whom now work in industry. Typically chemical engineering departments teach the use of MATLAB, Mathcad, Mathematica, or Maple but not the use of spreadsheets. Yet in industry, spreadsheet software (e.g., Excel) is the dominant computer package in use. Of course this may reflect the nature of many calculations that need to be performed by chemical engineers in industry, rather than a need to de-emphasize the teaching of sound numeri cal approaches in universities. Some faculty resist teaching spreadsheets, for example, because it is difficult to analyze the logic in the code, but this appears to be changing with the availability of VBA. Another objection is that a spreadsheet approach can encourage the use of inaccurate or inefficient numerical calculations (no error control, etc.) For complex calculations, it may be better to program spreadsheets using VBA, where programming logic is more transparent. The survey of industrial usage of computing in the appen dix also indicates that less than 50% of recent graduates in chemical engineering actually perform programming on the job (although there is no clear definition of what constitutes "programming" in industry). The use of spreadsheets in chemical engineering practice appears to be increasing. The Summer 2006 application of spreadsheets in university courses may be at tractive because of student-driven usage. David Clough (U. Colorado) and Brice Carnahan (U. Michigan) have developed many examples of spreadsheet applications, as presented at the 2002 ASEE Chemical Engineering Summer School. TEXTBOOKS AND AFFILIATED SOFTWARE The fragmented nature of software tied to leading under graduate textbooks makes integration of computing through the curriculum difficult, e.g., (a) material and energy balances: Felder and Rousseaul 21 EZ Solve; Himmelblau and Riggsl 31 POLYMATH (b) thermodynamics: Sandlerl 41 MathCAD; Kylel 5 1 POLYMATH; Elliott and Liral 61 various programs (c) separations: Wankatl 7 I Aspen ( d) process control: Seborg, Edgar, Mellichampl 81 MAT LAB; Bequettel 91 MATLAB; Riggsuo1_ MATLAB/Ex cel (e) chemical reaction engineering: FoglermI -POLY MATH; Ekerdt and Rawlingsl 121 Octave/MATLAB (f) product and process design: Seider, Seader, LewinmI -Aspen, HYSYS, CHEMCAD, PRO!! In addition to these courses, many departments are teach ing a statistics course, which involves still one more software package such as JMP, SAS, or Minitab. Clearly, using a subset of these textbooks sequentially through the sophomore,junior, and senior years will require a student to learn up to five or more different software packages. Adding software packages from outside of chemical engineering can push the total num ber of packages beyond 10, which becomes problematic for the typical student. It would be desirable to keep the number of software packages below three or four if possible [ note that Excel is only mentioned once in (a)-(f) above althoughitis used with many of the textbooks listed]. But usually textbooks are not chosen because of the bundled software. In addition, departments must address issues of software availability, licensing, cost, and providing software in computer labs vs. student-owned comput ers. A textbook that is closely coupled to a software package, a CD-ROM, or a Web site is clearly an attractive option. TEACHING PROCESS SIMULATORS THROUGH THE CURRICULUM Several departments have found that the difficulty of integration of computing tools mentioned above can be avoided by more extensive use of process simulators. It is quite common to expose students to a commercial simula tor in a thermodynamics or separations course. At Virginia Tech, ChE undergraduates have been using Aspen Plus and Aspen Dynamics to solve problems in all subjects, starting in the sophomore year. It is fairly straightforward to convert a steady-state model in Aspen Plus into a dynamic model (with PID control schemes) in Aspen Dynamics. The applicability 233

PAGE 92

of Aspen Plus to mass and energy balances, thermodynamics (physical and thermodynamic property analysis, estimation and regression), multicomponent separations, reactor design, and process flowsheet simulation is well known. In process control Aspen Dynamics enables students to evaluate con troller tuning, process dynamics, startup and shutdown, etc. HYSYS has similar features, and has been used at Rowan University for analysis in freshman-senior years. Recently, Version 2.0 of a CD-ROM, Using Process Simu lators in Chemical Engineering: A Multimedia Guide for the Core Curriculum,D 4 l has become available. Modules and tutorials are provided for self-paced instruction in the use of the process simulators to solve open-ended problems in courses on material and energy balances, thermodynamics, heat transfer, reactor design, separations, and product and process design. A 110-page document has been prepared for instructors suggesting the best instruction sequence and pro viding exercises and solutions, for each of the core courses (first introduced at the 2002 ASEE Chemical Engineering Summer School). NUMERICAL AND ANALYTICAL APPROACHES IN MODELING OF PHYSICAL BEHAVIOR Historically many engineering courses have been taught from an analytical viewpoint, but a transition is starting to occur in which numerical experiments are being gradually added in fluid flow or heat transfer courses. Problems and experiments should not be so simplified that they are not realistically formulated. Students are normally exposed to idealized fluid flow cases in the curriculum, for which ap plication of theoretical concepts results in a solution of a one-dimensional ordinary differential equation or an algebraic equation. Therefore it is very easy for them to come away with the notion that theory is useless for most real-life situations. Students should be able to select either analytical or numeri cal techniques to solve a problem, hence they should learn the advantages and disadvantages of either approach. Use of more sophisticated numerical tools such as CFD (computa tional fluid dynamics) will reduce the need to make many simplifying assumptions because you do not need as many assumptions to solve the problem numerically. Chemical engineering students should understand that there are both nu merical experiments and physical experiments. In some cases we can make observations from numerical experiments that you cannot see in physical data, but the converse is also true. This does not suggest that all derivations should be replaced by numerical simulation, neither should every experiment be replaced with a simulator. There should, however, be a bal ance of experimental fluid dynamics (EFD), analytical fluid dynamics (AFD), and CFD. To prepare students for industrial practice, there should be a department-level re-examination of the role of detailed analytical solutions. Is the purpose of some of these exercises 234 the preparation of undergraduates for graduate school or in dustry? Today practicing engineers are not expected to carry out complex derivations in project work. Once a fluid flow situation is analyzed theoretically or the governing principles are discussed, that same situation can be visualized using the computer. This visualization of the flow phenomena can significantly facilitate and enhance the learning process, especially for the visual learner. CFD software makes flow visualization easy. Students can simulate flow processes in a transient or steady-state mode. Flow patterns can be displayed via velocity contours, velocity vector plots, or graphs of ve locity profiles. A key element in flow visualization exercises is exploring the effects of different parameters. Using CFD, students can quickly change the size of the pipe, viscosity of the fluid, size of the particles, velocity of the feed, etc., and see the resulting changes in the flow behavior. This type of parametric analysis also ties in nicely with a discussion of di mensionless groups and geometric and dynamic similarity. While computing and visualization can increase understand ing, educators do not want students to view such simulators as black boxes. In the fluid mechanics course, simulations can become a mathematical exercise with little intuition, unless the instructor has the students solve a simple problem by hand first. More work on the software tools is needed, and it is critical to match the software tool to the student's knowledge base. Two specific recent packages that have been developed for educational usage are Flow Lab (a finite volume-based code) by Fluent, Inc., and FEMLAB (a finite element-based code) by Comsol, Inc. Based on a survey by Professor Jennifer Curtis at the University of Florida, about 20 departments of chemical engineering in the United States expose their undergraduate students to CFD software. FlowLab allows students to solve fluid dynamics problems without requiring a long training period. Using carefully constructed examples, Flow Lab allows students to get started immediately without having to spend the large time commitment to learn geometry and mesh-creation skills required by traditional CFD software. Current exercises that have been developed include sudden expansion in a pipe, flow and heat transfer in a pipe, flow around a cylinder, and flow over a heated plate, among oth ers. In addition, professors can create their own examples or customize the predefined ones. FEMLAB provides ready-to-use application modes, where the user can build his or her own model by defining the rel evant physical quantities rather than the equations directly. The software also allows for equation-based modeling, which gives the user the freedom to create equations. FEMLAB's programming language is an extension of the MATLAB language; this feature gives much flexibility to the user. FEM LAB 's graphical interface includes functions for automatic mesh generation of a user-defined geometry. Recently a kturbulence model has been added to its menu of options. Chemical Engineering Education

