Chemical engineering education ( Journal Site )

Material Information

Chemical engineering education
Alternate Title:
Abbreviated Title:
Chem. eng. educ.
Physical Description:
v. : ill. ; 22-28 cm.
American Society for Engineering Education -- Chemical Engineering Division
Chemical Engineering Division, American Society for Engineering Education
Place of Publication:
Storrs, Conn
Publication Date:
annual[ former 1960-1961]


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


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:
oclc - 01151209
lccn - 70013732
issn - 0009-2479
lcc - TP165 .C18
ddc - 660/.2/071
System ID:

This item is only available as the following downloads:

Kenneth R. Jolls, of Iowa State University, Janet Rohler Greisch ( PDF )

SUNY Buffalo, David A. Kofke ( PDF )

Separations: What to Teach Undergraduates, P.C. Wankat, R.P. Hesketh, K.H. Schulz, C.S. Slater ( PDF )

Book Reviews ( PDF )

A Field Guide to the Excess Functions, M.M. Abbott, M.V. Ariyapadi, N. Balsara, S. Dasgupta, J.S. Furno, P. Futterko, D.P. Gapinski, T.A. Grocela, R.D. Kaminsky, S.G. Karlsruher, E.W. Kiewra, A.S. Narayan, K.K. Nass, J.P. O'Connell, C.J. Parks, D.F. Rogowski, G.S. Roth, M.B. Sarsfield, K.M. Smith, M. Sujanani, J.J. Tee, N. Tzouvaras ( PDF )

A Bioreactor Experiment for the Senior Laboratory, Michael L. Shuler, Naheed Mufti, Michael Donaldson, Ronald Taticek ( PDF )

Exothermic CSTRs: Just How Stable are the Multiple Steady States? M. Shachman, N. Brauner, M.B. Cutlip ( PDF )

Meet Your Students: 5. Edward and Irving, Richard M. Felder ( PDF )

The M.I.T. Practice School: Intensive Practical Education in Chemical Engineering, Barry S. Johnson, Thomas A. Meadowcroft, Aleksander J. Franz, T. Alan Hatton ( PDF )

Designs to Demonstrate the Critical State, Ronald E. Marcotte, Luis C. Zepeda, Dale L. Schruben ( PDF )

Mathematical Modeling of an Experimental Reaction System, Aziz M. Abu-Khalaf ( PDF )

An Inegrated Design Sequence: Sophomore and Junior Years, Richard C. Bailie, Joseph A. Shaeiwitz, Wallace B. Whiting ( PDF )

Error Bars in Process Simulation, D.W. Thompson ( PDF )

Beware the Use of an Ideal Gas, Alan J. Brainard ( PDF )

Teaching as an Exercise in Project Management, Robin A. Chaplin ( PDF )

Assessing Student Presentation, David W. Edwards ( PDF )

A First-Year Introductory Seminar, Kevin Myers, Lawrance Flach, Amy Grosjean ( PDF )

A Pragmatic Approach to Grading Student Reports, Manfred Fehr ( PDF )

Full Text

aow aSg a te gnierit

an .h at ..

S.U.N.Y. BiTxfifa

t Statement of Ownership,
Management and
ftRequred by 39 U S C 3685)


Quarterly 4 See Attached

CHEMICAL ENGINEERING EDUCATION, Room,317, Chemical Encineeritn Department, Univers ty
of Florida, Gainesvllle, Alachua, Florida 32611-6005
Sc-",A l.,e Add.AI, ,ea.d.-a.. ., NA' 'o., ea.,",h .ao,,
Chemical En eete 8te Ditsian, A ertcan Sacety for fEni.eeritg Educattion,
11 DuPont Circle, Washington, DC 20030
6 II adc C uMa.ngtAdd.esoN Publn.At it..ot and.M.n.aoeaeEAora. (i ...ta orte..O

ASEE Chemical Engineering Division, 11 DuPont Circle, Washington, DC 20030
Edimi Mum, ui Captlr Mui. rew
Ray W. Fahien. Chem. Eng. Dept., Room 319. University of Florida, Gainesvlle, FL 32611

Carole C. Yocum, Chem. Eng. Dept., Room 317, University of Florida. Gainesville, FL 32611

Ful. N.... o,.. M.1. Au. .
Officialh blocain of Publisher as nY d g

l i s ondte oge *nde aeht sttunty Ho a o i Petcan 0 M or o al amount o. aends Mortunes or 0ver


can, icfeurNI c ut a cot fa ae 5.25

C 1 1. d. 2194 1980

a Tom 2 P ,ud 0 ceaot a 2198 198

O FtoADIANNAN ,o MN Caiei a heiMNN 48.25 49

raE T ciI ,ici.ln a.. c c a 1 i 2242.25 2029

ttNNNoa, AiN, a Noti,,l83.23 71

; R ,." ,,o N o- i g. ." r 0 [ 0 -

G TOTALNa I Fld--,UdJnn.. .. "" __ 2337.5 2100

Scerify that the statement made by
me above ae correct end compete
n 1.,- 3526. J,.i,

Chemical Engineering Education
Department of Chemical Engineering
University of Florida
Gainesville, FL 32611
PHONE and FAX: 904-392-0861
Ray W. Fahien
T. J. Anderson
Mack Tyner
Carole Yocum
James 0. Wilkes and Mark A. Burns
University of Michigan
William J. Koros
University of Texas, Austin

E. Dendy Sloan, Jr.
Colorado School of Mines
Gary Poehlein
Georgia Institute of Technology
Klaus Timmerhaus
University of Colorado
George Burnet
Iowa State University
Anthony T. DiBenedetto
University of Connecticut
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
J. David Hellums
Rice University
Angelo J. Perna
New Jersey Institute of Technology
Stanley I Sandler
University of Delaware
Richard C. Seagrave
Iowa State University
M. Sami Selim
Colorado School of Mines
James E. Slice
University of Texas at Austin
Phillip C. Wankat
Purdue University
Donald R. Woods
McMaster University

Winter 1994

Chemical Engineering Education

Volume 28

Number 1

Winter 1994

2 Kenneth R. Jolls, of Iowa State University, Janet Rohler Greisch
6 SUNY Buffalo, DavidA. Kofke
12 Separations: What to Teach Undergraduates,
P.C. Wankat, R.P. Hesketh, K.H. Schulz, C.S. Slater
52 An Integrated Design Sequence: Sophomore and Junior Years,
Richard C. Bailie, Joseph A. Shaeiwitz, Wallace B. Whiting
74 A First-Year Introductory Seminar,
Kevin Myers, Lawrance Flach, Amy Grosjean

18 A Field Guide to the Excess Functions, M.M. Abbott, M. V. Ariyapadi, N.
Balsara, S. Dasgupta, J.S. Furno, P. Futterko, D.P. Gapinski, T.A.
Grocela, R.D. Kaminsky, S.G. Karlsruher, E.W. Kiewra, A.S. Narayan,
K.K. Nass, J.P. O'Connell, C.J. Parks, D.F. Rogowski, G.S. Roth, M.B.
Sarsfield, K.M. Smith, M. Sujanani, J.J. Tee, N. Tzouvaras
30 Exothermic CSTRs: Just How Stable are the Multiple Steady States?
M. Shacham, N. Brauner, M.B. Cutlip
58 Error Bars in Process Simulation, D. W. Thompson
62 Beware the Use of an Ideal Gas, Alan J. Brainard
68 Teaching as an Exercise in Project Management, Robin A. Chaplin
70 Assessing Student Presentations, David W. Edwards
78 A Pragmatic Approach to Grading Student Reports, Manfred Fehr

24 A Bioreactor Experiment for the Senior Laboratory,
Michael L. Shuler, Naheed Mufti, Michael Donaldson, Ronald Taticek
48 Mathematical Modeling of an Experimental Reaction System,
Aziz M. Abu-Khalaf
36 Meet Your Students: 5. Edward and Irving, Richard M. Felder
38 The M.I.T. Practice School: Intensive Practical Education in Chemical
Engineering, Barry S. Johnson, Thomas A. Meadowcroft,
Aleksander J. Franz, T. Alan Hatton

44 Designs to Demonstrate the Critical State,
Ronald E. Marcotte, Luis C. Zepeda, Dale L. Schruben

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 32611-2022. Copyright 1994 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 copies replaced
if notified within 120 days of publication. Write for information on subscription costs and for back copy costs and
availability. POSTMASTER: Send address changes to CEE, Chemical Engineering Department., University of Florida,
Gainesville, FL 32611.

M educator


VCenneth R. Jolls

Iowa State University
Ames, IA 50011-3060

If Kenneth R. Jolls were the main character in
a novel, it would probably tell about a child
who took a toaster apart and used its compo-
nents to beat an intricate rhythm on the kitchen
table. In real life, the young Kenneth did take a
toaster apart (and couldn't put it back together),
but his musical talent didn't surface until high
school when he picked up a guitar his older brother
had left behind. Still, science and music (specifi-
cally, chemical engineering and jazz) shaped Ken's
novel life.
Ken started as a musician in his hometown of
Raleigh, North Carolina, in the era of big bands
and small combos. His first job, as a high school
sophomore playing guitar in a quartet, paid $9 a
night. Occasionally he worked with a vibraharp
Copyright ChE Division ofASEE 1994

of Iowa State University

player-a relatively new instrument that had its origins in vaudeville. "I
was captivated by the sound and also by the sight of the mallets-
usually two, sometimes four-racing over the bars, so I began
teaching myself to play," Ken remembers. He would eventually cap-
tivate his own audiences with the mallets-sometimes five of them-
dancing over the aluminum bars in complicated patterns. But that
rushes the story.
In the tenth grade, Ken was certain music would be his career. At
school he played the tuba and occasionally solo marimba; in profes-
sional country and commercial dance bands he played the guitar and the
vibraharp. After graduation he enrolled at North Carolina State "be-
cause it was there." He chose nuclear engineering as his major (his
father had encouraged him to "get into something more focused") but
worked part-time as a rural high school band director and played in
clubs on weekends. Then he won a scholarship to the highly regarded
music school at Indiana University, choosing the tuba as his major and
paying his bills by playing baritone horn in the marching band and
writing music for football half-time shows. He soon switched to a music
theory major with a percussion emphasis that included lessons on the
marimba but not the vibraharp, although he owned one by that time.
But bulbar polio intervened, and instead of returning to Indiana in the
fall of 1954, Ken found himself in an iron lung in a Raleigh hospital. As
he recovered he began playing again. One of the groups he played with
was the Duke Ambassadors, Duke University's popular big band, and
that contact with Duke students encouraged him to enroll at Duke. He
earned tuition fees by playing the university's carillon, writing music for
half-time shows, playing with the Ambassadors, writing many of their
arrangements, and, as a senior, leading the group.
Duke also whetted Ken's appetite for academics in general and sci-
ence in particular. So, Duke music degree in hand, he went back to
North Carolina State to earn a second bachelor's degree, this time in
chemical engineering. Dance jobs and even a stint in a country-western
band on a live weekly TV show paid his bills. During his five semesters
at NC State, Ken also sampled the professional world of chemical
engineering through summer jobs at DuPont and Sun Oil.
One of his favorite professors, Kenneth 0. Beatty, suggested graduate
school, and Ken was soon working for Thomas J. Hanratty at the
University of Illinois. "Tom was one of the great formative people in my
Chemical Engineering Education

life," Ken says, "but he wonders to this day how I 'kept the music hidden.' I
didn't think I was hiding it," Ken says-he played frequently during those
years, including once for Bob Hope. "I was playing the piano," Ken recalls,
"and he asked if I could play Thanks for the Memory. My response was,
'What key?'"
Ken was one of the first of Tom Hanratty's students to work on the now-
famous diffusion-controlled electrode technique to measure mass-transfer
rate and shear stress in flowing liquids. In Ken's case, the system was a
packed bed. The apparatus was a glass column six feet tall and twelve inches
in diameter, filled with 11,000 1-inch glass marbles. Most of their research
involved electrochemical measurements, but they also did some visualization
studies using dye to confirm the onset of turbulence. Their work became part
of the basic packed-bed literature and is still cited.
The first year of graduate school was awful, Ken says. "There I was, a
musician, in Harry Drickamer's class, of all places, competing with real
chemical engineers. But when I passed the oral prelim, I realized I would get
that PhD." He did, and a subsequent series of interview trips landed him a job
at the Polytechnic Institute of Brooklyn.
When his new department head, Jim Conti, asked him what he wanted to
teach, Ken replied, "Anything but thermodynamics." But Ken's first course
assignments had him facing thirty-five New York City seniors three times a
week to teach them-thermodynamics. "Fortunately, I had the second edition
of Smith and Van Ness," Ken says, "but I was never more than two or three
pages ahead of the students." John McFeeley, now a senior manager at
Polaroid who considers Ken the "most significant influence in my profes-
sional career," recalls some of Ken's initial teaching devices: "Maxwellian
demons, the Ideal Rubber Band Law, and even the little-publicized Zeroth
Law of Thermodynamics."
A few useful revelations resulted from that year, Ken says: "I realized
I didn't understand thermodynamics; if possible, my students understood
even less about it; and even when I did understand it, I didn't know how
to explain it."
Ken concluded that there had to be a better way to teach thermodynamics,
but he didn't start pursuing it because he had added electronics and instru-
mentation courses to his academic repertoire. A flair for instruments and

circuits, discovered in graduate school, led
him to associate with chemist Howard
Malmstadt who had been presenting "Elec-
tronics for Scientists" courses at Illinois since
the late 1950s. Ken went back to Illinois for
several summers to teach with Malmstadt and
also taught his own NSF summer electronics
courses at Brooklyn Poly while writing a se-
ries on the subject for Chemical Engineering.

Ken also contributed a session on "Harmonic
Oscillators" at Brooklyn Poly's annual Execu-
tive Technical Development Program. His
after-dinner session, scheduled near the end of
three weeks of intense continuing education
courses on statistics, math, and semiconduc-
tors, sounded oppressive to the executives in
the program. They were relieved, however, to
discover that the oscillators were vibraharp
bars, the harmonies were jazz "changes," and
the lecture was on the history of jazz. "I didn't
even mention Fourier," Ken says.

During his New York years, Ken also played
the "wedding-Bar Mitzvah-Scandinavian
circuit." At one memorable dinner-dance in
New Jersey, many of the guests came with
bodyguards, and at another job Ken was play-
ing in the Grand Ballroom of the Waldorf
Astoria while Robert Goulet sang in the room
next door.

Ken, equally hard at work preparing a demonstration of the Joule-Thomson
process (shown above with Ron Cotterman, John Prausnitz's grad student)
or manipulating the mallets (right) for his own and others' enjoyment.
Winter 1994

But then he left the Big Apple for Iowa State University. "I
was getting serious about teaching thermodynamics by this
time," he says. "I didn't think we needed another textbook,
but we did need better ways to explain fundamentally diffi-
cult concepts. We could demonstrate fluid flow in a pipe and
do experiments showing heat and mass transfer, for ex-
ample, but we couldn't demonstrate enthalpy or buy an en-
tropy meter." For him the better way was visual: "I was
convinced that many of the thermodynamic subtleties that
usually bewilder students could be represented graphically,
and that the phase diagram was the basic structure for doing
this. I may not have been very sophisticated in thermody-
namics," Ken recalls, "but I had discovered that textbook
authors liked to draw pictures of thermodynamic surfaces."
In the early 1970s, Ken saw a thesis that used a 3-D

"Gibbs postulated these connections
more than a century before machines were
invented to visualize them," [Ken] marvels. "It's a
stroke of fortune that we now have the
ideal scientific visualization
tool-computer graphics."

plotting program to locate the minimum points of a function
showing the economics of power-plant design, and the con-
nection to using the computer for thermodynamic drawings
became obvious to him. "I wanted to try 3-D pictures, and
the PVT diagram was the simplest I could think of," he says.
Over the next five years he and three students wrote com-
puter programs to draw phase diagrams for pure fluids, using
the geometrical operations associated with projecting 3-D
images onto a 2-D plane. With steam-table data they drew a
variety of thermodynamic functions and published several
papers. They also tackled mixtures. Kenneth Starling at Okla-
homa provided a program that generated mixture data, and
two more students joined Ken to do the graphics.
Ken's drawings were more than just cosmetic; they were
quantitative, and they were helpful in teaching. But they
were still just "a picture of this, and a picture of that," he
says. No conceptual links existed among the drawings until
he discovered Thermodynamics and Its Applications, by
Modell and Reid. The textbook changed the way he thought
about and taught thermodynamics.
Modell and Reid described "a fundamental equation that
sits at the top of the thermodynamic hierarchy and contains
all the information about the variables," Ken explains. The
Legendre transform connects the various fundamental forms,
and the applied functions, such as PVT, heat capacity, and
refrigeration diagrams, are derived from them." To learn
more about this connectivity, he went on leave in 1981 to
work with John Prausnitz at Berkeley.
Meanwhile, Ken hadn't neglected his aural art. During his

early years in Ames he played for Sonny and Cher and for
Helen Reddy, as well as numerous area events. He played in
shows such as Pippin and The Fantastiks, directed the or-
chestra for the musicals Hair and Jesus Christ Superstar,
and occasionally played percussion with the Des Moines
Symphony. At Berkeley he played a Noon Concert in Hertz
Hall with two Bay area musicians.
A letter Ken received at Berkeley changed his approach
to his visualization work. "When two minds separated by
many years ... arrive independently at the same conclusion
(concerning the usefulness of graphical methods)," wrote
North Carolina State chemistry professor Henry Bent,
"perhaps ... there is something compelling to that conclu-
sion. Indeed, the time is probably ripe for someone with the
proper gifts and interests and experience to go back over that
ground covered by Gibbs, from the standpoint of the capa-
bilities of modem computers. You may be precisely the right
person for that project."
That message sent Ken to the library for J. Willard Gibbs'
original accounts of his geometry-based formulation of ther-
modynamics. Gibbs' elaborate descriptions of lines, planes,
and contours described intriguing rationales for the state of a
system and for the tendency-through consideration of ther-
modynamic stability-for those states to change. For the
first time, Ken says, someone had proposed a geometric art
form to show the structure and logic of thermodynamics.
"Although Gibbs used very few drawings, he had an exact
geometrical analogy for phase transitions," Ken says. "It was
based on the energy-entropy-volume surface with its chang-
ing curvatures, and it used a plane rolling over the surface to
reveal not only the property values but also the phase transi-
tions associated with those changes. If you have the funda-
mental equation model for a particular system and the right
kind of geometrical tools, you can roll that plane and deter-
mine precisely all the thermodynamic properties one nor-
mally tabulates. I found that idea fascinating and learned that
other people found it fascinating, too, and had tried to model
it." But their efforts, all done by hand, were imprecise-and
as one-of-a-kind models, gave no hint of the connectivity
Gibbs had described.
Ken was convinced that the Gibbsian approach could ex-
plain how thermodynamic functions are related and thus
remove some of the abstraction that makes it such a difficult
subject. "Gibbs postulated these connections more than a
century before machines were invented to visualize them,"
he marvels. "It's a stroke of fortune that we now have the
ideal scientific visualization tool-computer graphics."
After returning to Iowa State, Ken discovered that Keith
Gubbins at Cornell shared some of his interests in graphics,
so he spent a semester there teaching thermodynamics and
working on the visualization processes. But he found that the
facilities, good as they were, could not accomplish what he
wanted to do.
Chemical Engineering Education

By the time he returned to
ISU, he was quite comfort-
able with the Modell and Reid
approach to thermodynamics
and the concept of using fun-
damental equations to reveal
the various levels of stability
and provide a logical flow of
information for developing
practical tools. Still, interest
in using the geometric ap-
proach for teaching was lim-
ited. "I couldn't understand it,"
Ken says. "All the compli-
cated things we talked about
we could now see-all the in-
teresting thermodynamic ef-
fects we deal with we could
now visualize."
"Thermodynamic diagrams
show many relationships that
otherwise require verbal, nu- Gibbs postulated geometric anal
medical, 'left-brain' process- century before machines, like
ing," Ken says. "I am utterly were invented
convinced, and enough oth-
ers who understand thermodynamics agree, that this is the
way, maybe the best way, to solve the long-standing teach-
ing problem in this difficult subject." As physicist Herbert
Callen wrote to Ken, "your elegant pictures" mean students
will "have a more concrete view of a subject which has
always been hampered by its abstract nature."
Of course, visualization skills are not distributed equally
throughout the population. "I didn't realize when I started
that people learn in different ways," Ken says. "Learning
through visualization was natural for me, and twenty years
ago I assumed it was natural for everyone else, too. Even
though it isn't ideal for everyone, visual learning does work
well for many people. But the pedagogical methods we use
in engineering tend to ignore them."
In 1984, a computer engineering student, John Ries, added
an interactive shell to one of Ken's PVT programs, allowing
users to input data more easily. Computer equipment had
also improved by this time, and color was available. Another
undergraduate student, Dan Coy, created a PC version of the
program, and these changes liberated the process of making
the drawings and changing them to fit specific situations.
Ken named the program "Equations of State," added tutori-
als and documentation, and began taking it to conferences
and writing about it. Although some viewers didn't "get it,"
others thought it was marvelous, and thermodynamics teach-
ers around the world (nearly fifty of them to date) bought it
for use in their classes.
"I knew what was hard about thermodynamics, and I knew
Winter 1994

the o
d to v

what would make the course
better and help students
learn," Ken reflects. "In this
case, I needed the power of
the computer, and I molded
the tool to fit the subject.
I think too often it hap-
pens the other way around:
the computer looks like a
neat tool and someone con-
trives a way to plug it into a
course. Unfortunately, it
may not add anything of sub-
stance; it may not be a genu-
ine improvement in the
course or in the way it's
taught. The judicious use of
technology to do some-
thing substantive in teach-
ing can enhance a good
textbook and even a good
in thermodynamics more than a teacher," he adds. "Too of-
ne above demonstrated by Ken, ten the computer is much ado
visualize them. about not very much. Stu-
dents rave about 'Equations
of State' because they can do things with the computer that
they can't do any other way." As Imre Zwiebel at Arizona
State puts it, Ken's program "makes students think before
they plunge into a problem."
Students also rave about another program Ken developed.
When he began teaching the separations course at ISU, he
realized he couldn't assign some of the most interesting
problems because the computations were too hard. "But with
process simulators like FLOWTRAN or ASPEN to do the
calculations, students could work on these more realistic
examples in absorption, distillation, and extraction," he notes.
Since those solutions almost always end up in graphical form
as stagewise operating diagrams, it seemed obvious to Ken
that he could use computer graphics for the drawings. One of
his graduate students, Deepak Lumba, wrote a program to
extract the numerical results from FLOWTRAN and pro-
duce the stagewise displays; another student, Michelle Nelson,
wrote a program to let students run the simulator interac-
tively. The combination produced "Simulation Graphics."
"I think it's an enormous advance in teaching separations
processes," Ken says. "You can do complicated separations
quickly, repetitively, and in different modes and then see the
results graphically." Students agreed, and Dick Seagrave,
Ken's department head during most of that development
period, told him "Simulation Graphics" was the best thing he
had produced.
And through it all was the music. Ken played at after-
Continued on page 11.




State University of New York
Buffalo, NY 14260-4200

he Chemical Engineering De-
partment at S.U.N.Y. is un-
dergoing great change. Some
mirrors the change that the profes-
sion itself is experiencing, but more
so it reflects the coming of a new
generation to the department. Our
founders succeeded remarkably well
in building a strong teaching and re-
search program, and now our genera-
tion bears the responsibility of con-
tinuing their tradition of excellence.
The changes present new opportuni-
ties as well as new challenges.

The University of Buffalo (UB) was
founded as a private institution in
1846, and Millard Fillmore served as
its first Chancellor. It grew over the
next century, and in 1962 it merged
into the State University of New York
(SUNY) system. Shortly thereafter,
planning and construction began for Furnas Hall, the
a new campus to be located three engineering a]
miles north of the original UB site.
Our department moved to this "North Campus" in 1977, and
it now occupies many of Fumas Hall's ten stories.
The campus continues to expand, most recently with the
dedication of an impressive 17,000-seat stadium, a student
union, a retail complex, and two spacious, modem facilities
for chemistry and fine arts. UB has the highest enrollment of
any campus in the SUNY system and offers the widest range
of academic programs of any public institution in New York
and New England. The Buffalo campus remains the only one
in SUNY to offer degrees in chemical engineering.

t Sul

Chemical engineering was instituted
at UB in 1961, and responsibility for
assembling a faculty was placed with
Joe Bergantz, our first chairman. Joe
recruited aggressively from all sources,
and indeed, three of the department
founders (Sol Weller, Paul Ehrlich, and
Bob Good) were chemists. The role
that they played in shaping the profes-
sional atmosphere that persists today
cannot be overstated. Each retired re-
cently, but they remain active in re-
search and are valued counsel in de-
partment activities.
Other founding members of the
department include Don Brutvan, Ken
Kiser, and Tom Weber. Don has also
retired, but returns regularly to help
with teaching duties. Ken continues
to serve as Associate Dean of Engi-
neering, a title he has held for sixteen
years. Tom is very active and well re-
garded as an educator, and his teach-
ing and service contribute enormously
to the department. He is also the au-
thor of Introduction to Process Dy-
namics and Control, a widely used un-
e of chemical dergraduate text.
Later arrivals were also instrumental
in the department's development. Eli Ruckenstein came from
the University of Delaware after emigrating from Rumania
in 1973. Ralph Yang arrived from Brookhaven National
Labs in 1978, and Vladimir Hlavacekjoined us in 1981 from
Czechoslovakia. All three have made tremendous contribu-
tions to the profession and science of chemical engineering.
Eli holds the rank of Distinguished Professor, and he has
received broader recognition for his many achievements
through election to the National Academy of Engineering.
Copyright ChE Division ofASEE 1994

Chemical Engineering Education

Ralph's very active research program is all the more remark-
able in light of the additional responsibilities he has assumed
as department chair. These three, along with Tom and Ken,
form the senior core of our faculty.
The long-term future of the department rests, of course,
with the junior faculty. The most senior of these is Mike
Ryan, with John Tsamopoulos and Carl Lund rounding out
the ranks of Associate Professor. Following them are no
fewer than five new hires over the past four years: Dave
Kofke, Lakis Mountziaris, Scott Diamond, Johannes Nitsche,
and Deborah Leckband. The eight junior faculty are estab-
lishing solid reputations in teaching and research: four are
National Science Foundation PYI/NYI awardees, and other
awards include an NIH First Award, a Whitaker Foundation
Award, and a Hackerman Young Author Award. All have
been successful in acquiring funds to conduct ambitious
research programs in a wide array of fields.
The success of the junior faculty owes much to the selfless
support and encouragement they receive from their senior
colleagues, and indeed, many significant changes have been
instituted at the behest of the newest arrivals. An atmosphere
of cooperation and camaraderie has blossomed in this en-
vironment, and morale in the department is very high.
Evidence is provided by two multi-investigator grants
awarded to the department by NSF in the past two years;
details follow below.

The State of New York has an unusually strong commit-
ment to providing accessible higher education to its citizens.
In-state tuition remains very low, and consequently a large
majority of our undergraduates are State residents. Fresh-
men arrive with 3.6/4.0 high school averages and SAT scores
averaging 650/545 (M/V). Recent graduating classes have
numbered approximately thirty, but it seems that future classes
will see this figure more than double.
The chemical engineering curriculum at UB is typical of
ABET-accredited schools, but some offerings are less com-
mon. Close relations with local industry allow our students
to gain, as a three-credit elective, the practical experience of
an internship. Each intern spends ten hours per week for one
semester at an industrial site. Lakis Mountziaris and Mike
Ryan have for several years offered undergraduate research
projects on paper- and tire-recycling. Scott Diamond has
taken this notion a step further by offering a university-wide
elective entitled "Biotechnology and Society" which attempts
to prepare the non-technical but educated person to make
informed decisions about some of the complex issues facing
society. The department can boast of ten local and national
teaching awards, including four instances of the SUNY
Chancellor's Award for Excellence in Teaching, the highest
teaching award given by SUNY.
The department is now embarking on an ambitious, col-

In and Around S.U.N.Y.