PAGE 93

Seider, et at.,[ 13 J present the design of configured consumer products, which usually involves 2-D or 3-D simulations. In Chapter 19 (Product Design), momentum and spe cies balances in a 2-D plasma CVD reactor are employed to produce thin Si fihns using CFD packages such as FEMLAB. This illustrates where it is very effective for students to use CFD packages to optimize designs-even without understanding all of the physical and chemical interactions in the transport-reaction processes. Even with these recent advances in educational CFD software, this computing technol ogy has been slow to penetrate undergraduate transport and reactor engineering courses. A 2002 CACHE survey of all chemical engineering departments in the United States on barriers to implementing CFD identified a lack of knowledge concerning available CFD resources, a lack of professor training in CFD, the relative difficulty of use and the long learning curve associated with using CFD software in a given course, and cost of CFD software. VIRTUAL LABORATORY EXPERIMENTS Laboratory courses are evolving, and new directions are being examined at specific universities, combining elements of simulation and also distance learning. In the chemical process industries, the high cost of pilot-scale equipment and operating manpower has led to more reliance on computer-based simulations rather than traditional pilot-scale experiments. During a typical day, the plant engineer works from a control room, or at least behind a computer screen. An engineer rarely is in the field adjusting valve posi tions, flow rates, and temperatures, because that is normally done using the computer interfaces of distributed control systems. The fourth-year unit operations laboratory at Texas Tech University is emulating industrial practice, by providing computer-generated simulations based upon math ematical models for laboratory equipmentY 5 l The unit operations laboratory can familiarize students with safety concerns and operational issues regarding each piece of equipment. Major pieces of equipment include a double-pipe heat exchanger, an am monia gas-absorber packed column, and a cooling tower. The Virtual Unit Operations laboratory (VUOL) complements the existing laboratory to give students a realistic experience with industrial operations. LabVIEW computer interfaces of the VUOL permit students to control the equipment in addition to physically turning valves and checking temperatures. In the Texas Tech course each student operates two physical and two virtual experi ments. Based on preliminary assessment data, students reported that this type oflabora tory class contributed either a great deal or considerably in all areas of ABET criteria a-k. Virtual and physical experiments complement each other and enhance student learning. In addition, there appears to be no significant difference in the student perception to their learning in using virtual vs. actual unit operations experiments, in 18 out of 20 ABET-related skill areas. While students believe both types of experiments are valu able, a total virtual unit operations laboratory would apparently not be well-received by the students. With the physical portion of the lab, students get a feel for the equipment and how it operates. With the virtual portion, the students become familiar with the computer interfaces that are similar to industrial control rooms, and learn to manipulate the equipment via those controls instead of manually turning valves and knobs. They can also explore operating scenarios which are not easily or economically investigated with physical equipment. Web-access of laboratory experiments enables real chemical engineering laboratory equipment to be controlled and monitored interactively by computers that are connected to the Internet, i.e., under the command of users over the Web. This capability is now available in the labs at University of Tennessee-Chattanooga as well as other schools Summer 2006 Today practicing eng1neers are not expected to carry out complex derivations in project work. Once afluid flow situation is analyzed theoretically or the governing principles are discussed, that same situation can be visual ized using the computer. 235

PAGE 94

such as University of Texas-Austin, Columbia University, University of Toledo, and MIT. Such labs permit faculty and students from any university to run Web-connected experi ments at any time of the day or night, any day of the week. The laboratory station computer operates the equipment (pumps, valves, heaters, relays, etc.), collects the data (pressure, tem perature, position, speed, concentration, etc.) and sends it to the Web user. The University ties-is far more difficult. The market may demand or need a new material with a specific set of properties, yet given the properties it is extremely difficult to know which monomers to put together to make a polymer and what molecular weight the polymer should have. Today the inverse design problem is attacked empirically by the synthetic chemist with his/her wealth of knowledge based on intuition and experience. A significant amount of work of Tennessee site is accessed through the Web address , and even includes audio and video of the operating equipment. While is already under way to develop the "Holy Grail" of materials design, namely, effective and powerful reverse-engineering software to solve the problem of going backwards from a set of desired properties to the realistic chemi cal structures and material morphologies that may have these properties. After this is com pleted, a subsequent step would involve how to manufacture the desired new product. computing and All established chemical engineering pro grams are facing increased financial pressure to keep existing laboratory experiments up to date and in satisfactory operating condition. Major operating costs of unit operations lab oratories include maintenance and teaching assistant support. Using highly automated experiments for remote operations will allow a drastic reduction in TA time requirements for those particular experiments. In addition, by sharing the operation of the experiments among several universities, there can be a pro rata reduction in maintenance costs. There is also the opportunity to add experi mental assignments to a lecture class using visualization can increase understanding, educators do not want students to A chapter on Molecular Structure Design in Seider, et al., [l 3 J contains simple optimization procedures using GAMS to determine poly merrepeat units, refrigerants, and solvents that have desired properties using group-contribu tion methods. Eventually, these will be replaced (and augmented) by molecular models. view such simulators as black boxes. Another subject related to product design is the scheduling of batch processes, which can be done using simple simulation techniques, as in BATCH PLUS and SUPERPRO DESIGNER. this technology. In a lecture class, it may be desirable to have students individually or in small groups carry out an experiment, much like a homework assignment; in contrast, a traditional experiment would require continu ous supervision by teaching assistants (e.g., one week of TA time for an entire class). Therefore, using an Internet-based experiment can greatly reduce the time commitment by the TA. It is clear that traditional experiments should remain in the curriculum to give students "hands-on" exposure, but they can be augmented with Internet labs. PROCESS AND PRODUCT DESIGN Historically there has been a process design emphasis in the curriculum that is now transitioning to a dual product and process design emphasis. This means that a framework is needed to make process decisions to make structured products. This has added a performance layer, i.e., not just purity of the product. Given a structure, we can often predict at some level what the properties of the material are likely to be. The accuracy of the results and the methods used to treat them depend critically on the complexity of the structure as well as the availability of information on similar structures. For example, various quantitative structure property rela tionship (QSPR) models are available for the prediction of polymer properties. The inverse engineering design problem, however-designing structures given a set of desired proper236 Hence design of optimal processing can be viewed as "product design" for specialty chemicals. Clearly, spreadsheets and optimi zation packages can also be used for many of these computations. Finally, the use of large databases and software systems, such as ASPEN IPE, for equipment sizing and purchase and installation cost estimation, is becoming common throughout the chemical industries for product and process design. MOLECULAR MODELING A molecular-level understanding of chemical manufac turing processes would greatly aid the development of steady-state and dynamic models of these processes. Process modeling is extensively practiced by the chemical industry to optimize chemical processes. One needs, however, to be able to develop a model of the process and then predict not only thermochemical and thermophysical properties but also accurate rate constants as input data for the process simula tion. Another critical set of data needed for the models is thermophysical properties. These properties include such simple quantities as boiling points and also more complex phenomena such as vapor/liquid equilibria phase diagrams, diffusion coefficients, liquid densities, and the prediction of critical points. A key role of computational chemistry is to provide input parameters of increasing accuracy and reliability to the process simulations. Chemical Engineering Education

PAGE 95

Under the NSF grant, "World Wide Web-Based Modules for Introduction of Molecular Simulation into the Chemical En gineering Curriculum," seven university experts in molecular simulations have developed Web-based modules to facilitate introduction of molecular simulation into the chemical engi neering undergraduate curriculum. These teaching modules can be integrated directly into chemical engineering core undergraduate courses, supplying for the instructor and the student the appropriate linkage material between macroscopic concepts currently taught in these courses and molecular simulations designed to aid student understanding of the molecular underpinnings of the phenomena. Modules are centered around Java Applets that run the molecular simula tions and provide an "experimental" simulation platform for students to explore concepts. In addition, modules contain instructor materials, fundamental tutorials, student problems, and assessment materials. A consistent Web-based interface has been designed that or ganizes all of the material in each module and develops scripts using perl; this eases the job of putting the written material into this common format. The developer of a module must construct simple text files, perhaps with HTML markup that permits inclusion of figures and tables. Then he or she runs the files through the perl script, which adds HTML formatting and links to put the set of files into the common configuration. The files are uploaded to the module site for anyone to access. This site is perhaps best accessed through the Etomica site. Etornica is a Java-based support environment developed for the modules project, which has now been expanded for other applications ( , contact is Professor David Kofke). Following is a list of phenomena and concepts for which modules are completed or planned: Chemical reaction equilibrium Osmosis Diffusion Molecular dynamics Normal modes of a solid Chemical reaction kinetics Dissipative particle dynamics Surf ace tension Crystal viewer Joule-Thomson expansion Self-assembly Chemical potential Multicomponent phase equilibrium Heat transfer Atomic billiards Viscosity CONCLUSIONS AND RECOMMENDATIONS One way to foster renewal of the curriculum is to identify departments where curriculum revision is being carried out Summer 2006 and to evaluate best computing practices and current trends. There may not be one answer because of different constraints under which various universities operate, such as number of faculty in the department and whether computing courses are taught outside the department. Contributions to this article came from nearly 20 universities, so we are aware of local issues. CACHE makes the following recommendations to enhance computing through the curriculum: ( 1) There is increasing pressure on the total number of hours in the curriculum, especially with the addition of life science courses. Departments should continue to re-examine whether a formal threeor four-credit hour computer programming course is required for the chemical engineering degree (vs. teaching how to use software or write m-files in MATLAB.for example). The chemical engineering computing course also pro vides students with a valuable experience in quantita tive problem solving. (2) The number of software tools that implement numerical methods used by students should be minimized; de partmental agreement on software used in each course should be reached within the faculty. Faculty need to reach consensus on how student computing skills can grow systematically through evaluating each course in the curriculum. (3) Courses such as transport phenomena and thermo dynamics offer new possibilities for introduction of computing physical and chemical behavior, such as with computational fluid dynamics or molecular model ing. Process design can add a product design emphasis by using such tools as well. (4) Internet-based and virtual laboratories offer a new means of strengthening the student simulation experi ence in order to reinforce theoretical concepts. (5) To prepare students to optimize process designs, it helps to expose students to process simulators for solu tion of a problem( s) in the core courses of the chemical engineering curriculum. Also, as software develops and product design is added at the senior level, instructors must select from among optimization packages ( such as GAMS), batch process simulators (such as BATCH PLUS and SUP ERP RO DESIGNER), and packages for estimating equipment sizes and installation costs (such as Aspen JFE). The use of comprehensive software packages and databases is common in industrial de sign and needs to be introduced in design courses and used for solution of design projects. ACKNOWLEDGMENTS Contributors to this paper include T.F. Edgar, W.D. Seider, D.E. Clough, J. Curtis, D.S. Dandy, B.L. Knutson, P.R. West moreland, J.J. Siirola, Chau-Chyun Chen, G.V. Reklaitis, R. LaRoche, J.B. Rawlings, E.M. Rosen, D. Kofke, M. Cutlip, M.J. Savelski, and M. Shacham. 237