Left: World-famous Niagara Falls attracts
locals as well as honeymooners from far
and wide.
Right: The area's pastoral beauty is
typified by this scene from UB's south
Below: The Albright-Knox gallery, home
of many treasures of modern art.

Winter 1994

laborative project to advance the use of computers in chemi-
cal engineering instruction. Seven faculty co-investigators
have been awarded funding under the very competitive Lead-
ership in Laboratory Development program in NSF's Divi-
sion of Undergraduate Education. The goal of the project is
to develop a Chemical Engineering Simulation Laboratory
(CESL) that will greatly expand the role of simulation as a
supplement to the laboratory and classroom experiences.
CESL will be an interactive simulation package that in-
stills in students an intuitive feel for chemical processes. The
simulation modules will provide virtual laboratory experi-
ences-the student being presented with "equipment" to be
characterized by "experiments" of their design. More inter-
esting will be situations in which the student operates a
simulated process, drawing on acquired intuition to respond
quickly to unexpected changes. We seek eventually to incor-
porate design elements in which the student must assemble a
process to accomplish a particular task.

We are very proud of the stature our department has ac-
quired for research in the short period of its existence. Each
year the department secures new external funding at a level
of about $1 million, graduates roughly ten PhD and eight
MS students, and publishes nearly a hundred articles in
refereed journals. The graduate program displays a healthy
balance between experimental and theoretical research, and
between fundamental and applied studies.
Catalysis and Reaction Engineering
Research in catalysis and reaction engineering is performed
by Yang, Lund, Ruckenstein, Mountziaris, and Hlavacek.
The efforts of Mountziaris and Hlavacek are materials-ori-
ented, and both of them rely on an unusually well-balanced
combination of theory, computer simulation, and experi-
ments to tackle their research problems. Vladimir Hlavacek
holds the C.C. Furnas Chair and is Director of the Labora-
tory for Ceramic and Reaction Engineering, which was es-
tablished in 1987 with the goal of bridging the gap between
chemical engineering and materials science. One of the
Laboratory's finest achievements is the development of com-
bustion synthesis methods to produce superhard ceramic
materials-nitrides, borides, carbides and the like. In this
technology, the reaction precursors are "ignited" to initiate a
self-propagating but controlled reaction which yields a very
pure, sinterable product without any additional energy input.
Hlavacek's group is also very active in studying the dynam-
ics of nonlinear reacting systems.
Hlavacek and Lakis Mountziaris both have research in-
terests in chemical vapor deposition (CVD). CVD delicately
balances transport processes with both gas- and surface-
phase kinetics to produce solid films from reactive gases.
Hlavacek uses CVD to make non-oxide ceramic fibers and
thick films, while Mountziaris is studying metalorganic CVD

as a means to produce thin films of compound semiconduc-
tors for advanced electronic and optical devices. There are
many novelties in Mountziaris' research, not the least of
which is a counterflow jet reactor for studying the decompo-
sition kinetics of CVD precursor gases.
Mountziaris is also applying CVD to grow diamond films
from hydrocarbons at low pressure. The value of such a
product is obvious, but Mountziaris takes it a step further-
by inserting dopant atoms in the growing film, he is attempt-
ing to produce diamond-based semiconductors; diamond's
electronic and heat transfer properties would make such a
material very useful in electronic devices.
The reaction engineering performed by Ralph Yang and
Carl Lund has a different flavor: it is oriented toward the
catalyst and the catalytic processes. Yang, the holder of the
Praxair Endowed Chair, is studying catalytic reduction of
NO for pollution control from power-plant emissions. Pres-
ently this task is accomplished by selective reduction of NO
with ammonia over a V205 catalyst. Yang has found that
pillared clays are better catalysts and, moreover, he has
developed a new sorbent-catalyst approach to NO reduction
without a reducing gas. Also, his work on carbon-gasifica-
tion reactions continues to be at the forefront of that field.
Lund is enhancing commercial reactions by applying a
strategy perfected long ago by biological systems: selective
removal of reaction intermediates via transport across a mem-
brane. Selective removal of reaction products increases re-
actor yield by exploiting chemical thermodynamics; this
notion is well established and is routinely exploited. Lund's
new approach exploits the kinetic features of the process
instead: timely removal of reaction intermediates prevents
them from participating in additional, undesired reactions
that degrade the product. The net result is an increased yield.
In one of several other projects, Lund is performing tran-
sient kinetic experiments with isotopically labeled reactants
to identify the pathways of coke formation in acid catalysts.
The enormous network of microscopic pores which endow
these catalysts with many wonderful properties also make
them susceptible to deactivation by coking. Lund's research
is needed to develop strategies that will minimize these
coking reactions, and thereby extend catalyst life.
The efforts of Eli Ruckenstein round out the department's
research in catalysis and materials. His research program is
remarkable for its astounding degree of diversity and nov-
elty. Eli has major thrusts in materials, colloids, separations,
catalysis, and bioengineering, and in his career he has made
significant contributions in transport phenomena and ther-
modynamics as well. His encyclopedic knowledge of the
literature forms the intellectual timber for the inspired works
that have earned him such acclaim. Eli's grasp of world
history and politics is equally amazing, but that is a story
(actually many stories, best told over lunch) in itself.
Ruckenstein's present work in catalysis reflects the diver-
Chemical Engineering Education

sity of his entire research program. He is building theories
for the activity and selectivity of catalytic reactions on sup-
ported-metal systems by combining sophisticated experi-
mental techniques with new concepts concerning the active
sites. In other work, he has developed bialkali-promoted
magnesium-oxide catalysts that are proving effective in many
reactions of industrial importance, including the conversion
of methane to ethylene, toluene to styrene, and acetonitrile
to acrylonitrile.
In the area of materials, Ruckenstein has had several break-
throughs in developing composite polymers with useful prop-
erties, such as electrical conductivity or great hardness. An-
other project has produced a new membrane that is highly
permselective yet possesses good mechanical properties. In
still another project, he has combined the Debye model of a
crystal with a Langevin model for diffusion to generate a
new, unified theory for mass transport in solids. His work
with colloidal systems continues, and is now directed to the
study of the properties of foams, focusing on film drainage
and Voronoi analysis of foam structure.
Transport Phenomena
The bulk of the department's research in transport phe-
nomena is done by Mountziaris, Nitsche, Ryan, and
Tsamopoulos. Mountziaris supplements his reaction-engi-
neering research with computational studies of multiphase
flows which display a rich variety of nonlinear phenomena.
Also, he-together with Lund-played a major role in initi-
ating the undergraduate simulation lab project described
above; before that he served as PI of a collaborative proposal
that led to the purchase of a state-of-the-art graphics system
for visualizing complex modeling calculations. Mountziaris,
perhaps more than anyone, personifies the spirit of coopera-
tion that has overtaken this department.
The behavior of drops and bubbles represents the essential
fluid mechanics underlying many multiphase phenomena of
interest to the chemical industry. John Tsamopoulos is
applying asymptotic theories and boundary- and finite-ele-
ment calculations to understand the behavior of these very
complex dynamical systems. His work can be used to de-
scribe the coalescence of bubbles suspended in liquids and
thereby aid in preparing or destabilizing emulsions and dis-
persions. The dynamics become particularly interesting when
a solid surface is introduced: fluids flowing rapidly near the
surface experience a local decrease in pressure, which causes
the formation of cavitation bubbles; subsequent collapse of
these bubbles contributes to erosion of the surface.
Tsamopoulos shares an interest with Mike Ryan in prob-
lems that concern the processing of polymers. Tsamopoulos'
focus is the fundamental fluid mechanics and finite-element
modeling, while Ryan concentrates on the process engineer-
ing. Ryan's experiments with injection molding and
thermoforming processes will help manufacturers predict
the ultimate properties of a finished part, knowing only the
Winter 1994

polymer's material properties and details of the forming
process. In another area, they are working to develop soft-
ware that can simulate the various stages of a blow-molding
cycle. Of interest here is the ability to predict the final wall
thickness of the product (such as a plastic cup): walls that are
too thin may fail, while overly thick walls waste material.
Ryan is also Director of the Business-Industry Affiliates
Program of the New York State Center for Hazardous Waste
Management, where he oversees research on the reduction
of hazardous waste generation at the source. Specific inter-
ests include recycling and reuse of post-consumer scrap
rubber and plastics, biodegradable polymers, and the use of
additives to enhance the degradability of a polymer material.
Arguably the busiest man in the department, Johannes
Nitsche has not let his enormous dedication to teaching
detract from his research program. He has great expertise on
Brownian transport in confined spaces, and he brings to bear
a wide range of mathematical and computational tools to
examine these problems in the context of catalysis and
separations technology. In diffusion of nonspherical par-
ticles in pores, the confining walls couple strongly with
particle hydrodynamics to produce unexpected behavior,
such as anomalous density distributions. Rotational dif-
fusion is a concept that is relatively unappreciated by
many engineers, yet it plays a key role in reactions in
porous media. The phenomenon is especially relevant to
proteins and other macromolecules that are reactive on only
a small portion of their surface. Nitsche is embarking on
several experimental investigations to guide and corroborate
his calculations.
Nitsche possesses several unique and en-
viable gifts, not the least of which is his
muse, Elroy Hutch. Nitsche has been
generous enough to share..
his "musings" with
Elroy with his
graduate and undergradu- Meet "Elroy Hutch," engineer
ate-and his colleagues "par incompetence!"
here at UB and, in
( fact, worldwide [see Fluid Phase
Equil., 78,157 (1992)]. Elroy's an-
\oFF tics have delighted our students in
e unit operations for several years
- now and have taught them sev-
eral lessons of how chemical en-
gineering should not be done.

Biochemical Engineering
The department has recently initiated a major concentra-
tion in biochemical engineering. Scott Diamond and Deborah
Leckband are the principals in the endeavor, along with Eli
Ruckenstein and another faculty member who will be re-

cruited soon. The facility assembled for this work is impres-
sive by any standard. Bioengineering occupies the entire
ninth floor (6000 ft2) of Furnas Hall, and it features seven
fully equipped laboratories: molecular biology, microbial
engineering, cell culture, separations, analytical surface sci-
ence, and video microscopy. Additional facilities, including
three environmental chambers, support these labs.
Mammalian cells exposed to laminar shear stresses (as
are, for example, the cells lining blood vessels) behave dif-
ferently than do cells removed from such an environment.
The response of living cells to mechanical forces is distinct
from the relatively well-understood phenomenon of recep-
tor-mediated signaling. Scott Diamond measures intracellu-
lar concentrations of key biomolecules and is assembling his
findings into a theory for the mechano-biological response.
The relevance of this work to the treatment and prevention
of coronary and vascular disease has been recognized tangi-
bly by the American Heart Association. The research is also
proving useful in the design and operation of bioreactors. In
other work, Diamond is examining the transport and kinetics
of proteolytic enzymes in entangled, cross-linked protein
gels, such as fibrin and collagen. While the work has several
direct applications, Diamond's primary interest relates to the
design of blood-clot dissolving agents and the design of
thrombolytic therapies.
Deborah Leckband's presence strengthens and expands
the department's reputation in colloids and interfaces. She
uses the surface forces apparatus in concert with sophisti-
cated biochemical and surface analytic techniques to probe
the nature of cell- and biopolymer-surface interactions. In
one application, Leckband examines the forces that govern
molecular recognition to guide her development of very
sensitive and localized biomolecular sensors. In another, she
uses her surface-forces measurements to improve protein-
separation techniques, both by chromatography and by parti-
tioning in aqueous solution. She also performs studies of
adhesion at the cellular level, with application to wound
healing, cancer cell metastasis, tissue engineering, and floc-
culation in bioreactors.

Much of the department's work in separations has been
discussed above, but it would be conspicuously incomplete
without highlighting the significant achievements of Ralph
Yang. He complements his work in catalysis with both
fundamental and applied studies of adsorption and adsorbent
materials. The author of Gas Separation by Adsorption Pro-
cesses, Yang is acknowledged as a leading expert in adsorp-
tion and its use as a separation technique. He is now studying
the molecular design and synthesis of new sorbents.

Molecular Thermodynamics
Finally, we come to David Kofke who, when he isn't

woodworking or speaking about himself in the third person,
conducts research in molecular thermodynamics. Com-
puter simulation is regarded by many as the third leg-
along with theory and experiment-upon which we build
our understanding of nature. Most of Kofke's group is busy
developing and applying Monte Carlo and molecular
dynamics simulation techniques-"experiments" on model
molecular systems. Much of their focus is on methods
for evaluating phase equilibria; other topics occupying
Kofke's attention include thermodynamics and transport in
anisotropic systems (such as liquid crystals) and theories to
predict properties of mixtures from data for the respective
pure components.

A local business official recently complained of two major
problems with operating in Buffalo-the number 2 problem
is getting people to relocate here, and the number 1 problem,
he said, is getting them to leave. Buffalo is New York's
second-largest city and is, according to recent rankings, its
"most livable." Housing is inexpensive, the roads are
uncongested, air travel is hassle-free, and (believe it or not)
the climate is moderate-temperatures below 15 or above
90oF are rare. We are also just north of the famed snow belt
beloved by skiers.
The city is rich in culture: the Albright-Knox gallery pos-
sesses one of the world's finest collections of 20th century
art; the Buffalo Philharmonic consistently ranks as one of
the nation's top orchestras; the theatre district, developed as
part of a recent renaissance of the downtown area, provides
opportunities to enjoy the best works of local and national
theater companies. If that's not enough, Toronto, one of the
world's premier cultural centers, is less than two hours away.
But Buffalo's biggest secret is its architecture: the region
boasts of a remarkable number of architectural masterpieces,
including five Prairie houses by Frank Lloyd Wright and
major works by Sullivan, Richardson, and others of similar
stature. Buffalo also posesses a beautiful system of parks
designed by Frederick Law Olmsted. Outdoor recreation
may be found in the waters of Lakes Erie and Ontario, or in
the many splendid woodland areas nearby. And, of course,
one of the world's greatest natural wonders-Niagara Falls-
is a mere twenty minutes from the campus.

The enthusiasm and optimism of the faculty, together with
their substantive accomplishments in both teaching and re-
search, paint a bright future for the department. This article,
we hope, has provided a convincing portrayal of the exciting
developments here. We urge any who are contemplating
their own future, either as a graduate student or an academic,
to give serious consideration to SUNY Buffalo and to the
opportunities that await you here. O

Chemical Engineering Education

Continued from page 5.
concert dinners for three major orchestras (Philadelphia, Bal-
timore, Leningrad). He has played "Jazz in July" concerts in
Des Moines each summer, a public TV special with a Des
Moines personality, and numerous club dates. He hosted a
reception for jazz vibes player Gary Burton, hosted a visit by
pianist and radio personality Marian McPartland, conducted
honors courses on jazz and improvisation, and presented pre-
concert programs about the Moder Jazz Quartet and Marian
And he kept thinking about Gibbs and his geometrical
formulation of thermodynamics, but those drawings were
harder to do. They needed more sophisticated equipment, for
one thing, and in 1986 Ken got it: an NSF grant provided a
Silicon Graphics IRIS 3030-the first advanced graphics
workstation at ISU. On it, graduate student Michael Schmitz
generated the first Gibbs surface-the Helmholtz energy
function for a pure fluid-using Movie.BYU software.
Later, graduate student Day Coy produced the complete
set of four pure-fluid images based on the Peng-Robinson
equation for ethylene. "Those first Gibbs images represented
a fixed amount of mass, but that isn't necessary in thermody-
namics," Ken says. "The mathematical description of a ther-
modynamic system can be scaled in a variety of ways, de-
pending on the ultimate use of the results. Dan later worked
out schemes so that drawings could be made according to all
scalings possible. To some people they seem almost like a
circular set of representations of the same thing through the
power of the Legendre transform. But these are necessary
redundancies that give us more useful forms.
"Dan was able to generalize both the thermodynamics and
the graphics schemes for producing the surfaces. He was
able to comprehend the three major aspects of the project:
the thermodynamic connectivity, the computer programming,
and the visualization. His dissertation is a masterful piece of
work," Ken says. "This is the first time these drawings have
been presented in any connected, quantitative, knowing--
even artistic-way. We thought about color, lighting, per-
spective-what we needed to make the contours convinc-
ing." Ken calls the drawings "the art of thermodynamics."
"If you take the time, you can look at every function in
Modell and Reid and see that 'this comes from that and with
good reason,'" Ken says. "Sometimes those reasons are pretty
subtle, but they're all rooted in the structure of classical
thermodynamics and in stability theory. That's what makes it
all work. If it weren't for thermodynamic stability, the world
would be a single-phase blah," he adds. "Stability theory
says that we have preferred states-things boil, condense,
freeze, form crystals. Dan's dissertation has drawings that
show those ideas."
But it hasn't always been easy. "Funds are hard to get,"
Winter 1994

Ken notes, "and funds for teaching improvements-espe-
cially for teaching improvements that don't involve laborato-
ries-are even harder to come by." Ken has had NSF funding
and grants from the Dreyfus Foundation and Union Carbide.
"I was warned against doing this work," he recalls, "and I
wouldn't advise a new faculty member trying to get tenure to
do it either. To do this stuff right takes a lot of time. And you
can't work in a vacuum; you need support from the depart-
ment and the department head, plus release time. It doesn't
work very well as an add-on."
Academic software needs to be documented, it needs to be
used, and it needs to be reviewed, just like other scholarly
efforts, Ken says. If the work is done right, the intellectual
effort to develop teaching software is no different than the
intellectual effort to do research. "That's not a popular opin-
ion," he adds, "but it's the truth."
Jazz requires intellectual effort, too. "I learned chords when
I started playing the guitar," Ken says. "And because I have a
mind that tends to organize things mathematically, I figured
out what the link was to other musical forms, and I saw
ultimately how jazz fits into general music theory.
"I also have a good ear. A friend in music school once
said I could 'hear around corners,'" Ken recalls. "That basic
ability to detect harmonies is crucial. I was born with the
nerve endings that create a good ear and a mathematical
mind," he adds. "Seeing those traits reproduced in my
son removes any doubt about the genetics," Michael Jolls
works as a programmer analyst in St. Louis, plays guitar
and piano, and does magic.
"I call it a quantitative ear," Ken explains. "I hear harmo-
nies, realize exactly what creates them, and can reproduce
them; I can connect what I hear-or even just what I think
about-to the way to play it. Because of my formal training
in theory, I can also talk about it, and often do-but it's still
easier just to play it."
Jazz on the vibraharp is geometrical, too. "There's no
magic," Ken says. "You just spread the two, four, or five
mallets and turn them so as to connect pairs of points with
straight lines." One of the things that keeps him playing is
that he can. Even though he doesn't practice very much, he
can shift into music mode very quickly. "Sometimes I feel
like I won't be able to find middle C," he groans, "but once I
get the mechanical thing going, I can do about 80% of the
rest off the top of my head."
In jazz, a composer writes chord structures to fit a melody,
and an improviser plays a line of notes that fits those chords
but is different from the melody. That requires a willingness
to take risks, to deviate from the formula, to step out of the
mold and try something new, Ken says. Of course, that's
what he's done in thermodynamics with his quantitative eye-
with his work to express scientific ideas visually. "There
must be a connection," Ken says, "or a very strange coinci-
dence." Well, truth is often more novel than fiction. O

R curriculum


What to Teach Undergraduates

Purdue University
West Lafayette, IN 47907

S eparations (also known as mass transfer or unit op-

erations) have always been an important area in
chemical engineering, but recently separations have
become a critical concern."~41 This concern arises because
separations are typically between 50% and 90% of the plant
cost and because the subject is so important in "hot" areas
such as biotechnology and the environment.
A workshop at the ASEE Chemical Engineering Division
Summer School in Bozeman, Montana, in August of 1992,
wrestled with the problem of what separations should be
taught to undergraduate chemical engineers. The workshop
consisted of an overview of separations, a short presenta-
tion on educational ideas and educational philosophy, and
small work groups discussing their curricula and the ques-
tion of what to teach.

The general consensus of the workshop participants was
that undergraduate students should be introduced to a broad
overview of separations. One way to present this overview
in a concise format is to use the classification shown in
Figure 1. [5 Some of the working groups found this scheme to
be useful for organizing their thoughts-the scheme should
also be useful to undergraduates, particularly global think-
ers. If this or a similar classification scheme is shown, then
each of the separations categories must be discussed. Sev-
eral concrete examples from each separation category should
be presented to the students. Within the category of equilib-
rium separation processes, the concept of a separating agent
is useful.
Two other concepts which are useful for an overview of
separations are the relationship between feed concentration
and selling price16'71 and the wide range of production rates
for different separations (see Figure 2). The Sherwood plots[6'7'
' University of Tulsa, Tulsa, OK 74104-3189
2 University of North Dakota, Grand Forks, ND 58202
SManhattan College, Riverdale, NY 10471
Copyright ChE Division of ASEE 1994

can be used to make the point that more concentrated feeds
are easier to separate. Also, the point can be made that not all
raw material sources are equal. Product rates for separations
vary enormously, as shown in Figure 2.[21 The type of separa-
tion used often depends on product rates. This can easily be
illustrated for oxygen production at different product rates.
For very small units membrane permeation and pressure
swing adsorption compete, for medium-size units pressure
and vacuum swing adsorption units seem to be favored, and
for very large units cryogenic distillation is least expensive.
Our experience is that these concepts are easily under-
stood by undergraduates. The concepts help the students
obtain a perspective on separations and to understand the
similarities and differences between separations. It is useful
to present Figure 1 both at the very beginning of the semes-
ter to illustrate the course structure and at the very end to
help review the material. Figure 1 also helps the students
realize that there are many other separations in addition to
those they studied. If one goal is to produce innovative

Phil Wankat received his BSChE from Purdue and his PhD from
Princeton and is currently a professor of chemical engineering at Purdue
University. He is interested in teaching and counseling, has won several
teaching awards at Purdue, and is Head of Freshman Engineering. His
research interests are in the area of separation processes, with particu-
lar emphasis on cyclic separations, adsorption, preparative chromatog-
raphy, simultaneous fermentation, and separation.
Robert P. Hesketh is an assistant professor at the University of Tulsa.
He received his BS (1982) at the University of Illinois Urbana, and his
PhD (1987) from the University of Delaware. He teaches undergraduate
mass transfer and chemical reactor design, and graduate reaction
kinetics, multiphase reactor design, and combustion. His research inter-
ests include incineration, combustion, and multiphase flows.
Kirk H. Schulz is an assistant professor at the University of North
Dakota. He received his BSChE and PhD in chemical engineering from
Virginia Polytechnic Institute and State University in 1986 and 1991,
respectively. His research interests include surface science and cataly-
sis, with teaching interests in mass transfer, chemical kinetics, and
physical chemistry laboratories.
C. Stewart Slater is a professor of chemical engineering at Manhattan
College. He received his PhD, MPh, MS, and BS degrees in chemical
engineering from Rutgers University. His research and teaching inter-
ests are in separation and purification technology, membrane pro-
cesses, and biotechnology.

Chemical Engineering Education

Electrostatic Separator
Emulsion Separator
High-Gradient Magnetic
Impingement Separator




- Electrodialysis
- Pervaporation
- Gas permeation
- Reverse Osmosis
- Ultrafiltration
- Gas Diffusion

D.L. Isomer Sep.
Ester Production


SKinetic Adsorption
- Molecular Distillation

-Mass Spectrometer
- Pressure Diffusion
-Thermal Diffusion
Isoelectrical Focusing

Ion Ex.

Freeze Drying
Molecular Sieve

Molecular Sieve
SZone Melting
SIon Exchange
Ion Exclusion
Drying Solids

-Absorption Extraction
-Distillation Dual Temp. Exchange
-Azeotropic Liquid Membrane
Foam Fract.

Coal Cleaning Plant
106 Natural Gas Dehydration
10s -- Cryogenic Oxygen Plant
Ethylene Glycol Refining
104 Fermentation Alcohol Plant

tenM Alpha Amylase Fermentation

S10 Kidney Dialysis

10 -
2 1.1 Medical Oxygen Unit
2 10-1
10-2 -
10-2 Various High-Value Proteins

10.-4 -

S Various High-Value Radioisotopes

106 -


Winter 1994

Figure 1 (Above). Classification of Separations. Modified from Reference 5.
Reprinted with permission from P.C. Wankat, Rate-Controlled Separations, Elsevier, Barking, England.
Copyright 1990, Elsevier Science Publishers Ltd.

Figure 2 (Left). Product rates for various Separation Processes.121 Reproduced by
permission of the American Institute of Chemical Engineering from G.E. Keller, III, AIChE Monograph Series,
83 (17), 1987. Copyright 1987, AIChE. All rights reserved.

engineers, then Figure 1 coupled with Figure 3121 are useful. Figure 3 is also
useful in design classes to help explain the strong preference of many compa-
nies for well-known technology. The Sherwood-type plot6'71 can be included
in the course when the economics of separations are covered, or as part of the
overview when Figure 1 is discussed. Figure 2 can also be part of the course
overview, or it can be profitably employed when discussing the use of com-
peting methods for the same separation problem.


The choices of what to teach and how to teach it are heavily influenced both
by which educational methods work and by our philosophy of chemical
engineering education. These topics were briefly covered in the workshop to
give all of the working groups a common basis.

Inductive reasoning starts with specific cases and is followed by the devel-



opment of general rules, while deductive reasoning starts
with a general theory and derives specifics from the general
theory. Both methods are useful and are powerful reasoning
methods; but inductive reasoning is the natural mode for
learning completely new areas. Thus, when the students are
exposed to separations for the first time, the use of inductive
reasoning will be more effective. In elective and graduate
student courses where the students are seeing some of the
material for a second time, deductive reasoning may be a
more efficient way to teach.
Four concepts of educational philosophy were presented
and discussed at the workshop. The first three of them were
extracted from Hougen's principles:181
1. The undergraduate program should be practical and
conservative, whereas the graduate program should be
imaginative and exploratory.
2. There should be a smooth flow of information from
graduate research to graduate teaching to undergraduate
3. If you can't find relevant problems to give the student,
then you shouldn't be teaching the material to students.

The fourth statement of educational philosophy was added
by the lecturer, Phil Wankat:
4. Different departments should do different things.

In the discussion following the presentation there was
general agreement with these four statements of educational
philosophy. The practical result of the first and third con-
cepts is that only separations currently used in industry were
included in the proposed undergraduate courses. The appli-
cation of the fourth concept led to a lack of consensus on
what to teach.