PAGE 96

REFERENCES 1. Clough, D.E., "ChE's Teaching Introductory Computing to ChE Stu dents-A Modern Computing Course with Emphasis on Problem Solv ing and Programming," ASEEAnnual Meeting, Montreal (2002) 2. Felder, R.M., and R.W. Rousseau, Elementary Principles of Chemical Processes, 3rd Ed., Wiley, New York (1999) 3. Himmeblau, D.M., and J.B. Riggs, Basic Principles and Calculations in Chemical Engineering, 7th Ed., Prentice-Hall, Upper Saddle River, NJ (2003) 4. Sandler, S.I., Chemical, Biochemical, and Engineering Thermodynam ics, 4th Ed., Wiley, New York (2006) 5. Kyle, B.G., Chemical and Process Thermodynamics, 3rd Ed., Prentice Hall, Upper Saddle River, NJ (1999) 6. Elliott, J.R., and C.T. Lira, Introductory Chemical Engineering Ther modynamics, Prentice-Hall, Upper Saddle, NJ (1999) 7. Wankat, PC., Equilibrium-Staged Separations, 2nd Ed., Prentice-Hall, Upper Saddle River, NJ (2006) 8. Seborg, D.E., T.F. Edgar, and D.A. Mellichamp, Process Dynamics and Control, 2nd Ed., Wiley, New York (2004) 9. Bequette, B.W., Process Control, Prentice-Hall, Upper Saddle River, NJ (2003) 10. Riggs, J.B., Chemical Process Control, Fernet Publishing, Lubbock, TX (2001) 11. Fogler, H.S., Elements of Chemical Reaction Engineering, 4th Ed., Prentice-Hall, Upper Saddle River, NJ (2006) 12. Ekerdt, J.G., and J.B. Rawlings, Chemical Reactor Analysis and Design Fundamentals, Nob Hill, Madison, WI (2002) 13. Seider, WD., J.D. Seader, and D.R. Lewin, Product and Process Design Principles, 2nd Ed., Wiley, New York (2004) 14. Lewin, D.R., etal., "Using Process Simulators in Chemical Engineer ing," CD-Rom for Seider, et al., textbook (2004) 15. Wiesner, T.F., and W Lan, "Comparison of Student Learning in Physical and Simulated Unit Operations Experiments," J. Engr. Educ., 195-204, (2003); see also APPENDIX 2003 Computing Survey of Recent Graduates In 1997 the CACHE Corporation carried out a survey of re cent graduates in chemical engineering from three universities to determine how that group used ( or did not use) computing in performing their jobs. Since that time, there have been con siderable changes in the field of information technology. Four universities volunteered to participate for the 2003 survey: Carnegie Mellon University, Clarkson University, McMaster University, and the University of Texas. A Web-based form was used to tabulate the responses using database software. The four universities used different approaches for contact ing their recent graduates, defined as students who graduated during the previous five years (1998-2003). Printed mail, e-mail, and/or Web forms were used depending on the spe cific school. The response rate for the four universities was estimated to be between 20% and 30%, which actually is quite good given the complexity and length of the survey (which took less than one hour to complete). The results of the survey are available in PowerPoint form on the CACHE Web site, . The questionnaire asked for the nature of the work carried out and the degree level of the respondents. No attempt was made to remove current graduate students from the sample even though they are technically not in the workplace. The 238 overwhelming majority of engineers value computing skills as critical to industrial problem solving. About 75% of recent graduates in the survey characterize their work as "technical." Compared to 1997, there was a gradual increase in the use of the computer as a general productivity tool. The personal computer is ubiquitous in all business and engineering work, including standard office tasks, with 70% of respondents using a computer actively at least half the day. For the range of using the computer 3/4 to all day, the percentages doubled from 19% to 44% between 1997 and 2003. Of the respondents, 99% report they use spreadsheets on a daily basis. Faculty have observed that spreadsheets are used by most if not all undergraduates, often with minimal formal instruction in the department. Industry clearly values the use of spreadsheets for a variety of applications based on the percentage of respondents who use them: data analysis (88%), numerical analysis (47%), material balances (25%), economic studies (23%), and other tasks such as financial modeling or emission calculations (17% ). Similar to spreadsheets, database software (70%) has the same level of penetration in daily work usage. It is noteworthy that even with continued improvement of packages such as MATLAB and MathCAD, they are used much more heav ily in academia than in industry (26% ). Numerical methods libraries are only infrequently used (6% ), which illustrates their general decline in popularity since the 1970s. Less than half of the survey respondents use a process simu lator in their work, probably because a growing percentage of students are working in nontraditional industries ( outside the CPI). Even in the CPI, not all chemical engineers are actively using simulators in the performance of their jobs. In 2003 there was more emphasis on and time devoted to training new engineers to use computing in their jobs (compared to 1997). There is a continued reliance by recent graduates on learning new computing skills on their own or with the help of colleagues. This supports the notion that universities should prepare their graduates to "learn how to learn." The amount of formal training to use computing tools continues to be fairly small. A majority of the respondents (78%) replied that under graduates should be exposed to some form of programming. This is not surprising even though a minority of engineers write programs in the workplace. Most people agree that use of programming logic is an important skill, whether it is C++, VBA, or MATLAB m-files. Of the respondents, 38% indicate they write computer programs at work ( compared to 20% in 1997), but it is not clear what actually constitutes programming in the workplace today (is running simulations considered to be programming?). Use of VBA along with spreadsheets is a dominant practice. The growth of usage of VBA to34% of the respondents is animportantdevelopment. C++ leads the rest of the progrannning options (24 % ). 0 Chemical Engineering Education