Five working groups, with six or seven professors in each
group, began the workshop portion by introducing them-
selves and discussing their current course structure in sepa-
rations. Separations courses included courses with titles
such as Separations, Mass Transfer, or Unit Operations,
among others. In addition, the separations parts of labora-
tory and design classes were included in the discussion.
Each group developed an undergraduate curriculum in sepa-
rations subject to the constraint of no increase in the credit
hours. A reporter recorded the group's comments, made a
presentation to the large group, and if desired became a
coauthor of this paper.
One of the functions of the working groups was to share
information about the wide variety of coverage on separa-
tions topics, both in the U.S. and the Canadian schools.
Schools reported from one to four lecture-type courses cov-
ering separations topics. There appeared to be approximately
an equal distribution between stand-alone separations courses
and mass transfer courses with a major separations compo-









Figure 3. Technological and Use Maturities of Separa-
tion Processes.[21 Reproduced by permission of the American Institute of
Chemical Engineers from G.E. Keller, III, AIChE Monograph Series, 83 (17),
1987. Copyright 1987, AIChE.

nent. In addition, all schools apparently include separations
in laboratory and/or design courses.
At a few schools a significant portion of separations mate-
rial was included in the capstone design course. Textbooks
mentioned included: Bennett and Myers; Foust, et al.;
Geankoplis; Henley and Seader; Hines and Maddox; King;
McCabe, et al.; Skelland; Treybal; Wankat; and notes by
Rousseau and by Tiller. This list of texts is similar to the
bibliography in the latest AIChE mass-transfer survey.191
Most schools supplement the textbooks with computer
programs, often in design classes. Computer programs
mentioned included Aspen, Aspen Plus, Flowtran, HYSIM,
and various spreadsheets for doing McCabe-Thiele cal-
culations. The working groups discussions showed a healthy
diversity in both the material presented and the method
of presentation.

None of the groups reached a consensus on what should be
taught (except that all groups agreed that there was too much
material to teach in the available time). This lack of consen-
sus apparently occurred because of the large variety of offer-
ings in separations and because everyone was explicitly
encouraged to retain diversity. Despite the lack of an overall
consensus, however, there were areas with substantial agree-
ment-as well as an interesting split of opinions.
There was a general opinion that newer separations are
under represented in the current curriculum. These processes
include membrane separations (reverse osmosis, ultrafiltra-
tion, gas permeation, and pervaporation), adsorption includ-
ing pressure swing adsorption, and chromatography. Since
hybrid applications coupling two or more types of separa-
tions are becoming increasingly important in industry, stu-
Chemical Engineering Education


Gas Absorption
.Ext. I/Azeo.
Crystallization Dist.
-on Exchange Solvent ExL
SAdsorption: Gas Feed

- Adsorption: Liquid Feed
Supercritical Membranes: Gas Feed
Gas Abs. / Ext. *0 Membranes: Liquid Feed
- Liquid Chromatography:
Membranes Liquid Feed
Field- Induced Separations
Atfinity Separations

l I I I I ] I I

dents should be made aware of these applications-perhaps
in a design course.
In addition to the underrepresented new separations, a
number of important existing mechanical separations (see
Figure 1) involving particulates have almost disappeared
from the curriculum at many, but not all, schools. Since
separation operations with particulates are extremely impor-
tant industrially, some way needs to be found to either retain
or reinstall them into an already crowded curriculum.

One of the chemical engineering laboratory courses was
proposed as a convenient place to cover those separations
which don't conveniently fit into existing courses. Examples
include mechanical separations such as filtration, centrifuga-
tion, or flotation. Many schools already have these experi-
ments in their laboratories, and they should certainly be
retained and updated. Newer separations such as chromatog-
raphy and membrane separations can easily be included in
laboratory courses.

These topics were discussed in considerably more detail in

Course Outlines at Authors' Schools

Class Meetings
Manhattan N. Dakota Purdue
Reqeqd eq'd2 Req'd Reqd El
Topic (3cr) (3 cr) (3 cr) (3 cr) (3

Introduction 1 1 1 1
Costs & Types of Separations 1 2 1
VLE 2 3 1
Flash Distillation 3 3 2
Distillation: McCabe-Thiele 8 12 9
Batch Distillation 1 3 2
Short-Cut Distillation Design 2 3 2
Computer Methods 2 3 1
Multi-Component Distillation 5 3 7
Adsorption/Stripping 4 3 3
Extraction 7 3 5
Leaching 2 -
Washing 2 1
Staged Column Design 4 2
Packed Column Design 3 3 1
Humidification -
Adsorption 11 -
Chromatography 3 -
Electrophoresis 4
Ion Exchange 2 4
Membranes 11 -
Selection and Sequencing 1 1
Diffusion 3
Interfacial Mass Transfer 3 -
Exams (includes final) 5 6 4 7 4
TOTAL 46 46 46 46 4
Textbook Treybal Wankat Wankat Wankat Wan
1980 1988&1990 1988 1988 19

Winter 1994

other sessions at the summer school. One problem with
using laboratory experiments is that at many schools stu-
dents choose laboratory experiments and thus many of them
will graduate without having done particular experiments.
But the chance of exposure to a particular separation method
was thought to be better than no chance of exposure. The
lack of formal instruction on a given laboratory project was
not thought to be a major problem since it forces the students
to learn new material on their own.
Although there were a few strong dissenters, most partici-
pants felt their current curriculum was too heavy in distilla-
tion. At the same time, there was an intriguing 50/50 split
among participants as to whether or not to retain the use of
Ponchon-Savarit diagrams in distillation. Paradoxically, many
professors wanted less coverage of distillation, but wanted
to retain use of Ponchon-Savarit diagrams. It was thought
that some room might be made in the curriculum by using
computer-aided instruction programs for helping the stu-
dents to visualize McCabe-Thiele and Ponchon-Savarit dia-
grams"101 and for doing "what-if?" calculations. This method
might increase the rate at which the stu-
dents learned this material. Of course, there
would be a strong temptation to use the
same amount of time and have the students
learn the material more thoroughly. Since
Tulsa writing educational software is extremely
et Reqd time-consuming, this approach is efficient
cr) (4 cr) if, and only if, the programs can be widely

S shared. Some participants also felt strongly
that their schools had too much coverage
1 of liquid-liquid extraction.
12 As an illustration, the coverage at the
2 authors' four schools is shown in Table 1.
2 The order of topics in the table is not nec-
essarily the order of presentation in the
5 course. One can analyze the four required
4 courses (not including Req'd. 2) in Table 1
S for the lowest common denominator of
course topics that is acceptable. The results
are: Introduction (1), VLE (1), Flash dis-
3 tillation (1), McCabe-Thiele (8), Batch
4 -distillation (1), Short-cut distillation (1),
S Multicomponent distillation plus computer
(5), Absorption/stripping (3), Extraction (3),
Packed design (1), and Examinations (3).
3 Thus, 28 of the 46 class periods (or 61%)
are common to all four courses. The re-
6 maining 39% could be up to the discretion
3 of the instructor. Obviously, this allows
6 57 significant opportunity to teach different
akat Henley/Seader material.
90 1981
Treybal An elective course in bioseparations, par-
1980 ticulate separation, or rate separations was


considered to be a good method for allowing students to
study additional separations without clogging the curricu-
lum. These electives can be dual-level courses and might
allow some faculty to teach in the area of their research.
Unfortunately, when the details of an elective course were
looked at there was considerable disagreement on the depth
versus the breadth of the course. The compromise of cover-
ing a few topics in depth while other topics are only sur-
veyed might be acceptable; but there was a strong feeling
that the course had to be integrated (perhaps using Figures 1
to 3) and not be a series of unconnected topics. The coverage
in an elective and a required advanced course at the authors'
schools is shown in Table 1.

Like many other areas of chemical engineering, knowl-
edge and application of new separation technologies are
expanding at a rapid rate. The problem of how to introduce
new separations into the curriculum is exacerbated by the
neglect of particulate separations. Modest adjustments in
current courses can probably be made by reducing the cover-
age of distillation and by adding new or neglected separa-
tions to the current course and in design and laboratory
courses. One or more elective courses in separations are also
highly desirable.

The authors thank all of the workshop attendees for their
enthusiastic participation during the workshop. In particular,
comments by recorders Alan Foss and Bonnie Tyler were
helpful in writing this paper. We also thank the industrial
supporters and the National Science Foundation for making
the Summer School possible.

* Bennett, C.O., and J. Myers, Momentum, Heat and Mass Trans-
fer, 3rd ed., McGraw-Hill, New York (1988)
Foust, A.S., L.A. Wenzel, C.W. Clump, L. Naus, and L.B.
Anderson, Principles of Unit Operations, 2nd ed., Wiley, New
York (1980)
Geankoplis, C.J., Transport Process and Unit Operations, 2nd
ed., Allyn, Boston, MA (1983)
Henley, E.J., and J.D. Seader, Equilibrium-Stage Separation
Operations in Chemical Engineering, Wiley, New York (1981)
Hines, A.L., and R.N. Maddox, Mass Transfer: Fundamentals
and Applications, Prentice Hall, Englewood Cliffs, NJ (1985)
King, C.J., Separation Processes, 2nd ed., McGraw-Hill, New
York (1980)
McCabe, W.L., J.C. Smith, and P. Harriott, Unit Operations of
Chemical Engineering, 4th ed., McGraw-Hill, New York (1985)
Skelland, A.H., Diffusional Mass Transfer, Krieger, Malabur,
FL (1985)
Treybal, R.E., Mass-Transfer Operations, 3rd ed., McGraw-
Hill, New York (1980)

* Wankat, P.C., Equilibrium-Staged Separations, Prentice Hall,
Englewood Cliffs, NJ (formerly from Elsevier, New York) (1988)
Wankat, P.C., Rate-Controlled Separations, Elsevier, Barking,
England (1990)

1. Baker, R.W., E.L. Cussler, W. Eykamp, W.J. Koros, R.L.
Riley, and H. Strathmann, Membrane Separation Systems:
A Research Needs Assessment, U.S. Dept. of Energy, DOE/
ER/30133-HI, April (1990)
2. Keller III, G.E., "Separations: New Directions for an Old
Field," AIChE Monograph Series, 83, 17 (1987)
3. Larson, M.A., (Workshop Director), Proceedings of Work-
shop on Opportunities and Challenges in Crystallization
Research, Iowa State University, Ames, IA, March (1991)
4. National Research Council, Separation and Purification:
Critical Needs and Opportunities, Washington DC (1987)
5. Wankat, P.C., Rate-Controlled Separations, Elsevier, Bark-
ing, England (1990)
6. Sherwood, T.K., R.L. Pigford, and C.R. Wilke, Mass Trans-
fer, McGraw-Hill, New York (1975)
7. Lightfoot, E.N., M.C.M. Cockrem, S.J. Gibbs, and A.M.
Athalye, "Recovery from Dilute Solutions," in N.N. Li and
H. Strathmann (Eds.), Separation Technology, Engineering
Foundation, New York, 122-154 (1988)
8. Bird, R.B., "Hougen's Principles: Some Guideposts for Chemi-
cal Engineering Departments," Chem. Eng. Ed., 20, 161
9. Eisen, E.O., "Teaching of Undergraduate Mass Transfer,"
AIChE Annual Meeting, paper 147j, New York, November
10. Jolls, K.R., M. Nelson, and D. Lumba, "Teaching Staged-
Process Design Through Interactive Computer Graphics,"
Chem. Eng. Ed., in press. O

Men^ book review

Non-Linear Chemical Kinetics
by Peter Gray and Stephen K. Scott
Oxford Science Publication, Clarendon Press, Oxford; 453 pgs.
$98.00 (1990)

Reviewed by
Massimo Morbidelli
Politecnico di Milano
Arvind Varma
University ofNotre Dame

Following an overview introductory chapter, this book
is divided into two parts which cover a total of fifteen chap-
ters. The first part is the broader one and is titled "The
Techniques," while the second part, "Experiments," consists
of only two chapters.
In the first part, the backbone kinetic scheme is the so-
called Autocatalator, whose complete version consists of the
following reactions: P-->A; A->B; A+2B--3B; B--->C. It is
Chemical Engineering Education

characterized by reaction rates following the law of
mass action and by the third step which is cubic and auto-
catalytic. This feature of the model provides the feedback
mechanism which produces instability and oscillations. In
the different chapters, the autocatalator is analyzed in detail
by considering some of its meaningful modifications and
several reactor configurations. This kinetic scheme has
been chosen because it exhibits all kinds of typical complex
nonlinear behavior, while it is simple enough to allow for
much exact analysis.
In Chapters 2 through 5, the closed homogeneous reactor
is considered. In Chapters 2 and 3, the isothermal case is
analyzed, first by determining the pseudo-stationary states
of the system. These are obtained by assuming that the
reactant concentration, e.g., the concentration of P, is con-
stant. Then, the concept of steady-state stability is intro-
duced and a linear stability analysis is performed to deter-
mine stability conditions on the parameters of the model. It
is shown that under the assumption of no reactant depletion,
the autocatalator exhibits oscillatory behavior (e.g., limit
cycles) for the values of parameters where the steady-state is
unstable. The importance of casting the model equations in
dimensionless form is strongly emphasized and this tech-
nique is applied throughout the book. Finally, the complete
model, accounting for reactant consumption, is studied, thus
proving that its behavior can be understood and predicted by
exploiting the previous analysis based on the pseudo-
stationary states and their stability.
Chapters 4 and 5 deal with thermokinetic oscillations in a
homogeneous reactor which exchanges heat with the sur-
roundings but is closed to mass transfer. In this case the
kinetic scheme simply consists of two consecutive first-
order reactions: P->A and A->B. Here the feedback mecha-
nism is provided by the thermal effect due to the exothermic-
ity of only the second reaction. Performing the same analysis
(e.g., first determining the pseudo-stationary states and their
stability and then describing the reactant consumption) yields
similar results as in the isothermal case. In order to account
for the temperature dependence of the model, the following
approach is used throughout the entire book. First the Arrhe-
nius law is approximated by a simple exponential expres-
sion, allowing one to obtain several analytical results. Then
the exact Arrhenius law is used and the results are compared.
In Chapter 5, the Hopf bifurcation analysis and techniques
for the quantitative analysis of relaxation oscillations are
described and applied to the thermokinetic oscillation model.
Chapters 6 through 8 deal with CSTRs. In Chapter 6, the
isothermal case is considered and steady-state multiplicity
patterns for kinetic schemes of increasing complexity are
determined. The effects of residence time and of the
reversibility of reactions are analyzed. The non-isothermal
CSTR is studied in Chapter 7 with first-order kinetics and
Arrhenius temperature dependence. Both adiabatic and non-

Use CEE's reasonable rates to advertise.
Minimum rate, 1/8 page, $100;
Each additional column inch or portion thereof, $40.

demonstrating excellence in both research and teaching are invited to
apply for a position at the Assistant, Associate, or full Professor level.
The area of research interest is open. Applicants should send a curricu-
lum vitae, list of three references, transcripts, and statement of research
and teaching objectives to: Dr. Isaac C. Sanchez, Chair of Search Com-
mittee, Department of Chemical Engineering, The University of Texas
at Austin, Austin, Texas 78712-1062. AA/EOE

The Department of Chemical Engineering at the University of Minne-
sota, Duluth, invites applicants with a PhD in chemical engineering or
closely related field to apply for a tenure track faculty position at the
assistant professor level commencing September 1, 1994. The success-
ful candidate is expected to develop a research program and attract
external funding, promote outreach, teach undergraduate core curricu-
lum, and advise undergraduates in an ABET-accredited program. Areas
of teaching emphasis include unit operations laboratories, thermody-
namics, hazardous waste processing, statistical anal analysis and design of
experiments. Industrial or research experience and teaching interests in
the area of hazardous waste management are desirable. Demonstrated
evidence of effective teaching and communication skills appropriate to a
faculty member is required.
Applications including a resume and three references must be received
by February 15, 1994. Send completed application to: Dr. Dianne Dorland,
Head, Department of Chemical Engineering, University of Minnesota,
Duluth, 231 Engineering Building, 10 University Drive, Duluth, Minne-
sota 55812. "The University of Minnesota is an equal opportunity edu-
cator and employer."

adiabatic reactors are considered. Here, the strong analogies
with the cubic autocatalytic kinetics in the isothermal
CSTR are illustrated. Also in the chapter, singularity theory
and its applications are introduced. In order to demonstrate
its power, some results about isothermal cubic antocatalysis
and non-isothermal CSTR are obtained again by applying
singularity theory methods.
Chapter 8 deals with the stability of stationary states in
the isothermal autocatalytic CSTR. In particular, the re-
sponse to transient perturbations is analyzed, illustrating
the exponential relaxation to stable states and the exponen-
tial growth from unstable states. The onset of oscillations
when the stationary state loses its stability is studied, with
particular reference to conditions for emerging stable or
unstable limit cycles.
Chapters 9 through 11 are concerned with spatially distrib-
uted systems where molecular diffusion and thermal con-
duction processes play a fundamental role. First, reaction-
diffusion cells are introduced, and spatially distributed sta-
Continued on page 28.

Winter 1994

e f classroom



Rensselaer Polytechnic Institute
Troy, NY 12180

Over the past thirty-five years, a substantial data base
has become available for the principal "energetic"
excess properties (gE, hE, and cE) of binary liquid
mixtures. A major use of these numbers is for incorporation
into group-contribution techniques (e.g., ASOG, UNIFAC)
for estimating liquid-phase activity coefficients. The appli-
cation here is quantitative; the ultimate goal is to predict the
composition and temperature dependence of the activity co-
efficient, 'y, in binary and multicomponent mixtures.
The excess-property data can also serve another, more
qualitative, purpose. Because they reflect differences be-
tween energetic and structural effects in a solution relative to
those in the unmixed components, the excess properties
serve as probes for elucidating phenomena at the molecular
level. The signs and relative magnitudes of the excess prop-
erties can therefore, with judicious interpretation, be used to
support or disqualify molecular theories. The desired gener-
alizations, however, must be based on a large number of
experimentally based results.
This "explanatory" role of the excess properties demands
that the most rational and communicative way of organizing
the data base be found. What simple ways exist to display gE,
hE, and cE which could highlight patterns and trends? How
best can we use these plots as aids for modeling and as props
for qualitative discussions of phase equilibria? Are there
patterns and trends that suggest important generalizations
connected to the chemical natures of the species involved?
Early work suggests that there is at least a qualified "yes"

* University of Virginia, Charlottesville, VA 22903

answer to these questions. Malesinskil" proposed classifica-
tions based on the signs of gE, hE, and sE; it is this simple idea
that all subsequent schemes share. Kauer, Bittrich, and Krug"21
used a plot of gE vs. TsE to display the then small g/hE
data base; they also proposed classifications and generaliza-
tions based on signs and mixture type. Gaube and cowork-
ers[3-51 employed a plot of gE vs. hE and also a modified
diagram in which gE is replaced by gE + TsE, where sC is the
combinatorial excess entropy. In these later efforts, the gE/hE
data base comprised about 200 points.
Most recently, Shukla, Chialvo, and Haile'61 studied the hE
vs. gE and hE/RT vs. gE/RT diagrams. While no data were
displayed, the authors discussed the classical-thermodynamic
features of these plots for miscible and immiscible systems
and also noted how molecular theory with different size and
energy ratios for the intermolecular potentials can lead to
various diagrams.
Since 1983, when the paper of Kohler and Gaube first
became known to us,171 we have explored the use of excess-
property diagrams for organizing data and for discussing
observed property and phase behavior. The diagrams are
particularly helpful as visual aids in the classroom. In fact,
early experiences with them were so positive that they sub-
sequently became vehicles for two comprehensive class-
room projects in which students scoured the literature for
excess-property data and then participated in the posing of
explanations and generalizations based on the results of their
searches. Thus, most of the coauthors of this paper are former
Rensselaer students who participated in these exercises, and
many of the data reported herein were gleaned by them.

Michael M. Abbott and John P. O'Connell are Professors of chemical
engineering at, respectively, Rensselaer Polytechnic Institute and The
University of Virginia. They share interests in thermodynamics and in
chemical-engineering education. Their twenty co-authors are former
BS, MS, and/or PhD students from Rensselaer who participated in the
collection, evaluation, and organization of the data upon which this
paper is based.

Copyright ChE Division ofASEE 1994

Chemical Engineering Education

At modest pressures, the excess properties of liquid mix-
tures depend only on composition and temperature. Isother-
mal data allow separation of these effects; many data are
reported at 298 K. At this temperature, comparison among
systems may be most representative at equimolar composi-
tion (the choice we make), but the way to display data still
must be chosen.
The excess Gibbs energy, excess enthalpy, and excess
entropy are related by

gE = hE T (1)
The experimentally accessible quantities are gE (via activ-
ity coefficients from vapor-liquid equilibria) and hE (from
calorimetry and temperature variations of gE). When these
are measured or estimated at the same conditions, a value
for sE can be found at the same conditions (there is no
entropy meter!). Equation (1) shows that only two of these
properties are independent, so any pair from the set {gE, hE,
sE} can be used as the coordinates. The three types of dia-
grams (gE vs. hE, gE vs. SE, and hE vs. SE) convey exactly the
same information.
Equivalent to Eq. (1), we may write

gE h sE (2)
which suggests choices from the set {gE/RT, hE/RT, sE/R) as
alternatives with exactly the same information.
We favor dimensionless ("scaled") coordinates (i.e., pairs
from the second of the above sets) for the following reasons:
The quantity gE/RT is the most natural dependent variable for
phase-equilibrium applications, because in y is a partial
molar property with respect to gE/RT.
The quantity hE/RT is cleanly related to gE/RT via the Gibbs-
Helmholtz equation

E (a(gE / RT)
S-T ,x (3)
\ /P,x
One can establish temperature-independent upper bounds for
gE/RT of a stable liquid mixture.
By scaling, values for gE and/or hE at different
temperatures are put on a more comparable
basis. Definitio
Experience shows that important generaliza-
tions and rules-of-thumb are more easily Region S
grasped and retained when expressed in
dimensionless terms. I
Again, which scaled coordinates should be
used? For everyday work, we favor the first
(gE/RT vs. hE/RT) and the last (hE/RT vs. sE/R),
which we call "engineering" and "modeling"
coordinates, respectively. Engineering coordi-
nates represent experimental quantities; they
Winter 1994

facilitate discussion of phase equilibria, especially liquid/
liquid equilibria (see Shukla, et al."61). Modeling coordinates
represent directly the enthalpic and entropic contributions to
gE/RT; they facilitate explanations of system-to-system trends
in gE/RT. Our goal in this paper is to present the data base we
have accumulated and demonstrate the engineering patterns
it shows. Thus, we will focus on the gE/RT vs. hE/RT diagram.
For convenience, we introduce the following notation for
equimolar binaries:

SgE / RT]
= hE / RT
S=sE /R
c E/RJ

At X=X2 =0.5

Equations (2) and (3) then become

q h- (5)
aT T
We also have, from classical thermodynamics,
a _-- & (6)
aT T
and thus, by Eqs. (4) through (6), we have
ah c-h (7)
Equations (4) through (7) are useful for analyzing the fea-
tures of the g vs. h diagram.
According to Eq. (4), there are just six possible combina-
tions of sign for g, h, and s. These are listed in Table 1.
Each sign combination defines a region on the a vs. h
diagram (see Table 1 and Figure 1) We number the regions
from I (counterclockwise) to VI.

The diagonal
line on Figure 1
corresponds to
= 0. In regions
to the right of the
diagonal (V, VI,

n of Regions on the
vs. h Diagram

iign g Sign h Sign i

+ + +
+ +

+ +




Figure 1. The g vs. h diagram.

and I), i is positive, and for regions to the left of the
diagonal (II, III, and IV), s is negative. Lines of constant
(nonzero) s are parallel to the s = 0 diagonal.
In Regions II and V the signs of g are preordained by
the signs of h and s according to Eq. (4). This is not so
for the rest of the diagram. Consider Region I where g,
h, and i are all positive. According to Eq. (4), the sign
of g is ambiguous, but the definitions actually require
that in Region I enthalpy dominates since h must
be greater than s. Similar arguments show that
enthalpy also dominates in Region IV, while entropy
dominates in Regions III and VI. (This pretty line of
reasoning, used by Malesinski,Em1 is purely classical and
So far, we have said nothing about magnitudes. How
large is large? We begin with g, defining a "large" g
as one for which phase-splitting (here, LLE) is likely.
In the simplest approximation (with gE/RT = Axx2),
g = 1/2 yields LLE; this model-dependent result con-
stitutes a practical lower bound on a for phase-splitting.
A greatest upper bound for stability obtains for
g = en2 = 0.6931 ..., corresponding to the Gibbs energy
change of mixing Ag being zero for an equimolar
binary mixture. Values in the range 0.50 < k < 0.69 are
thus "large," with g = /n2 chosen as the concrete limit
on "largeness."
What of h and i? Suppose that h = 0, as for an
"athermal" solution. According to Eq. (4), if ath = tn2,
then ia, = -tn2. Thus i < -en2 is a "large" negative s. In
fact, if i < -in2, h must be negative to produce a stable
liquid mixture. (Notice that the entropy change of mixing
As is negative for i < -in2. This perhaps counterintuitive,
but occasionally observed, behavior suggests "unusual"
phenomena in action.)
Suppose that i = 0, as for a "regular" solution. Accord-
ing to Eq. (4), if geg = in2, then ,,eg = tn2. Hence
h = fn2 is a "large" positive h; if h > tn2, i must be
positive to produce a stable liquid mixture.
We thus establish g > tn2, i < -fn2, and h > tn2 as
criteria of "largeness"; this is about as far as purely
classical reasoning can take us. The missing bounds (prac-
tical lower bounds on g and h, and a practical upper
bound on i), if they exist, must be supplied by Nature.
The temperature dependence of the excess properties
for a given mixture defines a trajectory on the g vs. h
diagram. Figure 2 shows a few examples. Some kinds of
trajectories are forbidden. For example, Eq. (5) requires
that g increase with T in Regions III, IV, and V (where
h is negative), and that g decrease with T in Regions VI,
I, and II (where h is positive).
From Eq. (6), the sign of ai/aT is determined solely by
the sign of 8. On the other hand, Eq. (7) shows that both

Figure 2. Trends with temperature for equimolar mixtures.
Symbols at 298 K; ranges are at least 100 K.

The gE/E Data Base
Classification by Region and Mixture Type

-- Region
Mix Type I II HI IV V VI Total
NP/NP 46 4 2 1 10 32 95
NA/NP 80 16 7 5 2 0 110
A/NP 29 54 0 0 0 0 83
NA/NA 11 6 6 24 0 1 48
A/NA 28 15 11 5 0 0 59
A/A 8 6 4 6 1 1 26
Totals 202 101 30 41 13 34 421

c and h contribute to ah/lT. (We will show later that negative h
usually implies positive E, so both i and h normally increase with
T for systems in Regions III, IV, and V.)
In any case, i is important in the analysis and prediction of
trends with T on the g vs. h diagram. Thus, we seek effective
ways of graphically displaying data for c. (We will show later
that a plot of i vs. i has advantages as an organizational and
explanatory aid.)

For organizing and discussing the data base, we (like many
others) find it convenient to classify mixtures by "type." We use a
coarse classification based on separate identification of the com-
ponents as nonpolar (NP), polar but nonassociating (NA), or polar
and associating (A). Here, "association" means association only
by hydrogen-bonding-though other association mechanisms ex-
ist. Hence, by our convention, acetonitrile/n-hexane is an NA/NP
mixture, whereas ethanol/n-hexane is an A/NP mixture. Notice
that this scheme gives us the same number of binary mixture types
(six) as there are regions on the g vs. h diagram. (This handy six-
Chemical Engineering Education

The "Extended" gE/hE/cp Data Base
Classification by Signs and Mixture Type


hEE EE hE@ SE@

hE sEQ hE SE

by-six mnemonic has no special thermodynamic significance.)
Data were collected in two separate sweeps of the litera-
ture. In the first effort we sought systems for which both gE
and hE had been measured (or could be estimated) at or near
to 298 K. In the second search, we also looked for ce data.
Both primary and secondary sources were consulted.
Table 2 summarizes the makeup of our gE/hE data; the data
themselves are in an Appendix that is available from the
senior authors (Abbott and O'Connell). In addition to the
approximately four hundred organic and aqueous/organic
mixtures classified in Table 2, equimolar gE/hE data were
found for twenty-two cryogenic mixtures at temperatures
ranging from 0.9 K (helium-3/helium-4) to 184 K (ethylene/
nitrous oxide and nitrous oxide/xenon). These data are also
available and can be obtained by writing the authors.
Although the cp data base is reasonably large (ca. 350
different mixtures), the overlap of systems with the gE/hE data
base is relatively modest. In many cases, however, the sign
of eE (if not the magnitude) can be unequivocally fixed by
Winter 1994

The "Extended" gE/hE/c Data Bas
Classification by Region and Mixture Tyj
Region and Sign of CpE
Mixture E E E E E E E E E
Type cpE cp cp cp cp @ cp cp cp cp(
NP/NP 3 25 0 0 0 0 0 0 1
NA/NP 10 13 3 0 1 2 0 1 0
A/NP 1 0 28 0 0 0 0 0 0
NA/NA 5 4 4 1 2 0 22 0 0
A/NA 6 1 7 0 3 0 0 1 0
A/A 1 0 1 0 3 0 0 1 0

Totals 26 43 43 1 9 2 22 3 1

Mixture Type

% of gE/hE

% of Extended g /hC /

Hence, about 65% of our mixtures contain a nonpolar

interpolation, by analogy, or by in-
spection of the temperature varia-
se tion of the data for hE. For example,
pe available data show that cE is al-
ways positive for 1-alkanol/n-al-
kane systems at 298 K. Similarly,
v VI cE is always positive for solvating
)c cpE Cp cpe TOTAL NA/NA mixtures. (Here, "solva-
1 1 15 46 tion" means that strong unlike at-
0 0 0 30 tractions occur even though asso-
0 0 0 29 ciation may not be found for one or
0 0 0 38 both of the unmixed components.)
0 0 0 18 Thus, we can define an "extended"
0 0 0 6 gEhE/c data base comprising about
_ ____ 150 systems for which gE, hE, and
1 1 15 167 either cp or its sign are known; it is
summarized in Tables 3 and 4.
Before presenting our findings,
we offer a few words of caution:
0 First: We make no claims of completeness. This collection is
largely the result of two classroom (i.e., time- and resource-
limited) searches of the mostly post-1960 literature.
0 Second: Our ground rules were to include only miscible
mixtures of nonelectrolytes at temperatures near to 298 K.
This delimits and biases the data base. Partially-miscible
mixtures are, of course, of great concern to designers of
separation processes, yet our collection excludes such systems.
0 Third: Although the relative proportions of data for the
various mixture types should reflect the relative numbers of
available gE/hE sets for these types, one must recognize that the
kinds of systems reported actually reflect the individual and
collective biases of thermodynamic experimentalists and their
customers. As a result, some classes of mixtures have received
disproportionately intense attention because of their interest to
correlators and theoreticians, and not because they are
particularly "representative" of Nature.
J Fourth: We note that many of the gE and/or hE values for 298
K are determined by extrapolation or derivation via the Gibbs-
Helmholtz equation. Such estimates are of course better than
no estimates at all, but they are, in the end, only estimates.
With these caveats in mind, we can briefly review the
statistical makeup of our data collection. For the gEhE and
extended gE/hE/cE data bases, mixture types are represented
approximately as follows (see Tables 2 and 3 for details):

species, 50% a nonassociating polar species, and 35% an
associating polar species. Of the binary mixture types, A/A
mixtures are the most poorly represented, accounting for
only about 5% of the whole.

Figure 3 is a g vs. h plot for the complete gE/hE data base.
Although this picture appears at first glance to be a scene
from an experimentalist's nightmare, it (and its companion
Table 2) delivers some important messages:
1. Regions I and II are very heavily represented, accounting
for 48% and 24% of the data base: positive gE and hE are
"the norm."
2. Region V is very sparsely represented. Nature appears to
abhor the most inherently stable of liquid mixtures-the
systems with negative hE and positive sE.
3. Only 59% of our mixtures have positive sE, whereas 80%
have positive hE. Thus, in a very gross statistical sense,
the regular solution (sE = 0) is a better approximant to
reality than is the athermal solution (hE = 0).
4. Negative hE implies negative sE, and positive sE implies
positive hE about 90% of the time. The converses are true
only about half of the time:

hEO sE)
sE@ -> hEe

(85% valid)
(95% valid)

hE -> sE (70% valid)
sE hEO (41% valid)
5. Nature seems to provide some of our missing bounds for
"largeness" of g and 9. Very approximately, according to
Figure 3, we may consider g < -0.4 and s > 0.4 as
additional criteria of "largeness." When combined with
the criteria presented earlier, these bounds define the
rectilinear region denoted by the dashed line in Figure 3.
Systems falling outside this region may be considered
Figure 4, the modeling plot of h vs. s for the complete gE/
hE data base, without most of the "unusual" cases, conveys
the same messages as Figure 3.

Plots of gvs. h
It is instructive to examine a vs. h relations for each of the
six binary mixture types. We do them in order of increasing
molecular complexity

NP/NP Mixtures
These are shown by the open circles of Figure 3 (and
separately in Figure 6 of the Appendix, available from the
authors). The systems fall mainly in Regions I and VI; g is
small to modest in size, rarely exceeding 0.2 in absolute
value. The Region VI mixtures mostly contain alkanes or
other "inerts" of greatly different molar volume; here, the

Boundary of "Large" g
Excess Properties 9


-1.5 -1.2 -0.9 -o.g. -e.3 o,a 0. 0.9

o % <0 x -o0
*"*' ^ -- --_ _- .--- -- _- -


Figure 3. The complete gE/hE data base for binaries at
298 Kand equimolar composition.

Boundary of Large"
Excess Properties



-1.0 -q.7 -0.4

0 -0.

0.5 0.8S

Figure 4. The complete hE/sE data base for binaries at
298 K and equimolar composition

negative g results from a relatively large positive s ("en-
tropy dominates"). The Region I mixtures mostly contain
two alkanes or two aromatics of modest molar-volume ratio,
or mixtures of an alkane with an aromatic hydrocarbon. Heat
effects can be quite large for the alkane/aromatic systems,
but these tend to be compensated by a large positive value of
s, leading to small or modest g.

NA/NP Mixtures
These are shown by the triangles in Figure 3 (and sepa-
rately in Figure 7 of the Appendix. Region I behavior is the
norm ("enthalpy dominates"), but one finds occasional ex-
cursions into Regions II through V. Most of the latter cases
are systems in which one of the substances is an aromatic
hydrocarbon, a tertiary amine, CC4, or acetonitrile. Mix-
tures containing acetonitrile tend to fall in Region II; their
location makes them appear to be "weak" relatives of A/NP
mixtures (see below). Notice that for NA/NP mixtures, both
g and h can be quite large.

Chemical Engineering Education


- / f

* 2.0

0 0

C = -S

* 0

iv -



-1.0 -0.7 -0.4 -0.

V *0





f 0.5 0.8
S-,, viii
0* *


C = -S

Figure 5. The complete cE/sE data base for binaries at
298 K and equimolar composition.

A/NP Mixtures
Data for these systems are the crosses in Figures 3 (and are
plotted separately in Figure 8 of the Appendix). All data fall
in Regions I or II. The stronger associators (alcohols and
carboxylic acids) tend to show Region II behavior when
mixed with alkanes. Here, s is negative and g is large and
positive: so large as to lead to phase splitting in extreme
cases. Mixtures of strong associators with aromatics exhibit
smaller values of g, owing to smaller negative, or even large
positive, values of s Mixtures of secondary amines with
hydrocarbons behave similarly to NA/NP systems: h and s
are positive and sufficiently comparable in magnitude so as
to produce small to modest values for g.

NA/NA Mixtures
As the diamond symbols of Figure 3 (and Figure 9 of the
Appendix) show, these systems exhibit one of two general
kinds of behavior, depending on whether the unlike species
can solvate by hydrogen bonding. If solvation occurs (ac-
etone/chloroform is the classical example), then gE, hE, and sE
are all negative, and Region IV behavior obtains (open
diamonds). If both species are proton donors or both are
proton acceptors, then Region I or Region II behavior is
common (solid diamonds). Quasi-ideal mixtures (very small
k, h, and s) are possible when the two species both have
high effective polarity.

A/NA and A/A Mixtures
Data for the A/NA systems are the box symbols in Figure

Winter 1994

3 (plotted separately in Figure 10 of the Appendix). A diver-
sity of behavior is seen, but a is usually positive and is often
large. Generalization is difficult because of the complex
molecular effects in operation; association or electrostatic
interactions between like molecules may be partially com-
pensated by solvation between unlike species. Region III
behavior is not uncommon, especially at low temperatures.
Aqueous systems are the open boxes in Figure 3, principally
in Region III.
The x symbols in Figure 3 (plotted separately in Figure
11 of the Appendix) show the very small data base for A/A
mixtures. Association or solvation can occur between all
pairs, sometimes leading to a near cancellation of polarity
or association effects (as in many alcohol/alcohol mix-
tures), and sometimes not. No easy generalization can be

Relationships with cE
Finally, we consider the excess heat capacity. The statis-
tics in Table 4 suggest a strong correlation between the signs
of sE and cE. Thus a negative s' (or cj) gives a positive cE
(or sE) about 90% of the time, while a positive sE or (c4)
implies a negative cp (or sE) about 70% of the time:

sEe E- cE (93% valid)

cp e sEG) (91% valid)

sE@ --- ce (68% valid)

ce P SEQ (73% valid)
Hence, we have the approximate equivalence

Sign (c) = Sign (sE)

Searching for correlations between the signs of hE and c ,
we find

hE -- (c (84% valid)

c4e -hE (91% valid)

hE -> ce (46% valid)

c@-> hEe (31% valid)
Negative signs on hE (or c ) imply positive signs on cE (or
hE) about 85% or more of the time, whereas positive signs on
hE (or c ) imply negative c (or hE) only about 40% of the
time. Thus, we have the relatively strong implications that a
negative hE (or c4) gives a positive c4 (or hE). The converse
statements are not generally true, however.
These arguments lead us to consider a plot of 6 vs. s
(Figure 5). Quantitative judgments are aided if one adds to
Figure 5 the dashed parity lines c = s and c = -s. These
lines, with the axes c = 0 and s = 0, divide the 6/1 plane into
Continued on page 77.





Cornell University
Ithaca, NY 14853

any chemical engineering students are not ex-

posed to biology through the standard curricu-
lum. But the use of concepts related to biological
reactors may be required of students since they apply not
only to the bioprocess industries but also to problems in
waste treatment and remediation of sites that are environ-
mental hazards. One approach for introducing these con-
cepts is to incorporate new examples within the existing
curriculum, and the senior laboratory course offers an excel-
lent opportunity to do just that. A biologically based experi-
ment also provides a forum for introducing material that
reinforces traditional chemical engineering principles.
The goals of our senior laboratory are given in Table 1
(based on a handout prepared by George Scheele). We will
show that these goals can be met through experiments
that test the oxygen transfer capabilities of two New
Brunswick Scientific Bioflow III fermenters during one class
period; we then examine the kinetics of yeast growth and
the effects of ethanol inhibition for a batch reactor during a
second class period.

In Cornell's senior laboratory course there are lectures,
briefings before a new laboratory, and a report session.
Experiments span two separate three-hour periods. Stu-
dent groups consist of three students each, and a group
comes to the laboratory the same days of the week for two
consecutive weeks.
Because of the relatively slow response time of biological
systems, designing experiments that will fit within two sepa-
rate three-hour blocks is difficult. Bioprocess laboratory ex-
periments at most universities require an extended one-day

Michael L. Shuler is the Samuel B. Eckert Professor of Chemical Engi-
neering At Comell University. He received both his BS (University of
Notre Dame) and PhD (University of Minnesota) in chemical engineering.
Naheed Mufti is a graduate student in chemical engineering at Comell
University. She did her BSc in biochemistry and BASc in chemical engi-
neering at the University of Ottawa. Her thesis research involves the
design of cell culture systems to study the effects of dioxin in cells of
human origin.
Michael Donaldson received his BS in chemical engineering from Wash-
ington State University and currently is a graduate student at Comell
working on insect cell cultures.
Ronald Taticek is a PhD student in chemical engineering at Comell
University. He received his BASc from the University of Ottawa and his
MASc from the University of Waterloo. He worked as a pilot plant engi-
neer with the Biotechnology Research Institute in Montreal prior to com-
mencing his PhD studies.

experience or make use of continuous culture devices requir-
ing sampling over a period of many days. Our approach
requires considerable prelaboratory preparation by the in-
structor and teaching assistant to circumvent this difficulty.
By using two fermenters in batch mode, with cells already in
exponential growth, and concentrating on transient responses
to specific stimuli, students can acquire sufficient data to
explore several important concepts.
Although we require that students practice good aseptic
technique when taking samples from the reactor, we wanted
a system that was sufficiently robust to minimize the
consequences of any mistakes. Consequently, we chose a
common yeast, Saccharomyces cerevisiae for the experi-
ments. It grows well at low pH (4.0), and if inoculated at
high cell density, it will dominate when another organism
accidentally enters the reactor.
The goal in the first laboratory period is to determine
volumetric mass transfer coefficients (kLa) for oxygen as a
function of air-sparge rates and agitator speed. Two methods
are used: unsteady-state with no cells and dynamic method
with cells. During this first period, students also determine
aerobic growth kinetic parameters and develop calibration
curves for measuring glucose and ethanol.
In the second lab period, the effects of ethanol inhibition

C Copyright ChE Division ofASEE 1994

Chemical Engineering Education

on aerobic (oxygen present) growth kinetic parameters are
tested. Four levels of ethanol are added to the fermenters: 15
g/L, 30 g/L, 45 g/L, and 60 g/L. Aerobic conditions are used
since confounding effects from endogenously generated etha-
nol can be avoided. If this system were being used for
ethanol production, it would be conducted under anaerobic
conditions (no oxygen present). The three-hour period makes
the use of anaerobic conditions impractical.
To rationalize the use of aerobic conditions, the students
are given a scenario in which a two-stage process is envi-
sioned. The first stage of the proposed process is to build up
cell mass as quickly as possible using aerobic growth condi-
tions, followed by a second-stage anaerobic fermentation to
ethanol. One common method of scale-up is to hold ka

Objectives for a Senior Chemical Engineering Course

A. In operating equipment, how to
Judge response time: i.e., bigger is slower to respond
Understand special properties or conditions: i.e., flooding of a
packed column
Develop an intuitive feeling for the behavior of chemical
engineering equipment and the magnitude of process streams
Develop a feel for data that are reliable and data that aren't
B. In interpreting the behavior of equipment, how to
Apply the theoretical and empirical equations discussed in the
lecture courses to real systems
a Develop a sense of appropriateness
n Recognize the limitations of theories
a Unify material from previous courses
Treat data
a Recognize the difference between independent and dependent
variables; between theoretical and empirical equations
n Develop the ability to arrange equations to analyze data
n Develop the ability to interpret results
C. In telling others what you have done, how to
Use generally accepted conventions of written reports
0 Know the format for a good technical report, magazine
article, or journal paper
n Know the conventions of writing, such as consistent tense and
person, proper paragraphing, pronoun-antecedent agreement,
and correct spelling
n Search for vague, overblown, or unnecessary words and
phrases, and to rectify problems
Use graphs and tables
n To analyze data-to fulfill your needs
a To convey ideas-to fulfill others' needs
Learn to synthesize the sometimes conflicting data and ideas at
your disposal into a clear, logical report

Lecture SubTopics

The cell as a chemical reactor Genetic engineering
What are proteins and Batch growth kinetics
enzymes? Kinetics of inhibition
Simple enzyme kinetics Experimental methods to
Major intracellular components determine cell mass and number
The genetic code and the Scale-up problems in bioreactors
Central Dogma How to determine kLa

Winter 1994

constant. This aspect justifies the first laboratory period; the
second period is justified by requiring the students to sug-
gest a kinetic expression that would account for the inhibi-
tory effects of ethanol. The presumption is that such a ki-
netic expression could be applied to both aerobic and anaero-
bic growth situations.
A general introduction to biochemical engineering is given
in two lecture periods focusing on oxygen mass transfer in
fermenters and summarizing kinetic models for growth and
inhibition. Students are given copies of a general review
article"m on biochemical engineering written for physical
scientists, an article on inhibition kinetics,'2] and a part of a
textbook describing oxygen transfer in fermenters.'3 The
topics covered in these two lectures are given in Table 2.


Bioreactor System
A diagram for the primary apparatus is shown in Figure 1.
We use two New Brunswick Scientific Bioflow III ferment-
ers with 3.3 L culture vessels, Phoenix Polarographic (Hous-
ton, Texas) dissolved oxygen electrodes, and Ingold
(Wilmington, Massachusetts) pH probes. An IBM PS/2 Model
30 286 computer controls the two fermenters, and an Epson
FX-850 (Torrance, California) printer prints out data at the
end of the experiment. To simultaneously control the two
fermenters and for data logging, we use the New Brunswick
Scientific Advanced Fermentation Software Package. Stan-
dard set points are pH at 4.0, temperature at 300C, and
agitation at 250 or 350 rpm. Antifoam C (Sigma Chemical
Company, St. Louis, Missouri) was added automatically.
The working volume for each fermentor is 2.3 L.
We added a check valve in the gas-supply lines to allow

Diagram of the Bioflow III with computer-data capture. Note the
three-way valve to change the gas supply. The filtered gas supply
can be directed to the head space or the ring sparger, as desired.
Temperature is controlled by the water bath on the base, and acid
or base is added automatically to control pH. The computer moni-
tors agitation rate, dissolved oxygen, pH, and temperature.

for a rapid switch from air to nitrogen gas. Also, a "Y" after
the flow meter and the use of clamps allowed the gas stream
to be directed to either the sparger ring submerged in the
liquid medium or to the headspace. For measuring kLa (with
no cells present), nitrogen gas is sparged through the me-
dium to remove oxygen; after oxygen is depleted the check
valve can be used to switch to air only. Measurement of the
dissolved oxygen upon re-aeration allows kLa to be calcu-
lated.131 A plot of ln(C* CL) versus time (t) yields a slope
equal to -kLa. C* is the saturated level of dissolved oxygen,
and CL is the value of dissolved oxygen in the medium.
Cells are present for the dynamic method. In this method,
air flow is stopped for a short period of time (the length
of time depends on cell concentration and initial level
of dissolved oxygen) and is then restarted before the dis-
solved oxygen drops low enough to alter cellular metabo-
lism (e.g., below 10% saturation). The rate of oxygen uptake
(or specific respiration rate) can be calculated from an
oxygen balance during the period when no air is being sup-
plied, and the kLa can be calculated from the re-aeration part
of the response by including a correction for oxygen con-
sumption by the cells.[31
A plot of CL versus qo2 X + dCL/dt has a slope of -1/kLa
and is plotted from the data from the re-aeration part of the
experiment. Here q02 is the specific rate of oxygen uptake,
while X is the total biomass concentration. The product,
q02 X, is determined from the rate of oxygen depletion when
the air is off. Since oxygen transfer by surface aeration
would invalidate these calculations, it is important to dis-
place air in the headspace with nitrogen as quickly as pos-
sible. The check valve and "Y" in the gas line make it easy to
switch from air sparging through the medium to nitrogen
flow to the headspace to displace headspace air and to pro-
vide a nitrogen "blanket." Clearly, nitrogen sparging in the
medium could not be used since oxygen removal from the
medium would then be due to both cellular respiration and
gas stripping.
Other pieces of equipment used in this laboratory include
a spectophotometer (Milton Ray, Spectronic 301; UV-vis-
ible wavelengths) and a bench-top shaker (New Brunswick
Scientific Company, Edison, New Jersey; G24 Environmen-
tal Incubator Shaker).
Organism and Medium
The organism used was Saccharomyes cerevisiae Cuy8,
obtained from Dr. Tim Huffaker's lab at Cornell. Most strains
of S. cerevisiae (a yeast) would be acceptable. The composi-
tion of the growth medium was
10 g/L glucose
1.5 g/L yeast extract
4.8 g/L (NH4)2 SO4
0.75 g/L KH2PO4
0.24 g/L MgSO4-7H,O

0.036 g/L CaCl2-2H20 pH to 4.0
Medium was autoclaved at 121C for 50 minutes in four
2L batches.

Start-Up Procedures
For the experiment to be completed in the allocated time,
the fermentation must already be under way when the class
begins. The inoculum is prepared by inoculating 100 mL of
sterile medium in a 250 mL Erlenmeyer flask with silicon
closure (for good gas transfer) using a loop of yeast from a
colony on a Petri plate. Three flasks are used; they are
incubated on the Model G24 shaker for 16 hours at 300C and
350 rpm. The bioreactors were inoculated 4 hours prior to
student arrival with 300 mL of inocula and 2 L fresh me-
dium. This procedure circumvents the lag phase and ensures
that the culture is in the early exponential growth phase. This
procedure is critical if students are to complete the experi-
ment within three hours.
Although the reactor itself should, in principle, be steril-
ized by autoclaving each day, we autoclaved the vessel for
Monday's laboratory but did not autoclave for subsequent
laboratory sessions. Rather, we emptied the vessel after the
laboratory and cleaned and disinfected it with 70% (by vol-
ume) ethanol acidified to pH 2 with HC1. The low pH of the
medium provides protection against significant contamina-
tion, and using a disinfection solution reduces the time and
labor involved in laboratory preparation.
Another aspect of laboratory preparation is calibration of
the dissolved-oxygen and pH probes using manufacturer
protocols. The dissolved-oxygen probes are particularly sen-
sitive to electrical interference and fouling by proteins and
medium components.

Assays: We used two enzyme-based assay kits (glucose
and ethanol) and directions for both were supplied by the
manufacturer (Sigma Chemicals, St. Louis, Missouri). The
assay requires a spectrophotometer capable of working at a
wavelength of 340 nm. We immediately filtered samples
from the reactor through a 0.22 gtm filter (Millipore) at the
end of a 10 mL syringe, and removed cells to prevent glu-
cose consumption before the assay was complete.
We made optical density measurements at 660 nm, using
unfiltered samples. Because of non-linearities in the rela-
tionship of optical density to dry weight, samples with O.D.
> 0.300 were diluted with sterile medium. Sterile medium
was used as the blank.

The objectives of the first period are to measure kLa and
base-line (e.g., zero ethanol) growth parameters. One
fermenter is used exclusively for unsteady-state kLa
measurements. Suggested conditions are two agitator
speeds (250 and 350 rpm) and three air-flow rates (1.0, 2.0,
Chemical Engineering Education

and 4.0 L/min) at each agitator speed. Of course, this fer-
menter is not inoculated with cells.
We inoculate the second fermenter and use it to determine
the maximum specific growth rate, substrate utilization rate,
and specific respiration rate under aerobic conditions with
no ethanol added. Additionally, we determine dynamic kLa
values for selected values of agitator speed and gas-flow rate
corresponding to two of the conditions used in the unsteady-
state experiments.
We inoculate both reactors in the second lab period, and
we determine growth substrate utilization and respiration
rates upon the addition of known amounts of ethanol. One
reactor is challenged at 15 g/L ethanol, then 45 g/L, and the
second reactor is challenged at 30 g/L and then 60 g/L.

For the briefing, we give the students a general statement
of the problem along with the goals for laboratory periods
one and two. We also provide selected instructions from the
manuals for the New Brunswick Scientific Bioflow III fer-
menters, the enzyme assay kits for glucose and ethanol, and
the Spectronic Model 301 spectrophotometer. A standard
curve relating optical density at 660 nm to dry weight of the
culture (developed by the teaching assistant) is also pro-
vided. We then introduce the students to the equipment and
demonstrate the use of the software for fermenter control
and data logging. We also demonstrate the use of Eppendorf
pipetters since many of the students have not previously
used them.
Based on this information, we then ask the student group
leader to formulate an experimental plan and to assign duties
to the other group members. Without this kind of good
preparation it is impossible for the group to complete the
laboratory on time.
We discuss problems in the experiments and the format of
the report (written in some cases and oral in others) during
the report session. We ask the students to calculate or to
provide the following:

1. Calibration curves for ethanol and glucose
2. Sample calculation for kLa from the unsteady-state
3. Sample calculations for kLa from the dynamic method'31
4. Calculation of lt, the specific growth rate (h '), which is
defined as

dX / dt
=- (1)
during the exponential phase of culture growth'" where
X is the dry-weight concentration of cells (g/L)
5. Calculation of qo2 specific respiration rate
(mgO2/g cells-h) for oxygen consumption. A value for
Winter 1994

qo2 X is found using the procedure for dynamic kLa'2'
and dividing this value by X yields qo2.
6. Calculation of the specific glucose uptake rate, qj
(g.glucose consumed per g cells per hour). Note that
dS / dt dS /dt t (2)
qgu X dX /dt Yxs

where S = glucose concentration in the growth medium
(g/L), t = time (h) and Yxs is the yield coefficient or
mass of cells formed per mass of substrate (glucose in
this case) consumed.
7. Calculation of Yx/s.
Items 4 through 7 above are calculated from all experi-
ments (0 to 60 g/L of added ethanol).
Based on these calculations, each member of the group is
expected to develop a correlation of kLa with agitator speed
and air-flow rate, and this correlation is compared to ex-
pected dependency of kLa on agitator speed and air-flow
rates. The expected dependency can be found from the com-
bining expressions'l,3'41 such as

kLa Pg0.4 Q0.5 N0.5 (3)
where N = agitator speed (rpm), Q is volumetric gas-flow
rate (L/min), and Pg is the power requirement in the aerated
fermenter, with constant impeller diameter
2 0.45
Pg QP 0.56 (4)
S0-( 0.56
where P is the power required in the unaerated fermenter,
and with
p oN3 (5)
where Eq. (5) applies in the turbulent region.'41
For the analysis of kinetics, the students test by fitting
several possible kinetic expressions to the five available data
points for m, q02, and qg,,. We also asked them to comment
on the possible effects, if any on the yield coefficient.
Many equations have been suggested to describe product
inhibition.'21 Three examples are

9=V-a[- ( p Yn
Jimx PI a i
L = max/e- ip

t = max Kixp

In the above equations, .,ma is the maximum specific growth
rate (h'-) and can be determined from exponential growth at
zero ethanol concentration; P is the extracellular ethanol
concentration. Students evaluate tima from growth data dur-
ing the first period (for the yeast used in these experiments,
9tmax is about 0.4 h '). For these equations, we assume that the

glucose substrate is present at sufficiently high concentra-
tions that growth is zero order with respect to glucose. In Eq.
(6), P~ax is a semi-empirical parameter corresponding to the
highest ethanol concentration which will allow growth. This
value can vary significantly from one strain of yeast to
another; values of 90 to 120 g/L are typical. The exponent, n,
is empirical, but often a value of 0.45 is used. In Eqs. (7) and
(8), k, and Kix are empirical parameters. For the yeast strain
we used, Eq. (6) was the most satisfactory.