PAGE 97

$5::j class and home problems ) The object of this column is to enhance our readers' collections of stimulating problems in chemical engineering education. Ideal problems, which may be "open-ended," are those that mo tivate the student either by the novel illustration of a particular principle, or by the elucidation of a difficult concept in a more traditional setting. Practical relevance is encouraged. The text portion of a manuscript (excluding figures) should not normally exceed 10 double-spaced pages (about 2,500 words). Please send manuscripts to Professor James 0. Wilkes (e-mail: wilkes@umich. edu), Chemical Engineering Department, University of Michigan, Ann Arbor, MI 48109-2136. Preliminary ideas may be discussed with Prof. Wilkes before submitting a manuscript. Design of a Fuel Processor System for GENERATING HYDROGEN FOR AUTOMOTIVE APPLICATIONS PANINI K. KOLAVENNU, JOHN C. 'I'ELO'ITE, AND 8RINIVAS PALANKI Florida State University and Florida A & M University Tallahassee, FL 32310-6046 F uel cell power systems for automotive applications have received increased attention in recent years because of their potential for high fuel efficiency and lower emis sions. [ll While there have been significant advances in fuel cell technology, this technology has not seen widespread applica tion in the automotive industry due to the lack of an efficient hydrogen distribution system.[ 2 l One option is to develop a system that utilizes a commonly available carbon-based hy drogenous fuel such as gasoline or methane to generate the necessary hydrogen in situ on an "as needed" basis. In this paper, the objective is to design a fuel processor system that utilizes methane to generate sufficient hydrogen of desired purity, generating 50 kW of power, or enough to drive a small carYl PROBLEM STATEMENT A schematic of the fuel cell system under consideration is shown in Figure 1 (next page) [ 3 l Methane enters the fuel processing system and is converted to hydrogen. Hydrogen enters the fuel cell where it reacts with oxygen to generate electrical power, driving an electric motor. The fuel processing system has a train of three packed-bed reactors: (1) the reformer, (2) the water-gas shift reactor, and (3) the preferential oxidation reactor. Copyright ChE Division of ASEE 2006 Summer 2006 Panini K. Kolavennu has a B. Tech in chemical engineering with specialization in biotechnology from Andhra University, Visakhapatnam, India. He is currently pur suing his Ph.D. in chemical engineering at FAMU-FSU College of Engineering, Tallahassee, Fla. His research interests include fuel cell and fuel processor design and analysis, model predictive control, and adaptive control. John C. Telotte has a B.S. and M.S. in chemical engineering from Tulane Univer sity. He also received a Ph.D. in chemical engineering from the University of Florida. He was a memberofthe chemical engineer ing faculty at the University of Wisconsin and Louisiana Tech University before join ing the FAMU-FSU College of Engineering in 1985. He is currently associate professor of chemical and biomedical engineering. His current research focuses on hydrogen storage problems and design of fuel cell systems. Srinivas Palanki received his B.Tech. in chemical engineering from the Indian In stitute of Technology, Delhi, and his M.S. and Ph.D. in chemical engineering from the University of Michigan, Ann Arbor. He joined the faculty of the FAMU-FSU College of Engineering in 1992. He is currently a professor of chemical and biomedical engi neering. His current research interests are in real-time optimization and nonlinear robust control with applications in the fuel cell and biomedical areas. 239

PAGE 98

Water Tank Combuster Heat Transfer _.("")Mediwn Tank PEMFuel Cell Reformer Switching Control System L I .------. Battery Backup Electric Motor (Load) Figure 1. Schematic of fuel cell system. Based on the Figure 1 schematic diagram, students are required to complete the following tasks: 240 1. Write the mole balance equations for the reformer, wa ter-gas shift reactor, and preferential oxidation reactor. 2. Calculate the volume necessary for 75% conversion in the steam reformer. Assume isothermal operation at 1000 K, with the reactor operating at 5 atm pressure. The flow rate of methane into the reactor is 9 moll min at room temperature, and the ratio of steam to meth ane is 3:1. 3. Calculate the maximum conversion in the water-gas shift reactor that can be obtained at 450 Kand 600 K, respectively, and the minimum volume required. 4. Calculate the volume of the water-gas shift reactor to obtain 90% conversion if: ( a) 20% of the total volume of the reactor is at 600 K and the rest of the reactor is at 450 K. (b) 60% of the total volume of the reactor is at 600 K and the rest of the reactor is at 450 K. 5. The input to the preferential oxidation reactor consists of exhaust from the water-gas shift reactor and air. The amount of air fed should be adjusted such that the amount of oxygen in the air is 2.1 times the amount of CO in the exhaust of the water-gas shift reactor. If the preferential oxidation reactor is operating at a temper ature of 47 3 Kand a pressure of 2 atm, calculate the volume required to bring the concentration of CO to below 100 ppm. Assume 90% conversion in the steam reformer operating at 1000 Kand 5 atm and a conver sion of 90% in the water-gas shift reactor operating at 500 Kand 2 atm. 6. Calculate the flow rate of hydrogen exiting the pref erential oxidation reactor. How does this flow rate change when the flow rate of methane entering the reformer is changed? 7. Energy Balance: Calculate the heat of reaction from the heat of formation for all the reactions, and list out the DATA exothermic and endothermic reactions. Calculate the enthalpies of all the feed and product streams and use this information to complete an overall energy balance for the reactor system. Steam reformer In this reactor, methane is converted to hydrogen and carbon monoxide. Part of the carbon monoxide reacts with water to produce carbon dioxide and hydrogen, and some methane is totally oxidized to carbon dioxide. CH 4 + HzO 3H 2 + CO (1) CO+ HzO CO 2 + H 2 CH 4 + 2Hz0 4H 2 + CO 2 (2) (3) Xu and Froment[ 4 l developed intrinsic rate expressions for steam reforming of methane, accompanied by the water gas shift reaction on a Ni/MgA1 2 0 3 catalyst. The following reaction rate laws were derived: kl p p H, co p2.5 CH 4 H 2 0 H 2 1 [ p3 p l ~=-(l_+_K_c_oP_c_o_+_K_H_P_H_+_K_c_H_P_c_H_+_K_H_o_P_H_o_i_P_H_)_ 2 2 2 4 4 2 2 2 (4) (5) [ p4 p l p p2 H 2 CO 2 p35 CH 4 H,o K H 2 3 f3 = 2 (1 + KcoPco + KH,PH, + KCH4PCH4 + KH,oPH,o /PH,) (6) where rl' r 2 and r 3 are the rates of formation of CO, CO 2 and CO 2 in the reactions represented by Eqs. (1), (2), and (3) respectively. The Pi are the partial pressures of the reactants. The values of the constants are given in Table 1. The adsorption coefficients can be found using the follow ing relations for the respective species ( -6Ho I Ki =A(KJexpl RT 1 J,wherei=H 2 ,CO,CH 4 ,Hz0 (7) The rate constants are given by a similar Arrhenius-type equation. ki = A1(ki)exp(-:;i} where j=l,2,3 (8) The equilibrium constants for the three reactions are given by the following expression, K =ex/A+ Bi I, where j = 1,2,3 (9) J l J T) Chemical Engineering Education

PAGE 99

Water-gas shift reactor In this reactor, most of the remaining carbon monoxide is converted to hydrogen. The following exothermic reaction occurs: CO+ Hp~ CO 2 + H 2 (10) Choi and Stenger[sJ proposed a kinetic model for the water-gas shift reaction on a Cu/ZnO/Al 2 O 3 catalyst operat ing between 400 K to 700 K. The following rate law was developed. (11) where r 4 is the rate of formation of CO 2 in the reaction repre sented by Eq. (10). The equilibrium constant Keq varies with temperature as follows: Keq = exp[ 45;7.8 4.33] (12) The rate constant k 4 follows an Arrhenius type equation as given below: k 4 = A'(k 4 )exp(-:.t) (13) Other constant values used are given in Table 1. Preferential oxidation reactor The stream exiting the water-gas shift reactor may still have significant amounts of carbon monoxide that can poison the Polymer Electrolyte Membrane (PEM) fuel cell electrocatalyst. For this reason, it is necessary to have a preferential oxidation reactor where the carbon monoxide from the water-gas shift reactor is reacted with air to form carbon dioxide. Some of the hydrogen reacts with the oxygen to produce water. CO+(l/2)O 2 ~CO 2 (14) (15) The following kinetic model was taken from Kahlich, et a[.[6J r = k poA2 [2Pa,] S S o, p co r6 = 1.5ksP~,42 [ 2P o, ] Pco (16) (17) where rs represents the rate of formation of CO 2 in the reaction represented by Eq. (16), and r 6 represents the rate of forma tion of Hp in the reaction represented by Eq. (17). The rate constant ks follows an Arrhenius-type equation: ks= A'(ks)exp(-:.;:s) (18) SOLUTION Each reactor is modeled as an isothermal plug-flow reac tor. It is assumed that no axial mixing or axial heat transfer occurs. This implies that the reactors are operating at high Summer 2006 TABLE 1 Kinetic Parameters for the Three Reactors Parameter Value A, 29.3014 A2 -4.35369 A, 25.225 A' (k 1 ) 9.886Xl0 16 [mo! atm 05 /(m 3 min)] A'(~) 4.665Xl0 7 [mo! atm1 /(m 3 min)] A' (Js) 2.386Xl0 16 [mo! atm 05 /(ltr min)] A (KH,) 6.209X109 [atm1 ] A(Kco) 8.339X105 [atm1 ] A (KH, 0) 1.77X 10 5 A (KCH,) 6.738X104 [atm1 ] B, -26248.4 [K1 ] B2 4593.17 [K1 ] B, -21825.28 [K1 ] E,, 240.1 [kJ/(mol K)] E,2 67.13 [kJ/(mol K)] E,, 243.9 [kJ/(mol K)] 6.Ho -82.90 [kJ/(mol K)] H, 6.H~o -70.65 [kJ/(mol K)] 6.H~ o +88.68 [kJ/(mol K)] 2 6.H~H -38.28 [kJ/(mol K)] 4 A' (k 4 ) 6.195Xl0 8 [mo! atm2 /(m 3 min)] A' (k 5 ) 2.333Xl0 11 [mo! atm04 /(ltr min)] E, 47.53 [kJ/(mol K)] E,s 71 [kJ/(mol K)] Peclet numbers for both heat and mass transfer. A more de tailed analysis incorporating these diffusive effects has been conducted by Bell and Edgar.[7l The automotive application puts a constraint on the total volume of the reactor train, since the entire system has to fit under the hood. The design equations are solved numerically in MATLAB. The process is also simulated in the process simulator CHEMCAD. The results are given below. Mole Balance Equations The general mole balance equation for a PFR is given by: dFi --~ 0~ dV J where j represents the species present in the reactor. It is necessary to determine the reaction rate for each species in the three reactors. 241