Equipment malfunctions, particularly with the dissolved
oxygen electrodes, should be anticipated, and the techniques
for preparing inocula should be followed consistently. Un-
der some stress conditions, this strain of yeast can develop a
pinkish pigment that can invalidate the optical density ver-
sus dry weight relationship.
We obtained generally good results for determining kLa,
particularly with the unsteady-state method. The kinetic data
are more problematic since few data points are available due
to the laboratory's time constraints. Further, the requirement
for a high level of precision in biomass and glucose mea-
surements is not met by every student group. Although the
actual techniques are straightforward, careful attention to
detail and sample handling are required. Groups with fewer

than three students are unworkable as the students are too
rushed for time to complete the assays carefully.
In summary, we believe that it is possible to introduce a
meaningful, challenging but doable, bioreactor experiment
to chemical engineering seniors who lack any prior exposure
to biology or bioprocessing. The use of two fermenters is
necessary if these goals are to be accomplished in two sepa-
rate laboratory periods of three hours. The student response
to the experience has been very positive.

Purchase of the equipment was supported, in part, by a
NSF Grant EID-9051404.

1. Shuler, M.L., "Bioprocess Engineering," in Encyclopedia of
Physical Science and Technology, Vol. 2, Academic Press,
Inc., New York, NY; 529 (1992)
2. Bajpai, R.K., and E.L. lannotti, "Product Inhibition," in
Handbook on Anaerobic Fermentations, L.E. Erickson and
D.Y-C Fung, eds., Marcel Dekker, Inc., New York, NY; 207
3. Wang, D.I.C., C.L. Cooney, A.L. Demain, P. Dunnill, A.E.
Humphrey, and M.D. Lilly, Fermentation and Enzyme Tech-
nology, John Wiley & Sons, New York, NY; Chap. 9 (1979)
4. Bailey, J.E., and D.F. Ollis, Biochemical Engineering Fun-
damentals, 2nd ed., McGraw-Hill Book Co., New York, NY;
Chap. 8 (1986) 0

REVIEW: Oscillations and Instabilities
Continued from page 17.
tionary states are calculated and analyzed, showing that
steady-state multiplicity can occur also in this case. The
effects of different kinetic mechanisms and boundary condi-
tions on the multiplicity pattern and the stability of the
steady-states are discussed.
In Chapter 10, the formation of stationary spatial patterns,
the so-called Turing structures, is considered for the
thermokinetic model, e.g., non-isothermal first-order kinet-
ics. First, the homogeneous steady-state is evaluated and its
stability character is determined. In the case where it is
stable to spatially homogeneous perturbations and the ratio
of mass and thermal diffusivities is sufficiently large, it is
demonstrated that stable spatial patterns can form due to
spatially inhomogeneous disturbances. On the contrary, when
the uniform steady-state is unstable to spatially homoge-
neous perturbations (e.g., the corresponding well-stirred sys-
tem exhibits limit cycle behavior), diffusion processes have
no stabilizing effect. Spatial patterns can form and survive
for a finite time, but eventually they decay to spatially ho-
mogeneous oscillations.
Chapter 11 deals with chemical traveling waves in a one-
dimensional space domain. The case of constant speed of
propagation is considered and its limitations are discussed. It

is shown how quadratic and cubic autocatalysis produce
traveling waves.
In Chapter 12, the broad issue of heterogeneous reactions
is addressed. The aim of this chapter is only to show that also
in this case, steady-state multiplicity and instability can oc-
cur due to non-linearities in the model equations. How these
non-linearities can arise is discussed in some detail. In par-
ticular, the cases of activated adsorption, multi-site reaction
mechanism and competitive chemisorption are considered.
The chapter includes some examples of steady-state multi-
plicity and oscillations.
The last chapter of the first part of the book, Chapter 13,
introduces the reader to the world of chemical chaos. After
clarifying that more than two state variables are required for
a continuous system to exhibit deterministic chaos, the au-
thors turn to simple discrete mappings to illustrate the strik-
ing phenomenon of the Feigenbaum cascade. Then the
periodic forcing of oscillatory systems is considered by
illustrating some techniques for their analysis and some ex-
amples of their behavior. Subsequently, complex oscilla-
tions and chaos in autonomous systems are dealt with and
the determination of the stability of limit cycles through the
computation of Floquet multipliers is described. Examples
are provided by a modified version of the autocatalator in an
isothermal batch reactor and by two consecutive exothermic
Chemical Engineering Education

reactions in a CSTR.
In the first chapter of the second part of the book, Chapter
14, the widely studied Belousov-Zhabotinskii (BZ) system,
as an example of a solution-phase reaction, is considered.
First, background information on the BZ oscillations,
its chemical mechanism and its simplified three-variable
model (e.g., the Oregonator) is given. Then the relaxation
oscillations of the Oregonator are analyzed in some detail
using the techniques developed in Chapter 5. The BZ
oscillations in flow reactors are considered next, introducing
the issue of bistability. The chapter is completed by the
analysis of the minimal bromate oscillator, where the or-
ganic substrate is omitted.
Chapter 15 presents results for several gas-phase combus-
tion reactions. In particular, hydrogen, carbon monoxide and
acetaldehyde oxidations are considered.
Indeed, the book reaches the dual aim stated by the authors
in the Preface. On one hand, it encourages more chemists "to
be less afraid of mathematics" by guiding the reader through
the colorful zoo of non-linear dynamics, using simple ex-
amples while avoiding the presentation and discussion of
theorems in rigorous mathematical terms. On the other hand,
those more familiar with the mathematical tools will find a
new opportunity to appreciate the richness of "the chemical
world" in terms of non-linearities.
The approach adopted throughout the book is rather
pragmatic. A clear indication of this attitude is in an intro-
duction to each chapter. It does not contain the usual list of
material treated in the chapter, but rather a list of items
which the reader should be able to accomplish "after a care-
ful study of the chapter."
In presenting each new dynamic phenomenon, first a nu-
merical example is presented in order to illustrate the physi-
cal picture, and then the mathematical tools for developing
an exhaustive analysis are described. The reader is never
involved in rigorous, high-level mathematical discussions,
while the mathematical developments are reported in full
detail. This permits smooth reading.
In the more complex area of spatially distributed reaction-
diffusion systems, the emphasis is placed on the representa-
tion of the various possible dynamic behavior. The descrip-
tion of the mathematical tools needed for the their determi-
nation becomes inevitably more vague.
The effort expended in relating the mathematical behavior
to the physico-chemical basis of the system is indeed re-
markable. For this, the authors analyze in parallel two differ-
ent models whose non-linearities arise from two different
sources. In the autocatalator, which has been introduced and
widely studied in the literature by the authors, the non-
linearity is in the cubic autocatalytic step. This model has
the merit, over the classical Brusselator and Oregonator, of
providing probably the best compromise between simpli-
Winter 1994

city and richness of dynamic behavior in an isothermal
closed system. In the non-isothermal model, where two irre-
versible first-order reactions are considered, the exotic dy-
namic behavior arises from the interaction of the reaction
thermal effects with the non-linear Arrhenius dependence of
the kinetic constant.
On the whole, this book provides an easy and convenient
entry into the difficult area of non-linear dynamics. While
there are not many independent courses existing which could
use this book as a text, it could certainly be used as a
supplementary text for graduate-level applied mathematics
and reaction engineering courses. It is most certainly a valu-
able reference for all who are interested in the dynamics of
reaction processes. 7

MR book review

by Phillip C. Wankat and Frank S. Oreovicz
Published by McGraw-Hill, Inc., New York, NY; 370pages, softcover
$32.95 (1993)
Reviewed by
C. Stewart Slater
Manhattan College
Teaching Engineering, by Wankat and Oreovicz, covers
all aspects of teaching engineering and does it in a clear and
concise manner. The text evolved out of an engineering
graduate course on educational methods taught at Purdue
University and a project funded by the National Science
Foundation. Materials from the text have been successfully
used in faculty-training courses given through the American
Society for Engineering Education as well as through other
Although it is oriented toward helping a new faculty mem-
ber to become an excellent and efficient teacher, the text can
also be quite useful for any engineering educator, and can be
used as the textbook for a graduate course for students who
are considering teaching as a career. The text can be used by
any engineering discipline and would certainly be suitable as
a resource for faculty outside of engineering.
The book is written in a pragmatic "how-to" style. This
method of concept-presentation allows an instructor to eas-
ily follow the points made on any of the various subjects.
Educational philosophy is incorporated when appropriate,
and extensive references are given for those who want more
information on a particular topic. Each chapter is effectively
divided into subsections and concludes with chapter com-
ments, summary and objectives, homework problems, and
The first chapter, "Introduction: Teaching Engineering,"
Continued on page 43.

M, -classroom


Just How Stable Are The Multiple Steady States?

Ben Gurion University of the Negev
Beer Sheva, 84105, Israel

his paper was prompted by a discussion with a col-
league who has been teaching chemical reaction
engineering for many years. During the discussion,
mention was made of the fact that when there are three
steady states in an exothermic continuous stirred tank reac-
tor (CSTR), the upper one can be unstable. The colleague
said that this is impossible, and he based his disbelief on the
classic plot of the multiple steady states, shown in Figure 1.
This particular plot was taken from Stephanopolous'" but it
appears in practically all the reaction engineering textbooks,
probably starting with the book by Levenspiel.[2'
This plot shows the curve of heat generated (A) and the
line of heat removed (B) versus the temperature in an exo-
thermic CSTR. The three steady states are the points of
intersection (P|, P2, and P3, Figure 1) of curve A and line B.
Let's assume that the reactor is started at temperature T2. At
this point the heat generated by the reaction (Q2) is greater
than the heat removed (Q2). This will cause the temperature
in the reactor to rise, and the rise will continue until the
upper steady state, P3, is reached. It is easy to show, using
similar arguments, that P, and P3 are stable steady states and
that P, is an unstable one. There is really no indication from
this plot that P3 could also be unstable.
1. Tel Aviv University, Tel Aviv, 69978, Israel
2. University of Connecticut, Storrs, CT 06269


Q; -- -- -


T, 2 Tr T3 Temperature

Figure 1. Three steady states of an exothermic CSTR
(adapted from [1]).

We can conclude that the use of a plot such as the one in
Figure 1 may lead to the misconception that the upper steady
state in an exothermic CSTR is always stable, but the roots
of this misconception are actually much deeper. They stem
from the mistaken belief that one can rely solely on results of
a steady state model to predict dynamic behavior. The steady
state model can certainly provide some guidelines, but a
dynamic model is needed to predict dynamic behavior.
It should be mentioned that there are textbooks (i.e.,
Westerterp, et al.,13' pg. 339) where dynamic analysis is
discussed in detail, based mainly on the pioneering work of
Aris and Amundson.'14 But in most reaction engineering
courses, only plots such as the one in Figure 1 are mentioned

Mordechal Shacham is Neima Brauner received Michael B. Cutlup re-
Professor and head of her BSc and MSc from the ceived his BChE and MS
Chemical Engineering at Technion, Israel Institute of from The Ohio State Uni-
the Ben Gurion University Technology, and her PhD varsity and his PhD from
of the Negev, Beer Sheva, from the University of Tel the University of Colorado.
Israel. He received his Aviv. She is currently As- He has taught at the Uni-
BSc and DSc from the sociate Professor in the versity of Connecticut for
Technion, Israel Institute of Fluid Mechanics and Heat the last twenty-five years,
Technology. His research -Transfer Department. She serving as Department
interests include applied teaches courses in Mass Head for nine years. His
numerical methods, com- and Heat Transfer and Pro- research interests include
puter-aided instruction, chemical process simula- cess Control. Her main research interests include catalytic and electrochemical reaction engineering,
tion, design, and optimization, and he is coauthor two-phase flows and transport phenomena in thin and he is coauthor of the POLYMATH numerical
of the POL YMATH package. films, analysis software.
Copyright ChE Division ofASEE 1994

Chemical Engineering Education

We can conclude that the use of a plot such as the one in Figure 1 may
lead to the misconception that the upper steady state in an exothermic CSTR is always
stable, but the roots of this misconception are actually much deeper. They stem from the mistaken belief
that one can rely solely on results of a steady state model to predict dynamic behavior.

as a practical means for analyzing CSTRs behavior. This
should not be the case any longer. The introduction of user-
friendly, interactive simulation packages which can solve
nonlinear algebraic or ordinary differential equations (ODEs)
has not only made the solution of dynamic nonlinear models
possible but has even made it easy.
In this paper we will demonstrate the use of one such
simulation package (POLYMATH) for analysis of the be-
havior of an exothermic CSTR. The POLYMATH package
is a numerical simulation package to be used with IBM and
compatible computers, and the current version (2.1.1 PC) is
distributed by the CACHE (Computer Aids for Chemical
Engineering Education) Corporation, a non-profit organiza-
tion for disseminating educational computer programs among
chemical engineering departments.*
POLYMATH has been used for almost a decade, and its
structure and possible applications have been described in
several publications."5-7] From among the programs included
in the package, the algebraic and ODE solver programs are
the most useful for exothermic CSTR analysis. (The alge-
braic equation solver was described in detail in reference 5.)
The main advantage of the POLYMATH ODE solver over
similar programs is that equations are typed in their math-
ematical form, and the user has to provide only infor-
mation regarding the mathematical model (equations, initial,
and final values). No technical information, such as integra-
tion method and step size, graph scaling, etc., has to be

Figure 2. Exothermic reaction in a CSTR.

* CACHE Corporation, P.O. Box 7939, Austin, TX 78713

provided. After the equations have been entered, the com-
puter time for solving even the most complicated problems
is only a few seconds.
For non-stiff equations, POLYMATH uses either an
explicit Euler's method or the fourth-order Runge-
Kutta method for integration. Euler's method is implemented
when the estimated integration error is less than 0.1 times
the error tolerance. For stiff equations the implicit Euler
method is used.
The structure of the rest of this paper is as follows:

In the next section we introduce an example prob-
lem. It is essentially the same problem as Luyben
presented.'' The problem definition is reproduced
for the reader's convenience.
In the third section, different combinations of
multiple steady states are demonstrated using a
steady state model, while the fourth section deals
with the analysis of the stability at different steady
states, using the dynamic model.
The model equations used for the reactor analysis
are given in the Appendices, in a form suitable for
use with the POLYMATH package.

The typical CSTR problems, in which a first order, exo-
thermic reaction is being carried out is presented in many
textbooks. We used a slightly modified form of an example
presented by Luyben.s81
An irreversible exothermic reaction A k B is carried
out in a perfectly mixed CSTR as shown in Figure 2. The
reaction is first order in reactant A and has a heat of reaction
X(BTU/mole A reacted). Negligible heat losses and constant
densities can be assumed. A cooling jacket surrounds the
reactor to remove the heat of reaction. Cooling water is
added to the jacket at a rate of Fj(ft'/sec) and an inlet tem-
perature Tjo(R). The volume of the reactor, V, and the
volume of water in the jacket, Vj(ft3) are constant.
The reaction rate coefficient changes as function of the
temperature according to the equation
k = a exp(-E / RT) (1)
The feed flow rate (F) and the cooling water flow rate (F)
are constant. The jacket water is assumed to be perfectly

Winter 1994


mixed. Heat transferred from the reactor to the jacket can be
calculated from

Q = UA (T T ) (2)

Q heat transfer rate in BTU/hr
U overall heat transfer coefficient in BTU/(sec)(ft2)(R)
A heat transfer area

The parameter values for the process181 are shown in
Table 1.
Taking into account that the inlet flow rate Fo is equal to
the outlet flow rate F, we see that dV/dt = 0, and the mole
and energy balances on the reactor and cooling jacket yield:

Vdt = F (CAO -CA)- VkCA

pCV =dT pCpFo (To T) ,VkCA UA(T Tj)

PjCVjTt PiCi(T^-T)+UA(T-TJ)
p1C~V3 -=- p3C3F3 (Tjo T ) + UA(T T )

At steady state these equations become


pCpFo (To T) ?VkCA UA(T Tj) = 0

pjCjF,(Tjo T)+UA(T T)=0

CSTR Parameter Values


40 ft3/hr
40 ft3/hr
0.50 mol/ft3
48 ft3
49.9 ft3/hr
3.85 ft3
7.08 x 1010 hr1
30,000 BTU/mol
1.99 BTU/mol R

(3) 8.000
6.000 -
(4) 1.o 10o-s 4.000
2. -QRX 10-5 2.000
(5) (Btu/hr) -ooo


Figure 3. Heat
(7) functions o


150 BTU/hr-ft2-oR
250 ft2
530 R
530 R
-30,000 BTU/mol
0.75 BTU/lbm-R
1.0 BTU/Ibm-R
50 lbm/ft3
62.3 lbm/ft3


0.000 540.000 580.000 620.000 660.000 700.0
T (-R)

removed (1) and heat generated (2) as
f temperatures when TO = 530 OR.

In the third and fourth sections we will discuss the number
of steady state solutions of these equation sets for the pa-
rameter values of Table 1.

There are several ways to solve the steady state equations
of the CSTR (Eqs. 6-8). The most obvious way is to solve
the three equations simultaneously, but this option has the
disadvantage that most solution algorithms will find only
one of the solutions. If there are several steady states, some
trial and error involving the initial estimates will be required
in order to find all the solutions.
Another option, one which will indicate all the steady
states, involves the preparation of plots similar to the one in
Figure 1. To accomplish this, we first must solve Eq. (6) for
CA and Eq. (8) for T3. This gives us

A F +Vk
Tjo + PT
TI= 1 + *3
Ti 1+P

where 13 = UA/(pjC1F).
Next, we define heat generated (Q,) and the negative

heat removed (-QR) as

QG = -VkCA (11)

-QR = -[pCpF(T- To)- UA(T- Tj)] (12)

In order to change the temperature in the reactor continu-
ously, a dummy differential equation
= 1 (13)

can be specified.
The set of equations consisting of Eq. (1) and Eqs. (9) to
(13) can be typed into the POLYMATH ODE simulator. The
form in which these equations are entered into POLYMATH
is shown in Appendix A. (Note that in the appendix the
notation "tr" is used for the temperature inside the reactor
and "tin" for the feed inlet temperature.) The numerical
values of the constants from Table 1 have already been
introduced into these equations.

The plot obtained by using the numerical values from
Table 1 is shown in Figure 3, above, which is very similar
(10) to Figure 1. The three steady states can be clearly iden-
tified as the points of intersections of the QR and Qo curves,
and the approximate temperatures at these points can
of be determined.
Chemical Engineering Education

The conditions can now be easily changed so as to get
different combinations of steady states. By changing the
reactor inlet temperature To, the heat removal line moves
parallel to itself and one or two steady state conditions can
be generated, as shown in Figure 4.
This is not the best way, however, to find the exact values
of the variables at the various steady states. To do that we
can rewrite Eq. (7) as a single nonlinear algebraic equation

f(T)= QG QR (14)

This equation, together with Eq. (1) and Eqs. (9) to (12) can
be entered into the POLYMATH nonlinear algebraic equa-
tions' solver program (as described in Appendix A). The
results for all three steady states for To = 530 R are summa-
rized in Table 2.

Once the steady states have been found, the most impor-
tant factor is how stable they are. We usually prefer to
operate the reactor at some particular steady state (most
often at the one with the highest conversion), but instability
at this steady state may cause many undesirable effects, such
as highly oscillatory response to small disturbances, or drift


Steady State T(R) Tj(R) CA(mole/ft3)
1. Lower 537.16 536.62 0.4739
2. Intermediate 599.99 594.63 0.2451
3. Upper 651.06 641.79 0.0591

to a different, less desirable steady state.
Stability at the different steady states can be determined
by calculating the eigenvalues of the state matrix of the
linearized model of the reactor. This method is widely taught
in process dynamics and control courses, but is not men-
tioned in any of the reaction engineering textbooks. Using
this method, the system of Equations (Eqs. 3,4,5) is linear-
ized at the vicinity of a steady state. Once the state matrix
which contains the multipliers of the state variables is con-
structed and its eigenvalues are calculated, the stability of a
steady state solution is determined by the sign of the real part
of the eigenvalues of the state matrix. If the real part is
positive, the steady state is unstable; a negative real part
indicates a stable steady state.
We have carried out such an analysis for the CSTR ex-
ample which was discussed earlier. We used two different
formulations of the problem. In the first formulation, we
assumed pseudo steady state with regard to the cooling wa-
ter temperature. That means that the differential equation,
Eq. (5), was replaced by the algebraic equation, Eq. (10).
The jacket's time constant is relatively small because of its
small volume, with the result that steady state assumption
reduces the stiffness of the problem and changes the result
very little. We will henceforth refer to this formulation as the
modified model.
In the second formulation, we used the basic set of equa-
tions, Eqs. (3), (4), and (5), and from this point on we will
refer to it as the basic model. The calculated eigenvalues are
shown in Tables 3 and 4.
For both formulations there is a positive real eigenvalue
for the intermediate steady state, indicating that this steady

1. Q x 10-5

2 -QRx 10-5


8.000 -
6.000 -

To =562R R
0 1 -' ^-

1. UUU
2.000 --

-0.000 -

-2.000 I I
00. 000 540.000 580.000 620.000 660.000 700.0
T( R)

8000 -

6.000 -



-0.000 -

7.000 -





- T =510R


T = 580' R

-3.000 1 I I I
500.000 540.000 580.000 620.000 660.000 700.0
T ( R)

Figure 4. Heat generated (1) and heat removed (2) as functions of temperature for various T, values.

Winter 1994

state is unstable. There is also a positive eigenvalue for the
upper steady state. In this case the absolute value of the
positive eigenvalue is much larger in the modified model
than in the basic model. That indicates that the upper steady
state will be unstable when using both formulations, but the
oscillations in the basic model will grow much slower than
in the modified model.

These results can be verified by simulation. The equations
that have to be typed into the POLYMATH ODE simulation
program are shown in Appendix B for both the modified and
the basic model.

Figure 5 shows the change of temperature inside the reac-
tor when it is started up at the three different steady states.
These plots were obtained using the modified model and
show that the reactor operation is as expected from the
theoretical analysis. The lower steady state is stable and the
intermediate state is very unstable, meaning that the tem-
perature starts to go down after about one hour and stabilizes
at the lower steady state after about five hours. In the upper
steady state the temperature first starts oscillating and finally
goes down toward the lower steady state.

The upper steady state can be further analyzed by looking
at the plot of heat generated versus temperature, shown in
Figure 6. It can be seen that for both the basic and the
modified formulation the heat generated creates a spiral
form where the growth rate of the spiral is much smaller in
the basic model. This is what is expected from the state
matrix eigenvalue analysis, but this plot is completely differ-
ent from the one in Figure 3 which was generated using the
steady state model.

It is interesting to note that when it is integrated for a long
enough time, the basic model will produce a limit cycle. 9i

This requires programs, however, that are "tuned" for inte-
gration of stiff equations for long time intervals with high
accuracy, and POLYMATH is not adequate.

The conclusion from these results is clear: using a steady
state model for predicting CSTR behavior at the upper steady
state can lead to wrong conclusions.

Several additional questions can be asked. First, is the
state matrix eigenvalue analysis really needed in order to
investigate the stability at different steady states? The an-

651.80 -
651.00 -
T (' R)
650.20 -
608.00 -
592.00 -
576.00 -
544.00 -
537.26 -
537.22 -

T('R) 537.18 -
537.14 -
537.10 -


I t


5.a. Upper steady


5.b. Intermediate
steady state

5.c. Lower steady

1.200 2.400 3.600 4.800 6.000

t (hours)

Figure 5. Temperature changes inside the reactor when
started at different steady states,
using the modified model.


( x 0- 5.305

(Btu/hr) 3oo -
5.292 -
5. 284 -


Qx l0-5

S.297 -
5.294 -


650.940 651.060 651.180 651.300 651.42

6.b. Basic

- I

5.285 i -

651.000 651.040 651.080 651.120 651.160
T ('R)

Figure 6. Plot of heat generated versus Tin the upper
steady state.

Chemical Engineering Education

Eigenvalue of the State Matrix Using
the Modified Model

Steady State 1st 2nd

1. Lower -1.446, 0 -0.953, 0
2. Intermediate -0.515, 0 3.504, 0
3. Upper 0.486, -2.86 0.486, 2.86

6.a. Modified

Eigenvalues of the State Matrix Using the Basic Model

Steady State 1st 2nd 3rd

1. Lower -188.7, 0 -1.267, 0 -0.976, 0
2. Intermediate -188.1, 0 -0.532, 0 3.049, 0
3. Upper -187.7, 0 0.00746, -2.754 0.00746, 2.754

i ll l

I m

swer in most cases will be no. Dynamic simulation can be
much easier and faster, and the real physical behavior of the
system can be observed, as opposed to observing indirect
indicators such as the eigenvalues.
Can the conditions in the CSTR be changed so that the
upper steady state is stable in the three steady state regions?
The reader can verify that such conditions exist by multiply-
ing the feed flow rate (Fo) by three and repeating the simula-
tion using the equations in Appendices A and B.
Is the instability of the upper steady state a result of the
varying cooling water temperature, and could it be pre-
vented if there were only two variables (T and CA)? The
reader can verify that this assumption is not true by fixing
the cooling water temperature at Tj = 5300R and using the
parameter values

Fo=40x10ft3/hr and cx=2x7.08x10"hr-'
instead of the values shown in Table 1. This set of param-
eters gives three steady states, with the upper one being

In this paper, we have demonstrated the applications of an
interactive numerical simulation package for location and
analysis of the steady states in an exothermic CSTR. We
showed that the use of a plot of heat generated and removed
versus temperature as the only means for analyzing the sta-
bility at the steady states may lead to wrong conclusions.
Also, that using this type of analysis sends the wrong mes-
sage to students, implying that they can rely solely on the
results of steady state models to predict dynamic behavior.
We have also shown that dynamic simulation is preferred
over other methods (such as state matrix eigenvalue analy-
sis) for testing stability at the steady states because it is easy,
it is fast, and the test is based on the real physical behavior
and not on indirect numerical indicators.

1. Stephanopoulos, G., Chemical Process Control: An Intro-
duction to Theory and Practice, Prentice Hall, Englewood
Cliffs, NJ (1984)
2. Levenspiel, O., Chemical Reaction Engineering, John Wiley
and Sons, New York (1962)
3. Westerterp, K.R., W.P.M. van Swaaij, and A.A.C.M.
Beenackers, Chemical Reactor Design and Operation, 2nd
ed., John Wiley and Sons, New York (1984)
4. Aris, R., and N.R. Amundson, "An Analysis of Chemical
Reactor Stability and Control," Chem. Eng. Sci., 7, 121
5. Shacham, M., and M.B. Cutlip, "A Simulation Package for
the PLATO System," Computers and Chem. Eng., 6(3), 209
6. Shacham, M., M.B. Cutlip, and P.D. Babcock, "A Microcom-
puter Simulation Package for Small Scale Systems,"

Microproc. and Microsyst., 9(2), 76 (1985)
7. Shacham, M., and M.B. Cutlip, "Application of a Microcom-
puter Computation Package in Chemical Engineering Edu-
cation," Chem. Eng. Ed., 121), 18 (1988)
8. Luyben, W.L., Process Modeling Simulation and Control for
Chemical Engineers, 2nd ed., McGraw-Hill, New York, p.
124 (1990)
9. Coughanowr, D.R., Process Systems Analysis and Control,
McGraw-Hill, New York, p. 502 (1991)

Steady State Model
Plotting Heat Generated and Heat Released versus Temperature
and Finding the Steady State Solutions.

(1) d(tr)/d(t)=l
(2) k=7.08*10**10*exp(-30000/1.99/tr)
(3) beta=150*250/(62.3*1.0*49.9)
(4) tj=(530+beta*tr)/(1+beta)
(5) ca=0.5/(1+48*k/40)
(6) qg=30000*k*ca*48
(7) tin=530
(8) rhocp=50*0.75
(9) qr=-(rhocp*40*(tin-tr)-150*250*(tr-tj))
t(0)= 0, tr(0)= 500
t(f)= 200
-To change feed temperature change tin value in
equation (7)
-To find the steady state solutions change Eqn
(1) to:
(1) f(tr) qg-qr
and use the algebraic equation solver program
To change feed flow rate change the number 40 ir
equations (5) and (9) to the desired value
To fix the cooling water temperature at Tj=530
change equation (4) to: tj=530.

Dynamic Simulation of the CSTR-Modified and Basic Models

(1) d(ca)/d(t)-40*(0.5-ca)/48-k*ca
(2) d(tr)/d(t)=(qg-qr)/(rhocp*48)
(3) beta=150*250/(62.3*1.0*49.9)
(4) tj=(530+beta*tr)/(l+beta)
(5) k 7.08*10**10*exp(-30000/1.99/tr)
(6) rhocp=50*0.75
(7) qg=30000*k*ca*48
(8) qr=-(rhocp*40*(530-tr)-150*250*(tr-tj))
t(0)= 0, ca(0)= 0.0581, tr(0)= 651.06
t(f)= 3
The initial values shown are close to the upper
steady state. To check additional steady states
use the values shown in Table 2 as initial
To change the feed flow rate change the number
40 in equations (1) and (8) to the desired
To fix the cooling water temperature at tj=530
change equation (4) to: tj=530.
To change from modified to basic model replace
Eqn (4) by the following equation:
(4) d(tj)/d(t)=49.9*((530-tj)+beta*(tr-tj) /3.85

Winter 1994

Random Thoughts...


5. Edward and Irving

North Carolina State University
Raleigh, NC 27695-7905

The scene is a dormitory room, shared by two senior
engineering students. Irving is hunched over his computer,
looking at an open manual next to the keyboard, as Edward
breezes in.
Ed: "Yo, Irv-shut it down and move it's party time."
Irv: (Silence)
E: "Come on, ace-the brew is losing its head...up and
I: "Chill out, Eddie-I'm trying to figure out how to in-
stall this upgrade on my operating system. Why don't
you go on ahead and I'll get there later?"
E: "Right-just like last week, when you were going to get
there in fifteen minutes and you never showed at all."
I: "I told you I got involved with the control homework
and lost track of time...anyway, you know I don't enjoy
these parties-you guys are lunatics."
E: "We can't be lunatics, we're engineers-we're all nerds,
we solve differential equations for kicks, most of us
wear glasses...besides, I knew the campus security guard
wouldn't really call the police last Friday-he just likes
to blow smoke. Here, I'll bet I can figure that out...a
few line commands here, a couple of mouse clicks
there, and we're off for the bright lights and the beauti-
I: "Eddie, get your grubby hands off that machine and let
me read the manual and do it right. Remember how
you were going to help me program my VCR to record
Star Trek last week, and you didn't need the instruc-
tions, and we ended up with a two-hour PBS special on
pancreas transplants?"
E: "That was only because I..."
I: "And how about that physics lab where you shorted out
the whole building? 'Let's just do it-lab manuals are
for weenies,' he says, just before the explosion."
E: "Yeah, but don't forget whose crazy idea got a patent
application on his summer job...your problem is you
spend so much time studying about what you're plan-
ning to do and worrying about why it might not work

@ Copyright ChE Division ofASEE 1994

that you never get around to doing it...but it's ok, read
all night if you can stand it, I'm out of here...oh, and
don't forget, I asked Jake and Marty and Amy and a
couple of the others to get together here tomorrow to
study with us for the design test."
I: "Dammit, Eddie, why do you keep doing this to me?
You know I study better alone-besides, you have an
attention span of about twenty seconds, and if those
jokers are over here you can forget studying or any-
thing else but..."
E: "No way-I'm really serious this time. I just like to
have people around-keeps things from getting too
I: "Too dull? You..."
E: "Later, my man. I'll save some foam for you..."
I: (Low growling noise)
Ed and Irv have been best friends since elementary school,
and no one was surprised when they enrolled in the same
engineering school and became roommates. What was sur-
prising was that they became friends in the first place, since
their personalities are polar opposites. Ed loves big parties,
and even if he doesn't know a soul when he walks in, every-
one knows his life story by the time he leaves. Irv, on the
other hand, doesn't like parties at all except for small quiet
gatherings of people he knows well. Privacy is a sacred
concept to Irv and a relatively alien one to Ed. They react
much differently when faced with unfamiliar tasks or situa-
tions. "Let's try this out and see what happens," says Ed, as
he dives in. "Hold on-let's think it through," responds Irv,
as he dips his toe in the water.
The two of them have dramatically different approaches to
schoolwork. Irv puts on some soft music, arranges his books
on his desk, and immerses himself. Even when Ed is there,
puttering around the room, fixing himself a snack, watching
TV, or even talking directly to Irv, Irv goes right on working,
occasionally mumbling responses to questions he really didn't
hear. Ed sometimes tries to work like that but can't do it; he's
constantly up and down, making comments about what he's
reading or asking Irv questions about it, and if he hears a
conversation down the hall or suspects that one might be
about to start, he's off like a shot to make sure he doesn't
Chemical Engineering Education

miss anything. He likes to see how others approach prob-
lems and to try out his solution ideas on them, and he drives
Irv crazy by assembling crowds to study or work on home-
work assignments when Irv wants to work in solitude.
Edward is an extravert and Irving is an introvert.* Al-
though the popular ideas of these terms (the extravert is the
one at the party wearing the lamp shade and the introvert is
the one hiding under the couch) are exaggerations, they have
some basis in reality. Extraverts tend to be gregarious and
active; introverts tend to be reserved and contemplative.
Extraverts are energized by being with people-the more the
better- while introverts find it draining to spend much time
with people they don't know well, and they may need to go
off somewhere by themselves afterwards to recharge their
batteries. Extraverts need to experience things to understand
them; introverts want to understand them first.
Science and engineering require the strengths of both
types-the thoughtfulness, capacity for sustained concentra-
tion, and desire for understanding of the introvert and the
quick thinking, verbal fluency, and willingness to take risks
of the extravert. Introverts may spend so much time thinking
about potential difficulties that they never quite get around
to trying out new ideas, while extraverts are comfortable
with trial-and-error learning and will not wait too long to
take action. Lacking the introverts' characteristic cautious-
ness, however, extraverts may get into trouble by jumping
into things before thinking them through, and being less able
to focus on one task for a long time, they are more likely to
accept superficial problem solutions. Extraverts are well
suited to jobs like technical sales and management that re-
quire strong interpersonal and communication skills and jobs
like consulting and emergency troubleshooting that require
quick thinking and responding, while introverts work better
in areas like research and design that allow them to take
information in, process it introspectively, and then respond.
While both extraverts and introverts can become excellent
scientists and engineers, the usual way these subjects are
taught-straight lectures, homework done individually, mini-
mal hands-on experience-stacks the deck in favor of the
introverts. Extraverts tend to have shorter attention spans
and find it hard to maintain their focus in long lectures. They
also do much of their best learning interactively-discuss-
ing, arguing, working out their ideas by bouncing them off
others; if they are forced to work individually all the time,

The degree to which one favors one or the other of these types can
be determined with the Myers-Briggs Type Indicator, a personal-
ity inventory based on Jung's theory of psychological types that
has been administered to over one million people, including many
engineering students and professors.r121 Ed and Irv are illustrative
of the two types but not all extraverts are just like Ed and not all
introverts are just like Irv. The two categories represent prefer-
ences, not mutually exclusive categories: the preferences may be
strong or weak, and all people exhibit characteristics of both types
to different degrees.
Winter 1994

they lose their most effective learning tool.
Several instructional techniques make classes more effec-
tive and enjoyable for both extraverts and introverts. Give
students several minutes of small-group exercises during
each class period-answering or generating questions, solv-
ing problems, or brainstorming. These exercises give extra-
verts occasions for activity and introverts opportunities to
reflect on the course material. Bring experimental demon-
strations-preferably hands-on-into lectures (for the extra-
verts) and give minilectures on interpretation of experimen-
tal results in laboratory courses (for the introverts). Use
interactive computer tutorials and simulations: extraverts
will enjoy the active learning they provide and introverts
will get practice in trial-and-error analysis in a relatively
risk-free environment. Assign some homework to teams of
three or four rather than to individuals. Some introverts may
complain about having to work in groups, but the extraverts
will appreciate getting to function in their preferred learning
mode for a change, and both types will learn the course
material better while improving their interpersonal, leader-
ship, and communication skills.[31
Epilogue: Ten Years Later. Following graduation, Ed
went to work as a product development engineer in the
polymer division of a chemical corporation and received
several patents for new membrane formulations. After two
years he decided that he enjoyed working with customers
more than synthesis reactors and extruders, moved into mar-
keting, and is currently associate marketing director in charge
of international sales. Irv went to work for an environmental
consulting firm, spent two years designing stack gas scrub-
bers, went back to graduate school for a PhD, and is now an
associate professor at a large university not far from where
Ed lives. They get together at least once a year. Ed always
proposes making the rounds of his favorite bars with some
drinking buddies he's sure Irv will like. Irv always looks
pained, makes some reference to lunatics, and counters with
a proposal to take in a chamber music concert or a poetry
reading. Ed rolls his eyes in mock disgust, says something
about "engineering nerds," and they compromise on dinner
with their wives at a good restaurant and drinks afterwards at
a quiet jazz lounge. They both thoroughly enjoy this routine
and wouldn't think for a moment of changing it.

1. Lawrence, G., People Types and Tiger Stripes, 2nd ed, Cen-
ter for Applications of Psychological Type, Gainesville, FL
2. McCaulley, M.H., E.S. Godleski, C.F. Yokomoto, L.
Harrisberger, and E.D. Sloan, "Applications of Psychologi-
cal Type in Engineering Education," Eng. Ed., 73(5), 394
3. Johnson, D.W., R.T. Johnson, and K.A. Smith, Cooperative
Learning: Increasing College Faculty Instructional Produc-
tivity, ASHE-ERIC High Education Report No. 4, George
Washington University (1991) 0

learning in industry

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 interns and coop
assignments typify such experiences; however, reports of more unusual cases are also welcome. Descrip-
tion of 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 Engi-
neering Department, University of Texas, Austin, Texas 78712.


Intensive Practical Education in Chemical Engineering

Massachusetts Institute of Technology
Cambridge, MA 02139

ne of the most satisfying engineering activities is
to apply an appropriate blend of theory, intuition,
and experience to the solution of a practical prob-
lem. In general, engineers learn their theory and principles in
the academic classrooms and laboratories; they cultivate
practical intuition and experience on the job. Naturally
enough, it may be asked if the academic training should also
include some practical development as well. The new engi-
neer could then approach a first job with some knowledge of
how it "really works." To this end, there exists a variety of
cooperative education and industrial internship programs.
In this article, we will describe a unique program of practi-
cal education at the graduate level. The David H. Koch
School of Chemical Engineering Practice (the "Practice
School") is administered by the Department of Chemical
Engineering at M.I.T. Students who are admitted to the
Practice School spend a semester working at off-campus
industrial stations. The semester at the stations replaces
the conventional Master's research thesis. Upon suc-
cessful completion of the Practice School, plus two
semesters of graduate lectures in Cambridge, students
are awarded the degree of Master of Science in Chemical
Engineering Practice.
The Practice School stations are maintained at host
companies, which provide office facilities and student
tuition and stipend support. Each station is staffed by a
Copyright ChE Division ofASEE 1994

Director and an Assistant Director (both faculty of the De-
partment residing year-round at the station) and a secretary
from the host company.
Practice School students work within the company, on the
company's problems, using the company's resources and
equipment, but they are not company employees-they work
for academic credit under the guidance of the station faculty.
They are given a good deal of responsibility for planning and
execution of the work, rather than perfunctorily performing
a predefined set of steps. For the host company, the Practice
School is like a small consulting firm, working for com-
pany clients within a format designed to accomplish a great
deal in a short time.

The Practice School is seventy-seven years old, and its
continuous operation has been interrupted only by the two
World Wars. It came about through the initiative of M.I.T.
alumnus A.D. Little and the support of Chemical Engineer-

Barry S. Johnston has been Assistant Professor and Director of the
Midland Station since 1992. He holds a PhD in chemical engineering from
Northwestern University. Prior to coming to M. I.T., he worked in the chemi-
cal and nuclear industries.
Thomas A. Meadowcroft has been Assistant Professor and Director of
the West Point Station since 1993. A Practice School graduate, he holds a
PhD in chemical engineering from M.I.T., where his thesis was on distrib-
uted control systems.
Aleksander J. Franz was recently the Assistant Director of the Midland
Station. A Practice School graduate, he is now a PhD student at the
University of Michigan.
T. Alan Hattan is Chevron Professor and Director of the Practice School.
He holds a PhD in chemical engineering from the University of Wisconsin.
He conducts research in novel separation processes and interfacial phe-

Chemical Engineering Education

ing Director W. H. Walker. Little, remembering his forma-
tive years in industry, sought to have industrial experience
made available to students on a regular basis. With funding
from George Eastman of Kodak, five stations were set up at
companies in the northeastern United States.
In the early years, each host company provided a labora-
tory and workshop for its station, but presently the Practice
School maintains no laboratories, and the students' time is
devoted entirely to company projects. What has not changed
is the residence of M.I.T. faculty at the stations and the
practical nature of the work.
There have been some thirty host companies for the sta-
tions over the years. There are currently two stations: the
Midland Station at the Dow Chemical Company in Midland,
Michigan, and the West Point Station at Merck Pharmaceu-
tical Manufacturing Division in West Point, Pennsylvania.
A comprehensive history of the Practice School and its
contributions to chemical engineering is available in a
monograph by Mattill.'"

Either between or after two semesters of graduate lectures
on campus, a class of five to eleven students spends one
semester in Practice School. These students all have
Bachelor's degrees (about half from M.I.T. and half from
other schools), predominantly in chemical engineering. They
stay eight weeks at one station, have one week off, and then
go to the other station for another eight weeks. At the end of
this schedule, each student will have worked on four techni-
cal projects, contributed to eight written reports, delivered at
least four talks, and led at least one technical team.
In the months before the students arrive at a station, the
resident faculty solicit project topics from
the host company, and together they pre- Sun M
pare projects, each designed to occupy a Informal Com
team of students for four weeks. They also Dinner at Orieni
identify the company resources that will be Director's Tot
needed to support the project. For each House Welc
project, the faculty write a Problem Din
Statement expressing the client's ob-jectives
and providing background information and
a suggested strategy for the students. The
students are divided into groups of two or
three, and in each group a group leader is
A typical work calendar is shown in Fig-
ure 1. After an orientation to the company
and the Practice School, each group re-
ceives a Problem Statement and is intro-
duced to its client. By the end of the first
week, each group submits a written Inves-
tigative Memorandum in which the stu-

Practice School students work within the
company, on the company's problems, using the
company's resources and equipment, but they are
not company employees-they work for academic
credit under the guidance of the station faculty.

dents demonstrate their understanding of the problem, present
relevant background information, and propose the method
by which they intend to accomplish their objectives. The
students must also exhibit a satisfactory understanding of the
safety requirements of their project. In the associated Pro-
posal Conference, one student from each group makes a
formal oral presentation to the clients, faculty, and students.
The Proposal Conference is a useful forum for review and
modification of project plans.
One week later, another student from each group presents
an oral Progress Report. By this time, significant progress is
expected, and members of the audience often respond with
valuable suggestions. With one week remaining, each group
leader chairs an informal meeting of the group, the client,
and the faculty in which the form and substance of the
remaining work are negotiated and agreed upon. The project
culminates in the writing of a Final Report (occasionally the
size of a small Master's thesis) and the oral Final Presenta-
tion by the group leader. The faculty critique two prelimi-
nary versions of the report before accepting the final draft. A
typical Final Report outline is shown in Figure 2.
The schedule is repeated for a second project, with new
group assignments and leaders. At the conclusion of the two
projects, the faculty give letter grades and evaluate each
student with respect to technical ability, creativity, motiva-
tion and initiative, leadership, and communication skills.

on Tue Wed Thu Fri Sat

pany MIT Slide Proposal Canoeing
station; Orientation; Reviews Conference;
urs; Students Investigative
:ome Meet with Memorandum
ner Clients Due
Revised Group Slide Progress Volleyball
Investigative Leaders Reviews Report; and
Memorandum Meet Group Barbeque
Due with Photo
Informal First Draft Movie
Progress of Final
Report Report
at Due
Second Slide Final Assistant Mackinac
Draft of Reviews Presentation Director Island
Report Conference; Trip!
Due Final Report
figuree 1. First-month Practice School calendar.

Winter 1994

For the host company, the Practice School is like a small
consulting firm, working for company clients within
a format designed to accomplish a great
deal in a short time.

The students perform a similar evaluation of each other. For many it
is the most thorough discussion of their strengths and weaknesses they
have ever received. After a week's break, the students go to the next
station for two more projects, following a similar schedule. Grading
and evaluation at the second station are independent of that at the first.
The objectives of a Practice School project are deliberately am-
bitious. Students normally find they must work sixty to eighty hours
a week. Careful planning of the work and effective organization of
the group thus become crucial skills. Working in groups is often
a new experience for the students. The group leader, in particu-
lar, can gain valuable experience in planning, allocating resources,
encouraging the team, and making decisions. Each student has at least
one turn as group leader.
The faculty attempt to keep abreast of the students' progress,
challenge their thinking, supply information and suggestions, and
direct them to resources. The students work as needed with a variety
of host company personnel such as librarians, laboratory technicians,
research scientists, technology specialists, computer experts, plant
operators, and process engineers. Both faculty and students are
bound by confidentiality agreements in their handling of the host
company's information.
The students live in apartments that cost about as much as they would
have to pay in Cambridge. To provide some diversion, the faculty
arrange weekend activities that vary with the preferences of the indi-
vidual class and can be as simple as dinner and a movie or as challeng-
ing as bicycling, cross-country skiing, or canoeing. During the week-
end between projects there is time for an overnight trip to an area resort
for skiing or sightseeing.

Consistent with the wide range of technologies and activities that fall
within the province of chemical engineering, students in the Practice
School may expect a diversity of work topics. The majority of projects
involve an operating process, but the work may also include research,
design, or simulation. Three case histories are described below, with
the descriptions written to give a clear idea of the students' work while
at the same time protecting the host company's proprietary information.
Each of the projects was accomplished in four weeks, including written
and oral communication of the results.

CASE 1: Refrigerated Separation
Figure 3 shows a continuous separation process. The column feed and
overhead streams are cooled by a cascade refrigeration unit that provides
coolant at two temperatures, the higher for the feed cooler and the lower for the
condenser. The clients wanted to increase production at an upstream reactor
and felt that the column and refrigeration unit were limiting.
The students began by determining the material and energy balances. They
obtained flows, temperatures, and pressures from plant instrumentation and

requested chemical analyses of several streams, using this
information to develop and validate an ASPEN PLUS model
of the process. A check of the column temperature profile
suggested that little separation was being achieved in the
upper portion of the column. After performing staging cal-
culations, the students concluded that the reflux ratio could
be reduced without affecting the overhead purity, as shown
in Figure 4. This would reduce the load on the refrigeration
system as well as increase column capacity.
The students then analyzed the operation of the refrigera-
tion unit. They quantified the split of refrigeration capacity
between the two sides of the cascade, showing that the
overall capacity is increased as the cooling is diverted from
the lower to the higher temperature side. The column stag-
ing calculations indicated that more feed cooling and less
overhead condensing would not adversely affect the sepa-
ration. Hence, they recommended operating conditions that
increased both column and refrigeration capacity.
Beyond the immediate production increase from changes
in operating conditions, the students also estimated the
capacities of several heat exchangers, both within the re-
frigeration unit and associated with the column. From this
they provided the clients with a list of equipment upgrades
to allow further production increases. Furthermore, the stu-
dents traced the service piping and recommended valve
settings that would improve coolant distribution without
requiring hardware modifications.
With the client's permission, the students supervised a
plant trial to test their recommendations. Following the
students' instructions, the plant achieved a record pro-
duction rate while maintaining product specifications.

LIST OF FIGU RES ................................ ........................4...
LIST OF TABLES ................................... .......................5...

1. SU M M A R Y ................................. .........................6...
2. INTRODUCTION ........................... ..................... 7
2.1 B background ....................................................... 7
2.2 O objective ................................ ..................... 11
2.3 M ethod of Approach ...................................... 11
3. PROCESS MODELING ........................ 12
3.1 ASPEN PLUS M odel..................................... 12
3.2 V alidation ............................ ........................ 18
4. PLANT TESTS ....................................................... 24
4.1 Proposed Operating Conditions .....................24
4.2 Results of Testing ......................... 26
5. PROCESS BOTTLENECKS ..................................32
6. CONCLUSIONS............................ ...................... 39
7. RECOMMENDATIONS ........................................ 40
8. REFERENCES........................ .............................. 41
9. NOMENCLATURE ............................................... 42
10. ACKNOWLEDGMENTS ......................................43
11. A PPEN D IX ................................. .......................... 44
11.1 ASPEN PLUS Input Files.............................. 44
11.2 Summary of Plant Data.................................. 51
11.3 Reactor-Separator Optimization Procedure...... 61
11.4 Location of Original Data .............................. 63
DISTRIBUTION LIST.................................................... 64

Figure 2. Table of Contents from a
typical Final Report
Chemical Engineering Education

Company engineers subsequently incorporated the stu- dent's
ASPEN PLUS model into a comprehensive model of the process,
which has proved useful in further op-timization studies.

CASE 2: Reactive Batch Distillation
A product is made by sequential substitution at active sites
A-B3 +3C A-C3+3B
The batch process is run in a steam-heated kettle, and volatile
compounds are separated in an overhead still. After an initial distil-
lation to remove by-product B, the kettle is run under vacuum to
remove excess reactant C, which is then recycled to the kettle to
begin the next batch. Sharply increased demand for the product
necessitated a production increase. In addition, there were unac-


Cascade Refrigeration Unit

Figure 3. Process schematic of refrigerated separation.


Reflux Ratio
* High

-o- Medium

- Low

10 20 30 40 50 60

I Temperature

Figure 4. Variation of column temperature profile
with reflux ratio
Winter 1994

ceptable variations in product consistency from batch to batch.
The students approached the problem in three ways. One stu-
dent began laboratory experiments to generate a consistent set
of kinetic data; another attempted to model the batch distillation
using BATCHFRAC software; and a third began reviewing the
process operating data. It sometimes happens that an initial ap-
proach must be abandoned. In this project, the batch modeling
was dropped and the students concentrated on process experi-
mentation to effect improvements.
Chemical analyses allowed the unsteady material balance to
be determined. The students discovered that by rearranging
process steps, several non-productive steps could be omitted
entirely. In particular, they could reduce the contamination of
recovered reactant C with by-product B. They reached their con-
clusions from the scrutiny of process data coupled with basic
stoichiometric calculations.
The students ran several trial batches in the process equipment to
test their recommendations, which required some adjustment of
sleep schedules to accommodate round-the-clock production. The
simplified batch scheme increased the purity of recovered reactant
C, which in turn improved product consistency from batch to batch.
In addition, product yield increased. Hydraulic calculations indi-
cated that the column pressure drop was significantly higher than
expected. Subsequent examination of the column proved that the
packing was crushed. The clients have realized significant produc-
tion gains from the students' work. In addition, the students' kinetic
data have proved useful to company engineers designing a new
continuous process for the product.

CASE 3: Quantitative Risk Assessment of a Storage
A volatile and flammable chemical is stored in a refrigerated
tank. The tank is protected from overpressure by rupture disks, but
the clients were concerned that the protection might be inadequate,
especially in case of fire.
The students examined the equipment and the safety procedures.
They constructed fault trees leading to a BLEVE (boiling liquid/
expanding vapor explosion) or a UVCE (unconfined vapor cloud
explosion). They made heavy use of the literature, as well as
interviewing plant safety organizations. From these sources they
assigned probability values to the steps in the fault trees and thus
derived the overall probability of each incident.
The students found and applied correlations for damage from
projectiles, blast waves, and thermal radiation. They deployed the
PHAST code to calculate the dispersion of the chemical in an
atmospheric release and thus estimated the consequential damage
to the plant and to the surrounding community.
Having specified the paths by which incidents could occur, the
students identified several improvements to operating procedures
and safety equipment. They designed a new pressure-relief system,
using an in-house code. Heat transfer calculations led to recom-
mendations for the number and placement of water deluge nozzles.
From their risk assessment, they could express quantitatively the
benefit to be gained from implementing each recommendation. The
estimated probability of incident was reduced by two orders of
magnitude. The final report was abstracted for inclusion in the host
company's process safety guidelines.

The Practice School is an educational program operated
for the benefit of the students, but it can only continue if the
host company feels that its money is being well spent. For
the company, the Practice School provides teams of talented
engineers who can mount energetic attacks on important
tasks. During their relatively short stay in Practice School,
students are undistracted by the multiple duties and concerns
of regular employees and are able to direct undivided atten-
tion to the problem at hand. The focus and intensity of the
students' efforts is often inspiring to company employees.
What the students lack in practical engineering ex-
perience, they sometimes make up for by a fresh approach
to problems. Often the solution to a problem is found in
the creative assembly of company resources. The students
do make original contributions, and occasionally they
demonstrate new techniques of analysis which are then
picked up by the company. Company personnel who de-
vote a few hours a day in supporting the student group
can see their investment produce a significant amount of
accomplished work.
Clients are generally pleased with the quantity and quality
of the work. Responses to surveys in recent years indicate
that 90% of the clients would like to have another Practice
School project. Many prospective clients, however, fear that
the students will require too much of their time, and this
possibility is especially troublesome for oversubscribed pro-
duction supervisors whose plants are running lean on per-
sonnel. While the short duration of a project is appealing to
some clients, others have requested that projects be longer,
or that less time be spent in writing and presentations.
Students attending Practice School may experience an
unprecedented level of professional involvement. The host
companies offer them important tasks; decisions will be
made and money spent, based on their work. Students are
excited to find that their efforts have resulted in significant
cost savings or production increases; they acquire the confi-
dence which follows the accomplishment of a demanding
task. From observing other groups, they appreciate the vari-
ety of activities and applications of engineering.
While knowledge from academic classrooms is offered as
separate subjects, a Practice School project is likely to re-
quire that this separately acquired knowledge be inte-
grated. Thus the reactor performance may be limited by the
heat exchanger area, and the distillation column fails be-
cause of reboiler piping. Furthermore, what the homework
problems normally gave as background information now has
to be obtained or estimated. The students must quickly as-
similate and deploy procedures and software that may be
new to them. Since different companies may use different
tools for similar purposes, the students must adapt to what is
available at each station.
The students benefit from working with experienced engi-

neers and scientists at the host companies. This can give
them a deeper understanding of particular technical con-
cepts, practical details of equipment operation, shortcut de-
sign and estimation techniques, or a sense of what consti-
tutes reasonable and realistic industrial practice. Experienced
engineers can illustrate how problems are best approached
and what to watch out for. The students may also observe
the sorts of jobs available in industry, which may influence
their own career decisions. In addition, the students have
more access to faculty guidance than in the typical campus
lecture course.
Finally, the students are able to improve their communica-
tion skills. Reports are not only written, but they also must
be revised. The faculty are concerned with content and pre-
sentation, questioning the choice of verb as well as the
accuracy of the energy balance. Each student's oral presenta-
tion is evaluated for delivery as well as for composition.
Before each talk there is a faculty review of the visual aids,
to assess both the organization and the clarity of the mate-
rial. The Informal Progress Report offers a chance to im-
prove meeting skills. Working in groups, the students gain
practical knowledge of interpersonal relationships, some-
times under stressful circumstances. The experience of be-
ing a group leader is particularly motivating to previously
shy or reserved students.