PAGE 100

0.7 0.6 0.5 .g 8 0.4 ., 0 ::S 0.3 \ .... ~--~ 2 --..... ___ --------------3 4 5 6 Volume (liters) (a) MATLAB 0.8 ~-----------------~ 0.7 0.6 0.5 0.4 U 0.3 Volume (liters) ---+-H2 - CH4 ---.co CO 2 HO 2 (b) CHEMCAD 7 Figure 2. Concentration profiles in reformer. 100 90 80 70 0: 60 ~ il 50 :> 0: 0 u 40 30 20 10 0 0 50 100 600 K 450 K 150 Volume (Liters) 200 Figure 3. Conversion of carbon monoxide in water-gas shift reactor. 250 In the reformer, the reactions taking place are represented by Eq. (1 )-(3). The reaction rates in terms of the species involved can be expressed in terms of the reaction rates represented by Eq. (4)-(6) as shown below. fCH4 = -fl f3 (20) rco = f1 f2 (21) fco 2 = f2 + f3 (22) rH,o = -r1 f2 2r3 (23) rH, = 3r 1 + r 2 + 4r 3 (24) There is only one reaction occurring in the water-gas shift reaction as shown by Eq. (10). We can express the reaction rate of each Eq. (10) species in terms of the reaction rate of 242 Eq. (11) as follows: (25) (26) (27) (28) In the preferential oxidation reactor, the reactions taking place are represented by Eq. (14)-(15). The reaction rates, in terms of the species, involved can be expressed in terms of the reaction rates represented by Eq. (16)-(17). r 02 = -0.5r 5 0.5r 6 (29) fco = -rs (30) fco 2 = fs (31) rH,o = r6 (32) rH, = -r6 (33) Volume of Steam Reformer Rate expressions for the different reactions are given in terms of partial pressures of reacting species. The given mo lar feed rate of the gases, F, should be converted to partial J pressures. Using the molar feed rates, we can calculate the mole fraction of the feed used to calculate the partial pres sures as follows: F X=---1.. J FT pj =XlT (34) (35) The partial pressure obtained is substituted into the rate expression to calculate the change in flow rate along the volume of the reactor. Chemical Engineering Education

PAGE 101

TABLE2 Effect of Temperature on WGS Reactor Volume the mole balance equations were integrated numerically for a variety of reactor volumes till the maximum conversion was attained. The conversion of carbon monoxide as a function of reactor volume at two different isothermal conditions (450 K and 600 K) is shown in Figure 3. If the reactor is operated isothermally Temperature(K) Volume (I) Conversion CHEMCAD 20%at600K 15.28 90% 60%at600K 19.74 90% 35 A 30 ] ell '-' 0 25 s 'O '-' 20 () ;::i 'O 0 !-< A. 15 A '-' Oil 0 'd >, ::r:: 5 5 Methane flow rate ( moles/min) Figure 4. Hydrogen produced as a function of methane feed. 9 The mole balance equations for the species in the steam reformer are given in the mole balance equation section. A volume for the reformer was estimated, and the mole bal ance equations were integrated numerically in MATLAB for this guessed volume using the initial conditions given in the problem statement. The methane concentration exiting the reformer was computed, and the methane conversion was calculated. The reactor volume was iteratively adjusted using a secant method until the computed conversion was 75% as specified in the problem statement. Students found a reformer volume of 6.38 liters resulted in the specified conversion. The concentration profiles, as a function of reactor volume, are shown in Figure 2(a). This reactor configuration was also simulated in CHEM CAD. Shown in Figure 2(b), the reformer with a volume of 6.38liters resulted in a conversion of 75%. Volume of Isothermal Water-Gas Shift Reactor The mole balance equations for the species in the water-gas shift reactor are given on pp. 241-242. This reactor was simulated in CHEMCAD and MAT LAB, producing identical results. The specific reaction rate constants were computed for the temperatures given in the problem statement first. Using the exit conditions of the re former as the initial conditions of the water-gas shift reactor, Summer 2006 Conversion MATLAB 90.14% 90.15% at 450 K, the maximum conversion possible is 98.8%. The minimum volume required to obtain this conver sion is about 250 liters. The maximum possible conversion at a temperature of 600 K is only 86.6%, and the minimum volume required is 9.4 liters. Volume of Two-Temperature-Zone Water-Gas Shift Reactor At low temperatures, the reaction is kinetically limited. At high temperatures, the reaction conversion is limited, i.e., the extent of reaction is limited by the thermodynamics. To mini mize the volume, the reactor is operated at a high temperature (600 K), and to increase the conversion it is operated at a lower temperature ( 450 K). The reactor volume for 90% conversion under two different reactor temperature regimes is calculated in CHEMCAD, and verified with MATLAB. The results are shown in Table 2. It is observed that the reactor volume can be substantially reduced as compared to the isothermal operation if 20% of the reactor is operated at 600 K. Volume of Preferential Oxidation Reactor The last reactor in the series is the preferential oxidation reactor. Here, the concentration of CO is brought to less than 100 ppm. Along with the combustion of CO, some hydrogen is also burned. For the gases that have a 75% conversion in the reformer and a 90% conversion in the water-gas shift reactor, the amount of conversion in the preferential oxidation reactor (PROX) is approximately 98.7%. The volume obtained for this conversion is 0.335 liters. Hydrogen Flow Rate From calculations performed in the section on "Volume of Preferential Oxidation Reactor," the flow rate of hydrogen exiting the preferential oxidation reactor is 28.08 moles/min. Using the reactor volumes given above, the reactor train was simulated for varying methane flow rates. A plot of hydro gen flow rate out of the fuel processor vs. methane flow rate into the reformer is shown in Figure 4. Note that the rela tion between the two flows is linear, and 3.12 moles of hydrogen are formed for every mole of methane entering the reformer. Energy Balance The standard heat of reaction can be obtained from the standard heat of formation of individual species involved in the reaction using Hess's Law. The standard heat of formation can be obtained from the NIST Chemistry Web book.[ 81 For oxygen and hydrogen, the standard heat of formation ( at 298 243