The intensity of the Practice School would be impractical
to maintain for more than a short time-students can find the
experience exhausting. Each group has written and revised
two documents and prepared three oral presentations each
month; they have often had to adjust sleep schedules to
follow production runs; they have met and listened and read
and discussed and debated and calculated and defended until
they want no more.
Appreciation of the benefits seems to grow with time.
Practice School graduates have been particularly strong in
their subsequent support of the program and the Depart-
ment.121 Students later tell us that through the Practice School
they gained a sense of how much can be accomplished, how
decisions can be made from incomplete and contradictory
information, how scarce resources can best be allocated, and
how new information can quickly be assimilated.
The Practice School does not pretend to be the only way to
learn; however we are confident that its contribution to chemi-
cal engineering education is unique and valuable.

1. Mattill, J.I., The Flagship: The M.I.T. School of Chemical
Engineering Practice, 1916-1991, David H. Koch School of
Chemical Engineering Practice, Massachusetts Institute of
Technology, Cambridge, MA (1991)
2. Mattill, J.I., "M.I.T.'s School of Chemical Engineering Prac-
tice: The Powerful Potential of Alumni Support," Chem.
Eng. Ed., 27(3), 154 (1993) 0
Chemical Engineering Education

REVIEW: Teaching Engineering
Continued from page 29.

sets the tone for the book by describing the importance of
teaching. An overview of the components of good teaching
is presented along with a list of learning principles.
Chapter 2 focuses on efficiency, which is important
for faculty in both teaching and research. Since most
faculty have a myriad of responsibilities, this chapter helps
by dealing with topics such as setting goals, establishing
priorities, and maintaining "to-do" lists, so that efficiency
can be increased.
Chapter 3, "Designing Your First Class," is an excellent
step-by-step guide for any new assistant professor and leads
the reader into the following chapters for more detailed
information. Course objectives and textbooks are the subject
of Chapter 4; the topics include taxonomies of educational
objectives, teaching approaches, textbook selection, and ad-
dressing ABET course requirements.
Chapter 5 covers problem solving and creativity, both of
which have been the focus of many engineering studies over
the years. Development of effective problem solving strate-
gies is very important to an engineering student's success
and the subject is concisely presented here. Teaching stu-
dents to be creative is also addressed. This chapter has a
particularly thorough reference listing.
Chapter 6 describes lecture format and style. Since lectur-
ing is by far the most popular style of teaching engineering,
this chapter is quite important-improving one's lecture style
significantly benefits any course. The chapter treats topics
such as the advantages/disadvantages of lecturing, improv-
ing lecture content, organization, performance aspects, and
interaction with students. The problem of how to effectively
teach large class sizes is also addressed.
Chapter 7, "Nontechnological Alternatives to Lecture,"
presents some options to the lecture format, such as discus-
sion, cooperative group learning, panels, debates, and "quiz
shows,"-all of which are used in other educational fields.
Independent study, mastery learning, and self-paced instruc-
tion are also covered.
"Teaching with Technology," Chapter 8, describes some
of the delivery techniques useful to teaching. The delivery
medias profiled include television and video, and comput-
ers. The audiotutorial method is also mentioned.
"Design and Laboratories" are featured in Chapter 9. Both
topics are quite important to engineering education and are
crucial to the accreditation process. Although they are not
extensively covered in this chapter, they are effectively sum-
marized. General aspects of incorporating design throughout
the curriculum, as well as teaching design projects, are pre-
sented. Laboratory structure for different student levels is
also reviewed.
Winter 1994

Chapter 10, "One-to-One Teaching and Advising," cov-
ers listening skills, tutoring and helping students, and
advising and counseling strategies. Proper advising of re-
search students from undergraduates through doctoral
candidates is presented.
Chapter 11 reviews the various aspects of "Testing, Home-
work, and Grading," while Chapter 12 explores "Student
Cheating, Discipline, and Ethics." It is designed to assist
faculty in preventing cheating and also addresses how to
incorporate the subject of ethics into the curriculum.
"Psychological Type and Learning," Chapter 13, focuses
on the natural differences among students which need to be
considered when teaching plans are made and in the teacher's
interactions with students. Both Piaget's and Perry's theories
of cognitive development and their application to engineer-
ing education are presented in Chapter 14. Learning theories
are explained in more detail in Chapter 15, which includes
further discussion of learning and teaching styles, Kolb's
learning cycle, and student motivation.
"Evaluation of Teaching" is covered in Chapter 16, and
topics ranging from promotion and tenture to professional
development are mentioned in Chapter 17. Information for
graduate students interested in finding an academic position
is presented in Appendix A, and a sample outline for a
course on teaching is presented in Appendix B.
Overall, this is an excellent book. It brings together all the
topics necessary for developing as a superb teacher. The
authors incorporate a significant amount of material into the
370 pages and do so in a way that is easy to follow and to use
for improving one's performance as an engineering educator.
I am sure that students would want to see this book on the
professor's required reading list for next semester! O

Me book review

Vol. 1: Mass Transfer
Edited by John J. McKetta
Marcel Dekker, 270 Madison Ave., New York, NY 10016
$350for set of Vols. 1 & 2 (1993)

Reviewed by
Scott Lynn
University of California, Berkeley

In the Preface, the editor of this handbook describes it as
"up-to-date" and "presented by world authorities in their
specialties." That statement is, perhaps, only half-accurate.
Many of the contributors are certainly the grand old men in
their respective areas-but the overside of the title page
notes that the contents of this volume were previously pub-
Continued on page 57.

MRa -class and home problems J

The object of this column is to enhance our readers' collection of interesting and novel problems in
chemical engineering. Problems of the type that can be used to motivate the student by presenting a
particular principle in class, or in a new light, or that can be assigned as a novel home problem, are
requested, as well as those that are more traditional in nature and which elucidate difficult concepts. Please
submit them to Professors James O. Wilkes and Mark A. Burns, Chemical Engineering Department, Univer-
sity of Michigan, Ann Arbor, Ml 48109-2136.



Texas A&M University-Kingsville
Kingsville, TX 78363

Emphasis on design considerations has led to a criti-
cal state apparatus that is easier to fabricate and use
than what has been offered in the literature. There,
focus has been on CO2 as the working fluid in capillary
tubes."'31 When Halpern and Lin1" followed the technique of
Banna and Mathews,t2] their tubes "failed to demonstrate
critical behavior." Halpern and Lin present an extensive
experimental portion that convincingly conveys that their
tubes work, but it is not clear whether this is due to usage
technique or fabrication technique. The previous two papers,
as well as one by Smith and Boyington,[31 all basically use a
capillary-fill technique that involves introducing CO2 gas
and then condensing it with external liquid nitrogen cooling.
Smith and Boyington also do a solid CO2 load, but this is
generally regarded as unsatisfactory due to difficulties of
quantifying granular solid CO2 volumes.

The essence of design, compared to, say, analysis or syn-
thesis, is the open-endedness of the question. It is recognized
at the outset that multiple solutions to a design problem
could work, and at least several should be initially consid-
ered. Contrast that philosophy to the uniformity in the scope
of the three cited papers-their exclusive consideration of
CO2, for example, may be due to the oldest paper in the
series131 or to a physical chemistry laboratory text.141 Perhaps
the roots of influence go back even further. Moore15' begins
his discussion of the critical region by pointing out that the
first gas to be studied in the critical region was carbon
dioxide-work done by Thomas Andrews in 1869. Dodge161
implies the industrial importance of CO2 in the introduction
@ Copyright ChE Division ofASEE 1994

to his chapter on refrigeration and reinforces that impression
with an extensive section on solid carbon dioxide processes.
Design, on the other hand, should respect history and in
this case might begin with the priority criterion to make
critical phenomena visible, perhaps even to highlight the
vanishing meniscus or "critical opalescence"17' that occurs in
transition through the critical state. Realizing that these ef-
fects could be expected in any critical state situation, one
might find the question shifting: "What working fluids might
have critical conditions most readily obtainable in the labo-
ratory, and can they be made visible?"
Some working fluids with mild critical conditions are
gaseous at ambient laboratory conditions and have a particu-
lar disadvantage in that special efforts are needed to handle
them compared to solids or liquids. Thus, true to the genuine

Ron Marcotte is a professor of physical chemis-
try at Texas A&I University. He earned his BS
and PhD at the University of Florida, and his
research interests include the study of gas be-
havior and gaseous reactions.

Luis Zepeda is a computer services engineer
for Mobil Chemical Company. He is currently
working on a process control modernization
project. He attended Texas A&I University and
earned his BS in chemical engineering from the
University of Texas at Austin.

Dale Schruben is a professor in the Chemical
and Natural Gas Engineering Department at Texas
A&I University. He holds BS degrees in physics
and nuclear engineering from Kansas State Uni-
versity, and chemical engineering MS and PhD
degrees from the University of Minnesota and
Carnegie-Mellon University, respectively.

Chemical Engineering Education

design process, easing the difficulties in attaining a critical state
for a system must be balanced against the difficulties of preparing
the system. Ambient liquids, unlike ambient gases, could be
injected into capillaries by syringe. Solids at ambient conditions
have the quantification difficulties noted earlier. Ambient liquid
and other working fluid candidates with critical temperatures well
below the annealing point of Pyrex (about 823 "K) are repre-
sented in Table 1, along with some comments.
The foregoing comment on Pyrex annealing leads to the ques-
tion of system-container materials of construction. Glass, quartz,
or, at milder conditions, a synthetic could be considered. We
proceeded with glass for reasons of performance and economy.
Too often in the past our profession has suffered by its emphasis
on performance and economy-safety must also be a factor.

Critical pressures are above atmospheric and so safety from
explosion must be considered in addition to toxicological safety.
Pressures that will be generated can be calculated from a thick
wall formula involving an inside and outside radius as well as a
simpler thin wall capillary tube formula"81

2 2
S= Rpi
R2 2 Pi
0 -Ri

- piD

S = maximum stress that may be experienced in the tube material
Ro = outside rube radius, 3.5mm
R, = inside tube radius, 0.75mm
p, = pressure inside the tube
D = tube diameter, equal to 2 Ro, 7.0mm
t = tube thickness, equal to R, R,

For the 7-mm capillary tubes used and with the 33.7 bar critical
pressure of n-pentane as an example, the two formulas in Eq. (1)
yield (with 1.0133 bar/atm and 14.7 psi/atm) comparable results
of 536 and 622 psi, respectively, for the experienced stress. Perry's
Chemical Engineering Handbook, pages 23-60, gives the maxi-
mum stress (a yield stress or close relative, modulus of rupture) of
6,000 to 10,000 psi for borosilicate (Pyrex) glass. This represents
a safety factor of about ten and is intended to secure occasional

Working Fluid Candidates
Crical Parameters
TempJK PresJbar Comments

Carbon Dioxide 304.2
Isobutane 408.1
n-Pentane 469.6
Diethyl ether 466.7
11 -Trichlorotri-
fluoroethane 422.2
Perfluoroethane 487.2

Solid (STP); hard to measure
Low T,; gas (STP)
Liquid (STP); low toxicity
Liquid (STP); low toxicity

20.4 Liquid (STP); low toxicity; not common*
34.1 Liquid (STP); low toxicity; not common*

mild excursions above the critical pressure.

The "Safety" section above explicitly mentions the cap-
illary, and the earlier sections hinted that a capillary glass
tube was used. Thermal transport questions include:
Can the capillary volume be heated uniformly and can
molecules exchange, on the microscopic scale, between
liquid and gas states, sufficiently to ensure quasi-equilib-
rium? On both points a spherical cavity might be best.
Thus, if a capillary is used, shortness would be desirable.
However, sufficient length should be provided so that
definition of a meniscus is promoted and meniscus activ-
ity is easily followed.
The cited papers11-31 produced capillary tubes of approxi-
mate respective lengths 20, 35, and 20 cm. It is not clear
why the long lengths were used (since shortness is desir-
able) unless the manifolding arrangements in some way
required it.
Our end product was a 10-cm capillary, and fabrica-
tion began by sealing one end of the capillary with a
gas/oxygen torch used for glass blowing. The small por-
tion of the other end was drawn out in a tapered neck. This
made the sealing step easier and may have helped in the
purging process. Presence of the slight restriction due to
the tapered neck may have promoted arrest of backflow.
The length of the tube was measured and the required
(quantitative information follows) height of the liquid
meniscus noted. At that height (with pure saturated
vapor above it) the contents in the constant volume capil-
lary will have the critical specific volume when heated to
the critical temperature.
The detailed technique is to inject more than the re-
quired amount of liquid with a syringe through the capil-
lary neck. Low toxicity fluids like n-pentane or diethyl
ether will, at ambient, partially vaporize and purge the
capillaries of all gases except the pure vapor. This purge
continues until the correct liquid amount is obtained. The
tube is then quenched in a cold bath, and the tube neck is
sealed with a blob of molten glass and annealed with a
second or two of reducing flame. With some practice, the
slightly drawn neck is easily reheated, quickly and specifi-
cally, without disturbing the rest of the container. The
molten glass blob at the end of a glass rod is applied with a
tamping acting that again thickens out the previously
slightly drawn neck. Previous wall thicknesses (or greater)
upon which the safety calculations were based, can be
achieved. Again briefly, Figure 1 illustrates the situation
within a second of sealing. The shape of the drawn neck is
indicated. Filling with the syringe has been straightfor-
wardly accomplished. The tube is in its cold bath with
flame and partially melted glass rod ready. The shape after
the closure, tamp, and anneal steps is also shown. With


From Refs. 10 and 11; other from Ref 7.

Winter 1994

practice, one person of normal talent can perform the purge and
seal step, but it is easier with two people.
Since the working fluids have critical temperatures of about
2000C, tubes can be placed in a glycerol or silicone oil bath. (As an
aside, another advantage of these working fluids over CO2 is that
special cooling, as with a fire extinguisher,13' is not required to
rapidly cool from supercritical down to temperatures at the upper
edge of the critical state. The ambient air and higher critical
temperature here provide sufficient gradient for cooling, the point
being to save time in the set up, demonstration, cool off, and take
down.) Slow heating with stirring of the bath causes the tubes to
move slowly through the critical state. These shorter tubes evi-
dently have none of the mass transport resistance to equilibrium
(and the attendant need to be rotated) noted by Halpern and Lin.
This in itself is a significant simplification of procedure.
In cases where the critical composition is not obtained during
fabrication, the meniscus slowly rises or falls until it disappears,
which indicates a saturated liquid or saturated vapor condition,
respectively, the temperature of which can be noted. Thus the
saturation boundary ("steam dome") can be constructed directly on
a temperature versus specific volume plot. Figure 2 shows this
with our data. We have estimated the uncertainty involved in
reading, simultaneously, the slowly rising temperature and the
slowly moving meniscus. Specific volume, v, is found from room
temperature liquid and vapor specific volumes v, and Vg, along
with meniscus height, h, and the tube length, L, from the relation
v= xvg +(1-x)v,; x = I (2)
1+ hv
(L -h)v
There is no need to introduce the new parameter"'
a= 1-x
because it is trivially related to x, the "quality" parameter as
defined and widely used in many engineering thermodynamics
texts as mass of vapor over total mass for a pure component in a
control volume. A plot of saturated liquid and vapor specific
volume vs. temperature may offer better economy of data than
plotting either alpha or x in the saturation region vs. temperature.
The first is a line while the second two would yield areas, and as
such have a manifold increase in data-base size. From another
point of view, however, the extra data may be reassuring and even
helpful if one is uncertain what the plot is turning out to be in a
continuous sense as the discrete, necessarily discontinuous, ex-
perimental points are plotted. Essentially the same information is
displayed in either plot.
When a tube of critical composition is carefully heated or cooled,
the meniscus of Figure 3 gently vanishes, as shown. Then the so-
called "critical opalescence" is reached,171 varying from a smoky to
a twinkly appearance. Notice also, that the curvature of the liquid
surface relaxes as the critical state is reached.
With practice, the tube can be mounted on a stand with a test-
tube clamp and a hand-held propane torch can be used to heat it,

circumventing the need for a temperature bath. Hand
movement compensates for uneven cooling sufficiently
to demonstrate the critical opalescence. A stand-up trans-
parent reach-around shield should be used; ours is ap-
proximately 1/4-inch thick Lexan or similar material that
is standard issue in many laboratory supply catalogs.
This apparatus makes a short (five- to ten-minute) dem-
onstration that can be offered as a vital lecture stimulus to
a small number of students in a classroom setting, al-
though it is ideally suited to the laboratory.
Several ideas have been mentioned for expanding this
to a mini-exercise for the Class and Home Problem sec-
tion of this journal, such as calculating the liquid height,
h, to give a critical specific volume. A starting point
might be the quality,171 identified here as
x =Mg (L h) Apg
M (L- h)Apg +hAp,
(Dividing the right side by the numerator yields Eq. 2.
The letter v, for specific volume, is just the reciprocal of
the density; the gas and liquid densities are indicated,
again, by the g and 1 subscripts, respectively.) Note that
Mg and M are the gas and total mass, respectively, and
that h and L refer to liquid height and total height in the
closed capillary. Capillary cross-section is A. Some re-
arrangement of Eq. (2) yields the L over h ratio
S1+ hp,
x (L h)pg

.-+( (3)

The densities and quality must still be found. Quality can



n7 Y





Figure 1. Illustration of the drawn neck, flame, and
molten glass rod at the moment of quench and seal.
Chemical Engineering Education

be obtained from Eq. (2)
vc =xVg+(1-x)v, (4)
The quantity 304 x 106m3/mol[7] can be combined with the
72 molecular weight to yield the critical specific volume, v,.
The Handbook of Physics and Chemistry (HPC)1oi value
for the liquid density, about 0.63 g/cc, agrees well with the
Rackett equationt71
V1 = V Z(1-Tr)02857
of 0.66 g/cc; where
T, = 298/466.7
Z, = 0.262
v, = 304/72 cc/g
Iterative approaches such as the Redlich/Kwong approach
could have also been used.
In addition to having a remarkably complete tabulation of
quantitative thermodynamic property data, Smith and Van
Nessl7' also has a wealth of such information in the problems.
On page 319, for example, Antoine's equation is given for n-
pentane. It predicts to many places the saturation pressure,
temperature pairs given on page 2378 of HPC. (That data
would make a wonderful basis for constructing an Antoine-
type equation. That is not our purpose here, but it could be
included-students need to go through such an effort at least
once in their career to appreciate correlations and other sources
for such information.) The saturation pressure at, say, 25C,
could be used with the ideal gas law to find Vg = 503 cc/g, but
we cannot expect this to be very accurate. It could be used as a
starting value for an iterative scheme171 for a more accurate



160 -

1 2 3 4 5 6 7

Figure 2. The temperature vs. specific volume plot
by the described techniques.

Figure 3. Progressive meniscus relaxation as it vanishes
at critical conditions and "critical opalescence" appears.
Winter 1994

value. It is on the order of the 3-to-4 g/1 that one might expect
from Section 3-72 of Perry's Chemical Engineering Hand-
book,"' where ethane and butane are 1.4 and 2.6 g/l, respec-
tively. n-Pentane is not listed there, but let us assume the 4 g/1
value, or vg = 250 cc/g. This could also be used in an iterative
scheme (3 cycles with Redlich/Kwong yields 318 cc/g). With
v, = 1/0.63 cc/g and vc having the only value before calculated,
304/72 cc/g, Eq. (4) will yield x = 1.1% with Vg = 250 cc/g or
x = 0.83% with vg = 318 cc/g. In either case, Eq. (3) then will
lead to h = 0.37 L when rounded to two significant digits.
Other issues could be pursued. We could start with Eq. (3)
and form

L 1+ l-- = I+ P I(5
h i-x p i-x -y

We might like to ignore the gas-to-liquid density ratio here by
considering the parameter y as unity minus the density ratio.
Dropping the density ratio is introducing error into y. Propa-
gation of error (y approaching 1) in standard fashion

( )ay
produces a singular form, as would direct comparisons of
exact and approximate expressions. The singularity can be
cast in terms of densities, and 1'Hopital's rule will close. Prob-
ing density uncertainties, however, might be more meaning-
ful, with the characteristic form
8(L / h) 8p + 5Pg
(L/h-1) pg p

Professor Van Ness showed encouraging interest and of-
fered wise comment in this endeavor.

1. Halpern, A.M., and M.F. Lin, J. Chem. Educ., 63, 38 (1986)
2. Banna, M.S., and R.D. Mathews, J. Chem. Educ., 56,838 (1979)
3. Smith, S. Ruven, and R. Boyington, J. Chem. Educ., 51, 86
4. Shoemaker, D.P., and C.W. Garland, Experiments in Physical
Chemistry, McGraw-Hill, New York (1962)
5. Moore, W.J., Physical Chemistry, Prentice-Hall, Englewood
Cliffs, NJ (1965)
6. Dodge, B.F., Chemical Engineering Thermodynamics, McGraw-
Hill, New York (1944)
7. Smith, J.M., and H.C. Van Ness, Introduction to Chemical
Engineering Thermodynamics, 4th ed., McGraw-Hill, New York
8. Singer, F.L., Strength of Materials, 2nd ed., Harper & Row,
New York (1962)
9. Perry, R.H., ed., Chemical Engineering Handbook, 5th ed.,
McGraw-Hill, New York (1984)
10. Handbook of Chemistry and Physics, 41st ed., Chemical Rub-
ber, Cleveland, OH (1960)
11. Dean, J.D., ed., Handbook of Organic Chemistry, McGraw-Hill,
New York (1987) O





King Saud University
Riyadh 11421, Saudi Arabia

Usually, one objective of a laboratory experiment is
measuring the values of a dependent variable as it
changes with an independent variable, or determin-
ing the values of some characteristic constants or parameters
of a system. It is also useful to teach the students how to
build up a mathematical model of the system under investi-
gation and to think of possible analytical or numerical solu-
tions to the model.
Models which describe the transient period of the system
(e.g., start-up or shut-down) contain differential equa-
tions which can be solved either analytically or numerically
(if the model is relatively complex). Steady-state models
(with uniform dependent variables) usually contain alge-
braic equations.
The objective of this paper is to simulate the dynamics of a
CSTR during different stages of its continuous operation.
Mathematical models will be developed, along with analyti-
cal and numerical solutions, and together, these will be com-
pared with experimental results.

The reaction between ethyl acetate and sodium hydroxide
in aqueous solutions is demonstrated in many chemical engi-
neering labs. It is characterized by its constant density, and it
is safe and easy to operate and analyze. The reaction rate is

well established, and the mechanisms have been discussed
in detail. The overall reaction rate is regarded as a second-
order reaction, particularly at low temperatures." 3"
In this study, the experiment is operated batchwise to
determine the order of the reaction and the reaction-rate
constant. These kinetic data are used to simulate the dynam-
ics of the continuous mode operation.

The saponification of ethyl acetate occurs quickly, so it is
more convenient to measure the concentration by following
the change in conductivity than to titrate aliquots since the
latter need sampling, quenching (to stop the reaction in the
sample), and back-titration.
In a batchwise mode and at zero time, one liter of sodium
hydroxide (0.1 N) and one liter of ethyl acetate (0.1 N) are
thoroughly mixed at room temperature (23'C). The conduc-
tivity is recorded at suitable intervals of time, and the con-
centration can be read from a suitable calibration curve. The
order of the reaction and the rate constant can be deduced
from these batchwise data by a suitable differential and/or
integral analysis. Both methods give a second-order and a
reaction-rate constant, k = 6.1 1/mol-min.
Continuous operation, which is the objective of this study,
is illustrated in Figure 1. Solutions of sodium hydroxide and
ethyl acetate (at equal concentrations) are pumped first to
the head tanks (to eliminate fluctuations in flow rate caused
by direct pumping to the reactor) and then to the reactor at
suitable, but equal, flow rates. The speed of the stirrer is
adjusted by the stirrer speed control, while temperature con-
trol is achieved through the heater temperature control and
the cooling coil. The conductivity is recorded at suitable
intervals of time. Notice that at zero time the reactor is
empty. During the startup and filling of the reactor, both
concentration and volume change with time until the reactor
overflows; then the concentration will change with time
until steady state is reached.

Chemical Engineering Education

Aziz M. Abu-Khalaf is a member of the chemi-
cal engineering teaching staff at King Saud
University. His main interests are in mathemati-
cal modeling, corrosion, and controlled-release
@ C systems.
Copyright ChE Division ofASEE 1994

Once students determine the order of reaction and the rate
constant from the batch data and become familiar with the
system, they are in a position to mathematically model and
develop a transient analysis of the continuous mode opera-
tion of the stirred reactor. Three stages can be modeled:
1. From beginning to overflow
2. From overflow to steady state
3. Steady state operation
Obviously, the first and second stages are transient, while
stage three is represented by a steady-state model. In this
study, analysis is restricted to identical concentrations and
flow rates of reactants. We let
C = molar reactant concentration in the reactors;
Co = molar reactant concentration in feed; mols/liter (M)
F = flow rate; liter/min
k = reaction rate constant; liter/mols-min
r = kC2, the rate of reaction; mols/min-liter
t = time; min
V = volume of the reacting system; liter
Stage One
This stage is semibatch. There is no output because the
reactor contents do not yet reach the overflow level. A
material balance on either NaOH or ethyl acetate (both reac-
tants are at the same concentration and flow rate) gives:

rate of accumulation = rate of input rate of consumption.
d (VC) = FCo VkC2 (1

V +C = FC VkC2 (2
dt dt
But V is a function of time, and since the system is o
constant density and flow rate, a total mass balance gives

Feed tanks

Figure 1. Reactor set-up
Winter 1994

Once students determine the order of reaction and
the rate constant from the batch data and become
familiar with the system, they are in a position
to mathematically model and develop a
transient analysis of the continuous
mode operation of the stirred reactor.

dV= F

or V = Ft

since at t = 0, V = 0. Equation (2) then becomes

Ft dC + CF = FC0 FtkC2

dC Co C kC2 (3)
dt t t
Equation (3) is subject to C = Co at t = 0. This equation can
be solved using the substitution

u = exp(k C dt)
Equation (3) becomes, after some manipulation,

t2 d+ t du kCotu=0 (4)

which can be solved via Bessel functions by using the substi-
tution z = t112. Equation (4) becomes

2 du + z du 4 kCOZ2u = 0 (5)
dz2 dz
which is a modified Bessel equation, the general solution of
which (in terms of t) isE41

u = AIo(4kCot)+BKo(4 kCot) (6)

where Io and Ko are the modified Bessel functions of the first
and second kind of order zero, respectively, and A and B are
constants to be determined by the boundary conditions and
f the nature of the problem.
At this point, students can be asked to carry out several
interesting exercises:
1. Perform the steps required to get Eqs. (4) and (5).
2. At t = 0, u = 1, greatly simplifying Eq. (6). Show that B must
3. We are looking for a relation between C and t; thus Eq. (6)
should be written in terms of C.
4. If task 3 above is performed, it can be shown that

C = I(2 kC-t)

where I, is the modified Bessel function of the second kind
of order one. This equation shows that C is not defined at
t = 0, although at this initial time C = C0. It is a good exercise
for the students to show that this condition is implicitly


satisfied by expanding the modified Bessel functions, con-
sidering only the first few terms.

Stage Two
The second stage is continuous, but not yet steady. The
concentration is changing with time but the volume of the
reactants is constant. A material balance takes the form:
rate of accumulation =
rate of input rate of output rate of consumption
S= FC -FC kVC2 (8)
and therefore
dC Co C kC2 (9)
dT k (9)
T=t-T = time in minutes
T=V/F = time constant
At steady state, C = Cs, which is a particular solution to Eq.
(9). Substituting
C = Cs + (10)
changes Eq. (9) to a linear first-order differential equation
which can be solved by the integrating factor method. The
final solution is

C 1C, (11)
B exp(AT) -
A = 1+2kCs

B k
C, -C, A
and C, is the concentration at the beginning of stage two
(e.g., at t = r).
Students can be asked to derive Eq. (11) by starting from
Eqs. (9) and (10). It is interesting to check Eq. (11) at T = 0
and at steady state (e.g., as T -> o) and to think of a special
form of the equation suitable for short times.

Stage Three
This is the easiest stage to model. A material balance
requires that
rate of input = rate of output + rate of consumption

FCo = FC + Vr


FCo = FC, + kVC2
k tC +C -C = 0 (13)
where Cs is the steady state concentration. If k is known, C,
can be predicted, or if Cs is measured, k can be calculated.

Notice that Eq. (13) can be derived directly from Eq. (9)
where at steady state dC/dT = 0 and C = C_.

Numerical solutions are required whenever analytical so-
lutions are not possible or are difficult to obtain. A suitable
Runge-Kutta subroutine (e.g., IVPRK or DIVPRK from
IMSL) can be used to solve the initial value problems, Eqs.
(3) and (9). Concentrations from Eq. (7) can be calculated
using IMSL special functions (DBSIO and DBSI1). Equation
(3) is not defined at t = 0, but it can be shown (by Taylor
series expansion) that

C-)| =- 1 kC2
dt I t=o 2
This condition should be considered in the numerical solu-
tion. Notice that the initial condition for stage two is that
C = C, and T = 0 (or t = T), where C, is the concentration at
the end of stage one.

ITime, min
Figure 2. Analytical and numerical profiles (perfectly
matched) ofreactant concentration for different initial con-

Figure 3. Analytical and numerical profiles (perfectly
matched) of reactant concentration for different initial con-
Chemical Engineering Education

Equations (3) and (9) were solved numerically and the
solutions were compared to the analytical solution from Eqs.
(7) and (11). Figures 2, 3, and 4 show these theoretical
profiles for different cases of initial concentrations and for
different values of rate constant (F = 0.23 1/min, V = 2.8 1).
Notice that both numerical and analytical profiles coincide,
i.e., they match perfectly. Both the analytical and the
numerical solutions are compared to experimental results
(shown in Figure 5) for the case of 0.1 N initial con-
centration, a volume of reactants of 2.5 1 and a total flow
rate of 0.24 l/min.

Continuous operation of the CSTR setup as shown in
Figure 1 can serve several useful functions:
1. It can foster a realization of some practical problems, such
as fluctuations inflow. A possible remedy is the use of head

0 10 20 30 40 50 60 70
Time, min

Figure 4. Analytical and numerical profiles (perfectly
matched) of reactant concentration for different rate
constant values.

(gravity) tanks. This could serve as a practical design
2. It can spur students to think of start-up procedures that
achieve efficient transient operations, e.g., how to introduce
the reactants into the vessels, and when and how to heat (or
cool), if necessary.
3. It can encourage development of reliable mathematical
models that describe the performance of the continuous
system. This reliability should be checked against experi-
mental data and reasons for any discrepancies between
theory and experiment should be discussed.
Theoretical performance of the continuous mode opera-
tion is shown in Figures 2, 3, and 4. There is almost no
difference between numerical and analytical solutions, as
can be clearly seen from the figures. They match perfectly.
This confirms the correctness of the analytical solutions.
Both solutions are compared with experimental results, as
shown in Figure 5. A discrepancy between theoretical and
experimental results is noticed in the initial period of the
start-up operation. It is not difficult to predict, theoretically,
the concentration at any time, but limitation of solutions in
the reactor at the beginning of the start-up makes it difficult
to accurately measure the concentration. Stirring during this
initial period causes air bubbles, and this will affect the
performance of the conductivity meter. A possible solution
to this problem is to increase the flow rate (in new runs) or to
measure the concentration using other means. Students should
think of possible limitations to these solutions.
Students will find it interesting to consider the possible
reasons for this discrepancy. They should first think of ex-
perimental errors and then check the assumptions of the
model in order to think of new models, if they are required.
They will quickly come to a realization of the difference
between models and reality.
Another interesting point is the approach to steady state
concentration. As can be seen from the figures, a steady state
condition is approached asymptotically. Thus, depending on
the significant figures required by the experiment, different
values of experimental steady-state concentrations can be
recorded, and they will represent a percentage of the theo-
retical steady-state value which can be achieved from stage
three by using Eq. (13).

The author wishes to acknowledge helpful discussions
with Professor M.A. Soliman.

1. Bender, M.L., Chem. Revs., 60, 53 (1960)
2. Tsujikawa, H., and H. Inoue, Bull. Chem. Soc. Japan, 39,
1837 (1966)
3. Potts, J.E., and E.S. Amis, J. Am. Chem. Soc., 71, 2112
4. Wylie, C.R., and L.C. Barrett, Advanced Engineering Math-
ematics, McGraw-Hill Book Company, NY (1982) 0

Figure 5 Theoretical profile of reactant concentration as
compared to experimental results.
(Co = 0.1 M, V= 2.5 1, F = 0.24 1/min).
Winter 1994

[9@ curriculum
-- .____________lo



Sophomore and Junior Years

West Virginia University
Morgantown, WV 26506-6101

he recent information explosion in science and engi-
neering requires careful selection of what topics to
present to the undergraduate. In the limit, as the
amount of material covered in class approaches infinity,
student comprehension approaches zero; the inverse is also
true. Both limiting conditions are intolerable. It is important
to back away from saturation coverage and to focus on the
material that future engineers will need.
The future engineer will need to understand (analyze,
optimize, synthesize, evaluate) the interaction of a series of
steps that work together to achieve some desired result. The
engineer (today and in the future) must be able to
Work in teams to solve a problem
Identify the need to know information
Locate material to satisfy these needs
Learn the content material
Solve the problem with appropriate methodology
Communicate effectively
The design sequence presented in this paper requires stu-
dents to practice and develop all of these skills. They learn to
recognize the interrelationship between courses; they learn
that the courses are not isolated, only to be forgotten as soon
as a grade is received.
Our faculty members agree that a common goal of produc-
ing graduates with the above skills is the most important

aspect of our undergraduate program. Under the leader-
ship of two or three faculty, all other faculty teaching the
affected courses participate in the design projects. We dis-
cuss the undergraduate curriculum and adjust course empha-
sis if necessary; if an instructor feels that students do not
have adequate mastery of a prerequisite subject, similar ad-
justments are made. We believe that any design project
which requires students to learn for themselves, which dem-
onstrates that engineering is more than a series of well-
defined problems at the end of some chapter, which demon-
strates that engineering knows no traditional course-work
boundaries, and which forces students to synthesize material
as it is learned, will benefit the student and, ultimately,
produce a better engineer.
We recently introduced an integrated design project that
spans the sophomore and junior years. A single chemical
process design is analyzed, synthesized, and evaluated over
the course of these two years. The first-semester sophomore
project focuses on material balances and applies simple eco-
nomic criteria based on the costs of raw materials, products,
and by-products. Each subsequent semester requires addi-
tional knowledge, and by the end of the junior year the
design yields an economic optimization of an improved pro-
cess which the students synthesized and which required se-
lection of operating conditions and sizing of chemical reac-
tors, heat exchangers, pumps, compressors, separators, and
recycle rates. The economics include both capital costs and
operating costs such as pay-back period (time value of money
is not introduced until the senior year).
These are group design projects that are integrated into

Richard C. Bailie received Joseph A. Shaelwitz re- Wallace B. Whiting is As-
-. 1 his degrees from Iowa ceived his degrees in chemi- sociate Professor of Chemi-
State University (PhD), cal engineering from the Uni- cal Engineering at West Vir-
Wayne State University versity of Delaware (BS in ginia University, where he
(MSChE), and Illinois In- 1974) and Carnegie Mellon has taught for the past de-
stitute of Technology University (MS in 1976 and cade. He is active in ASEE
(BSChE). His interests are PhD in 1978). His research and AIChE, and his re-
in fluidization and energy interests are in mass trans- search and teaching inter-
utilization, and he has pub- fer, especially in pharmaceu- ests range from thermody-
lished many articles and a tical systems, and in design namics to process safety
book in these areas, and design education. and process design.
Copyright ChE Division ofASEE 1994
52 Chemical Engineering Education

existing courses-three in the sophomore year and five in
the junior year. Different grades are assigned for all team
members on the project, based on an evaluation of the con-
tent of the design project relevant to that course. More de-
tails are presented elsewhere.P]
The rationale for this approach is uncomplicated. We want
to prepare our students for a group senior-year design project,
for a sequence of individual comprehensive problems re-
quired in the senior year, and ultimately, for a forty-year
career in chemical engineering. These senior-year design
activities are described elsewhere.[2'31
As a result of the design projects, students develop a
number of personalized strategies for life-long learning-
they learn self-evaluation and experience team work, they
recognize the role of economics in decision making, they
appreciate the need to understand basic principles, and they
understand the various sources of engineering information.
Since (group) oral and written reports are required most
semesters, they also learn the importance of developing
communication skills.
In addition to the direct impact on students, faculty are
provided with the necessary input to assess student perfor-
mance in their application of engineering principles. The
feedback provided by the senior-year design projects affect
the content of the sophomore/junior design projects. This is
one mechanism we use to ensure that graduates meet mini-
mum standards of knowledge and skills. The feedback di-
rectly measures learning and aids in curriculum develop-
ment and improvement.
ABET is in the process of changing the design and engi-
neering science requirements to eliminate the "bean count-
ing" in favor of an "integrated design experience" through-
out the curriculum, ending in a capstone course. The pro-
gram described in this paper is one type of integrated experi-
ence which appears to satisfy these new criteria.
This paper will briefly summarize the design sequence,
review some of the experiences the students had during the
sequence, describe changes in student development, report
on the "student culture" that has evolved, and attempt to
explain why these changes took place. The paper will focus
on the first-semester junior design for the 1992-93 year to
illustrate how the process operates.

A single chemical process is the basis for the design se-
quence during the sophomore and junior years. Each subse-
quent semester's design requires additional knowledge and
more detail, including mastery of the previous design. All
chemical process designs used for this sequence are symbol-
ized by a generic process block diagram (see Figure 1).
These include four essential elements: a pre-reactor system,
a chemical reactor system, a recycle stream, and a post-
reactor system (i.e., one or more separation units).
Winter 1994


Figure 1. Generic chemical process

We introduce first-semester sophomore students to a simple
process flow sheet that includes a reactor, a separator, and a
recycle stream. We give them cost data for feed and product
streams; we provide several feed stocks and recycle rates
and require the students to select operating conditions.
In the second semester we give the students a more com-
plicated flow sheet that includes heat and work units. Utility
costs are provided and are included in the evaluation. Stu-
dents learn that heating, cooling, and power cost money. The
advantage of high conversion at elevated pressures is offset
by the high cost of running the gas compressors. This affects
the selection of operating conditions. As the students' under-
standing of the process is enhanced, the quality of their
decisions improves.
In the first semester of the junior year, we cover thermody-
namics, heat transfer, and fluid flow. For the first time,
students learn how to calculate the area of heat exchangers,
to evaluate the work/heat requirements for systems of staged
compressors with intercoolers, to determine the number of
adiabatic (equilibrium) reactors in series, to handle the non-
ideal behavior of gas mixtures, and to determine the size of
process piping. All of these studies are needed for the new
design. For the first time, capital costs are considered in the
analysis. The optimum operating temperature and pressure
changes because of this change in the objective function.
The final design in the second semester of the junior year
differs in one major aspect. We do not give the students a
flow diagram as a starting point. We provide kinetic infor-
mation that yields different reactor performances and require
the students to examine the separations units of the post-
reactor system. Combining their experience from the previ-
ous designs with the new information on separations and
kinetics, they synthesize a new, improved, process.

Details of the
First-Semester Junior-Year Design Project
A few details, taken from the first-semester junior year,
are presented here, and they illustrate the types of activities
that result from the design sequence. Recent first-semester
juniors involved in their third design continued their investi-
gation of a process for the production of ammonia from
synthesis gas. Figure 2 shows the process flow sheet provided

to the students with the problem statement. It is a "cari
ture" chemical-process diagram and includes certain attribi
that have been distorted. They are not tricks, nor are tl
necessarily realistic. They are often naive design choi
meant to focus students' attention on a specific concept. T
examples are absence of any heat integration and compre
ing a vapor when pumping a liquid is possible. From I
flow sheet, we expect students to discover a need to kn
more about separations and reaction kinetics to develop
credible design. This established groundwork for the folk
ing semester's content. The "caricatures" included in Fig
2 and the problem statement are:
1. High concentration of CO2 in the feed
2. Questionable solubility data for CO2 in liquid NH3
3. Fixed reactor cost independent of temperature, pressure,
concentration, and flow rate
4. Single-stage compressorforfeed
Students started this design with valuable experience gail
from previous projects with a somewhat simpler ammo
process flow sheet. They had discovered the high cost
gas compression and the effect of conversion on proc
profitability (that the higher conversions were obtain
at lower temperatures and higher pressures). The probl
statement had provided cost information on feed mater
equipment, and a wide range of utilities along with introd
tory relationships for estimating the capital costs for ma
equipment units.
The students soon realized they were not able to evalu
chemical conversion for this new feed material and did
understand how to reduce the cost of compression. Also,
relationship for heat-exchanger costs required knowlec
of the area, and they could not calculate the area. They co
not deal with non-ideal gases. They had unmasked seven
"needs-to-know." These needs were satisfied by the ci
current class work.
Figure 3 is a composite design that represents the
students' response to the project. The major character-
istics are

The reactor operated at low pressure (in spite of lower
A three-stage compressor replaced the single-stage
A 3-7 stage adiabatic reactor with intercooling was
used to obtain high conversion
A single-stage condenser replaced the two-stage
High recycle rates were required.

While group solutions differed in detail, all the groups
presented a modified flow sheet that was an improve-
ment over the one they were given. They had all cor-
rectly identified the need for class content, had learned

that content, and had applied it to the design. They had
gained the confidence to challenge and to make changes to
the original flow sheet.
Students are traditionally trained to mimic what they have
seen and heard, but this seldom translates into an ability to
make meaningful changes. Individually, they fear looking
foolish-but as a group they are braver, more willing to
express their ideas and opinions. Previously, they had prob-
ably had little practice in challenging things that were pre-
sented to them-the "caricature" problem helped them rec-
ognize that they can contribute better ideas.
Students wanted to use a chemical process simulator to
perform the many required calculations, but prior to their
junior year, our students are not allowed to use a simulator
for class problems or projects. Having observed seniors work-
ing with the software, the juniors wanted to exploit this tool.
Faculty encouraged them to use it whenever appropriate;
however, no class time was spent in instructing students on
its use. The juniors solved this problem by finding knowl-
edgeable colleagues (in this case, seniors) and asking them
how to use it. In response, the seniors set up individual
tutoring sessions and were available to help as needed. (This
required a user-friendly, interactive, readily available simu-
lator.) Given this self-discovered need-to-know, the juniors
picked up the fundamentals quickly, gaining proficiency
with practice. They took the initiative and learned what was
necessary on their own. This is how our sophomores and
juniors learn about plotting, spread-sheeting, and graphics
software; no formal class time is devoted to these activities.

In navigating a passage from the starting point represented
by Figure 2, and the destination represented by Figures 3,
students encountered many obstacles that had to be over-
come. How the students hurdled these barriers is revealing.
Some of their solutions follow:


FEED -1 D-1 I
TEMP. 6008C
PRE .. i BR I
wc J Rnn NIR

Figure 2. Ammonia (caricature) process flowsheet
Chemical Engineering Education

Event 1. Most design groups established a computer simu-
lation as a first step. They studied the effects of changing
recycle ratio, pressure, and temperature on the conversion. It
was only after many hours (days) of toil that they considered
economics and found that the system was an economic di-
saster. Unlike those groups, one group chose to consider
economics first. They contrived what they labeled a "magic
black box" containing a chemical process to produce ammo-
nia. This box cost nothing to build and operate, it converted
100% of the limiting reactant to product, and all the energy
released in the reaction was converted to steam that was sold
at the highest value provided in the problem statement. This
was the best conceivable case that satisfied a material bal-
ance; the simple analysis took little time and required no
simulation. This group came directly to the conclusion that
the system was an economic disaster.
The other groups did not focus on the goal and developed
information on the performance of a system that could not
possibly achieve the goal. They eventually came to the same
conclusion, but they wasted significant time in the mean-
time. The professor had advised them to focus continuously
on the goal (economics) and not to set it aside for later
consideration, but the advice had little impact. When the one
group presented their "little black box" concept to the class,
the other groups realized (without being told or criticized)
how they had wasted time by not focusing on the goal.
Event 2. The design problem, as originally stated, had no
payback. The students were shocked! (This was probably
their first experience with a totally unexpected result.) At
this point the design objective was modified, with the new
objective of minimizing losses. Students were also encour-
aged to suggest the maximum price that could be paid for
feed material in order to get a positive payback and to
consider any recommendation that would make the process
profitable. The groups considered several alternatives to re-
duce the feed costs:
Paying only for the reactants and not for the inert
CO2 portion of the feed




A --e T O Mf iILi PD l
M MT I...N.


Figure 3. Improved (student composite) ammonia flow sheet
Winter 1994


Removing the C02from the feed and selling it as a
A combination of the above
The "magic black box" analysis showed that there was a
possibility of a five-year payback period if high yields and
low capital costs could be achieved. Among the alternatives
considered for making the process profitable were:
Separating the CO2, H2, and N2 from the purge
stream and selling them all in pure state
Reacting the NH3 and CO2 in the product gas to
produce urea
The students found references showing that the composi-
tion of the product gas, the temperature, and the pressure are
all appropriate for feeding directly to the reactor unit of a
urea plant. Applying the "little black box" showed that there
was an opportunity to obtain a reasonable payback period. It
was now time to consider details on how to reduce the
capital costs and increase the yield.

Event 3. Many groups recommended the removal of CO2
from the feed, and the most common reason given for doing
this was to reduce the amount of CO2 in the product ammo-
nia. Several groups concluded that the reduction in equip-
ment and operating costs would be significant. One group
found their way into the library and came up with flow
sheets for ammonia plants showing a water scrubber that
removed the CO2 (they also found design information on
the scrubber, but did not understand how to use it). Another
group found out that CO2 could not be tolerated in the
reactor because it poisoned the catalyst-this uncovered
the fact that no design would work unless the CO2 was
completely removed before the reactor or unless a new
catalyst was found. The critical importance of catalyst be-
havior was identified.
It became evident to the students that their design was at
risk until they understood more about reactor design and
separations. The direction the students took at this point
depended on their assumptions, but whatever direction they
took, the merit of their design depended upon the
validity of the assumptions they made. Until the as-
sumptions were justified, their solution offered a
higher-than-necessary risk of failure. To reduce the
risk, students saw the need for future course content
covering reactor design, separations, and the proper-
ties of the catalyst.

Event 4. The ammonia is removed in a partial
condenser. This presented a major problem for the
students. It required a condensable material to be
removed from a non-condensing gas-a situation for
which calculation of the heat transfer rate involves
simultaneous heat and mass transfer considerations.
Initially, the students thought they could not deal with
this complex system because it was not included any-

where in the curriculum. Once they recognized the problem
and brought it to the attention of the faculty, however, it
was added to the concurrent course material. The cover-
age of radiation and numerical solutions to unsteady state
heat transfer had to be reduced, but this was not judged to
be a serious problem.
This last event demonstrates that when a major prob-
lem arises, it can be taken care of by the faculty and the
necessary principles can be added to the curriculum. For
lesser problems which are brought to the faculty's attention,
the students are referred to other sources of information
(such as the seniors). Many smaller problems, however, are
never even brought to the faculty's attention and are simply
solved by the students on their own-obviously, a very
satisfying development.

The design sequence represents a framework for the cur-
riculum. Students need to know where their education is
taking them and how the materials they study fit in-seldom
do they simply accept the professorial assertion that, "Trust
me, it is critical for you to know this material." After discov-
ering the importance of course material on their own, stu-
dents usually pursue an understanding of all course materials
more aggressively. In other words, students who know where
they are going are more likely to get there!
The goal of this comprehensive sequence is made clear to
the students. They will be expected to learn how to analyze,
design, and handle comprehensive chemical processes on
their own; they are told that problems in their senior year and
beyond, throughout their careers, will require this ability.
They observe the effort the seniors put into these problems
and understand that they, too, will at some point be present-
ing and defending their solutions before a panel of profes-
sors or supervisors.
Each semester begins with all the ideas from all the groups
on the table. Then each individual group has to analyze this
new information, rejecting the poor ideas and selecting those
ideas that are worthy of follow-up. The teams listen to all of
the other presentations, acquiring a good understanding of a
variety of feasible solutions to their problem.
In our department, the program's effectiveness is enhanced
by the department "culture." Graduates and seniors, excited
about what they are doing, pass their enthusiasm on to the
other students. Also, we find that motivation is not a prob-
lem if students can observe their progress toward a final
goal. In a sense, our department may be viewed as a one-
room schoolhouse; faculty offices, the classrooms, and the
undergraduate computer room/work area/lounge are all on
the same floor. In this close-knit setting, students can see
that they are doing things that they could not do the previous
semester, and that seniors are more capable than juniors and
juniors are more capable than sophomores. The culture pro-

vides beneficial peer pressure as well as a student network
where upper-level students support lower-level students.
Although the culture described above has existed in our
department for many years, it is not a prerequisite to suc-
cessful use of such coordinated design projects. If the projects
are viewed by faculty and students as a coordinated frame-
work for the curriculum, features of the culture will sponta-
neously emerge. A few faculty may become champions, but
all will actively participate. Sophomores will seek assistance
from juniors, who in turn learn from seniors.
It is not necessary to develop design projects from scratch.
We seek new project ideas from colleagues in other depart-
ments as well as from those in industry and government. We
recently got an idea from a departmental seminar speaker,
and we have also obtained ideas from former students. Over
the years we have developed several successful projects for
the sophomore-junior sequence, and we would be happy to
share them with others.
Questions often arise regarding the opportunity for
students to free load and be carried along by the group. This
has not been a problem-probably because of peer pressure
and knowing that the time will come when each of them
must appear alone to answer questions from a faculty or
employer panel.
Introduction of the design sequence has not reduced the
amount of course content provided. While our students still
have ample opportunity to solve differential equations and
integrate the Navier-Stokes equations, some changes in sub-
ject material were made. The major impact on course con-
tent was made previously when we decided to incorporate
modest design activity in each course. The subsequent inte-
gration of the design into a single problem for each semester
and to retain a single theme for the problems over four
semesters was expected to reduce the time taken away from
formal course work. The amount of time the students elected
to devote to the design has increased and exceeded our
expectations. There is some concern about the excessive
time many students spend on these problems.

The integrated design process involves features that are
largely overlooked (or dismissed), but which can signifi-
cantly impact a students' overall development. Some of these
features are
Focus Students must focus on a goal. Once they under-
stand what they are expected to accomplish, they can appre-
ciate the process used to achieve the goal, the relevance of
the course work, and the progress they have made toward
that goal each semester.
Group Projects Students benefit from group participa-
tion in the design sequence in several ways. Groups are more
willing than individuals to express new ideas and to make
Chemical Engineering Education

judgments. They find comfort in numbers and are less fear-
ful of looking foolish. While individual students are unlikely
to have sufficient experience to manage a design problem,
the situation is different for several students working to-
gether. The sum is truly greater than the parts.
Culture (One-Room Schoolhouse) Students benefit a
great deal from observing the performance of other students
at the same level as well as in the levels above and below
theirs. It is obvious to them that the higher-level students can
do what they cannot do, and that they can now do what they
could not do earlier. Upper-class students can be a great help
to other students in their courses, telling them what to expect
in future courses and why they need to know the subjects
they are studying.
Need-to-Know Students are more motivated to under-
stand and to retain knowledge and principles when their
studies are the result of a sequence of events that begins with
a need-to-know. The steps following the need-to-know are
to gather information, learn the necessary principles, and
apply principles to an original problem.
Depth and Breadth It is essential that students be ex-
posed to a wide range of knowledge. It is also essential for
students to pursue some knowledge in depth and to under-
stand how it is applied to the development of chemical
All the items listed above have a common element: they
require the student to take an active, rather than a passive,
role in learning. Their design skills (not the subject of this
paper) are significantly advanced by this process. Students'
success in this program provides them with additional at-
tributes resulting from focused participation in activities in-
variant with time-attributes that will serve them well
throughout their professional careers.

1. Shaeiwitz, J.A., and R.C. Bailie, "Incorporating Design into
the Sophomore and Junior Years," Proceedings of 1992 ASEE
Conference, p. 1266
2. Turton, R., and R.C. Bailie, "Chemical Engineering Design:
Problem-Solving Strategy," Chem. Eng. Ed., 26, 44 (1992)
3. Gardner, A.A., P.H. Whiting, and A.F. Galli, "From Raw
Materials to Profit: Career Role-Playing in a Senior Design
Project," Paper #74c, presented at Annual AIChE Meeting,
Los Angeles, CA (1982) 0

REVIEW: Unit Operations Handbook
Continued from page 43.
lished in Encyclopedia of Chemical Processing and Design,
edited by McKetta and Cunningham, first printed in 1976
and reprinted at one- to two-year intervals through 1990.
Many of the sections do not appear to have been updated
since their first appearance. The most recent reference found
by this reviewer in any section was from 1986; most of the
Winter 1994

sections cite nothing more recent than the 1970s. The section
on batch distillation cites only a single reference that was
published in 1958, while the section on packed towers short-
cuts cites only material from two chapters in the 4th edition
of Perry's Handbook, which appeared in 1963.
As a result, many of the sections are seriously out of date.
For instance, the section on absorption presents an overly
long, highly empirical example for calculating steam
stripping that makes no use of computer techniques. The
section on packed column internals (as well as the intro-
ductory section on distillation) contains nothing about
packing that have been introduced in the past decade. The
section on the costs associated with gas adsorption cites the
price of activated carbon that prevailed in 1977, which may
or may not correspond with current costs when the M & S
cost index is employed.
As is to be expected in a multi-authored handbook, the
sections are uneven in quality. Many (but by no means all) of
the sections are highly tilted toward petroleum processing.
The section on estimating naphtha cuts in distillation, for
instance, uses so much oil-company jargon that it is almost
unintelligible to someone who wasn't working in that area in
the 1960s. A diagram of the VLE data for methanol/water
shows a non-existent tangent pinch, and absorption is de-
scribed as a purely physical phenomenon (despite the use of
the alkanol amines to remove acid gases). To illustrate the
separation of azeotrope-forming compounds by distillation,
using benzene to break the ethanol/water azeotrope, a four-
column sequence is presented even though the use of three
columns is more common and two columns can do the job.
Some of the examples used to illustrate principles are
curious: The case of an absorber with a pinch at the bottom
tray is illustrated with a column in which an insufficient
stream of pure water is used so that only a specified fraction
of the SO2 is removed from a flue gas. No mention is made
of the improbability of a) using water, with its low capacity
for SO2, as the absorbent, or b) not using a stream of absor-
bent that is sufficiently large to shift the pinch point to the
top tray when designing a scrubber to remove SO2.
In a similar exercise, natural gas is used to strip H2