PAGE 102

TABLE3 Standard Heats of Formation Species Standard Heat of Formation (kJ/mol) co -110.53 CO 2 -393.51 H 2 O -242 CH 4 -74.5 TABLE4 Standard Heats of Reaction and Type of Reaction Standard Reaction Heat of Type Reaction (kJ/mol) CH 4 + H 2 O ---"3H 2 + CO 205 endothermic -sCO+Hp CO 2 +H 2 -41 exothermic CH 4 + 2H 2 O 4H 2 + CO 2 164 endothermic co+ (1/2)02 ---"CO 2 -283 exothermic -sH 2 + (1/2)0 2 H 2 O -242 exothermic K) is zero. The heat of formation for other species is given in Table 3. Using heat of formation data, the standard heats of reaction were found. These results are given in Table 4. There are two feed streams going into the reactor train and one product stream coming out. The enthalpies of these streams can be computed easily in CHEMCAD. The enthalpy of the stream containing steam and methane is -8.261 MJ/min, the air stream is -0.003 MJ/min, and the product stream is -6.042 MJ/min. Air is assumed to be 79% nitrogen and 21 % oxygen. Enthalpies are calculated based on the same reference states as those used for the heat of reaction. The net heat to be supplied to the fuel processor is 2.22 MJ/min. DISCUSSION The purpose of this work was to demonstrate that standard reaction engineering principles could be utilized to design a fuel processor that uses methane to generate hydrogen of the required purity for a PEM fuel cell used for automotive applications. The total volume required for the fuel processor to generate hydrogen of the required purity is small enough to fit under the hood of small cars. Furthermore, the energy bal anced reactions involved indicate that there is a net depletion of energy, and so it is necessary to provide a small amount of heat to the reactors. This project gives the instructor the opportunity to discuss several reactor design issues such as: 244 The importance of changing the temperature in the water-gas shift reactor to minimize the volume. This is a common approach when one looks at equilibrium limitations in exothermic reactions. The necessity of having the preferential oxidation reactor. While the reactor volume is small (the size of a soft drink can), it is necessary to have this reactor so that the concentration of carbon monoxide is reduced to a level acceptable for the PEMjuel cell. The importance of doing an overall energy balance on the process. This exercise demonstrates it is necessary to burn some of the methane feed or the hydrogen generated for the reactions to proceed at the specified conditions. For every mole of methane utilized, more than 3 moles of hydrogen are produced. This demonstrates that hydrogen is produced not only from methane, but also from water. The equivalence of results in MATLAB and CHEM CAD for reactor design calculations. This problem was given as a reactor design problem in our department in summer 2005. All design teams were ultimately successful in producing an acceptable design report for this problem. A discussion with the students at the end of the semester indicated a great deal of enthusiasm in tackling this problem. Students appreciated that a practical design problem had been assigned instead of a problem from a textbook. All the reactor design calculations were done assuming steady state. In an automotive application, however, the de mand for hydrogen will fluctuate, so it is necessary to compute the dynamic response of the reactor train to sudden changes in methane flow rate and utilize this information to design a suitable control system. This requires the solution of a set of partial differential equations,Pl and is beyond the scope of an undergraduate class in reaction kinetics. NOMENCLATURE A (K) pre-exponential factor Ea Activation Energy, kJ/mol k rate constant K equilibrium constant or adsorption coefficient P partial pressure, Pascals r reaction rates, mol/min/1 R gas constant, kJ/mol/K T temperature, K REFERENCES 1. Zaic, J.M., and D.G. Loffler, "Fuel Processing for FEM Fuel Cells: Transport and Kinetic Issues of System Design," J. Power Sources, 111, 58-64 (2002) 2. Lovins, A.B., and B.D. Williams, "A Strategy for the Hydrogen Transi tion," 10th Annual U.S. Hydrogen Meeting, Vienna, VA (1999) 3. Kolavennu, PK., J.C. Telotte, and S. Palanki, "Modeling and Control of a Fuel Cell Power System for Automotive Applications," European Symposium for Computer Aided Process Engineering-14, Lisbon, Portugal (2004) 4. Xu, J.G., and G.F. Froment, "Methane Steam Reforming, Methanation And Water-Gas Shift 1. Intrinsic Kinetics," AIChE J., 35(1), 88-96 (1989) 5. Choi, Y., and H.G. Stenger, "Water-Gas Shift Reaction Kinetics and Reactor Modeling for Fuel Cell Grade Hydrogen," J. Power Sources, 124, 43 2-43 9 (2003) 6. Kahlich, M.J., H.A. Gasteiger, and R.J. Behm, "Kinetics of Selective CO Oxidation in H 2 -Rich Gas on Pt/Alp 3 "J. Catalysis, 171, 93-105 (1997) 7. Bell, N.H., and T.F. Edgar, "Modeling of a Fixed-Bed Water-Gas Shift Reactor," J. Proc. Cont., 1, 22-31 (1991) 8. 9. Larminie, J., and A. Dicks, Fuel Cell Systems, Wiley, New York (2000) 0 Chemical Engineering Education


xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID E2BCAK6ZQ_5IGO3D INGEST_TIME 2011-09-09T21:37:12Z PACKAGE AA00000383_00167
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES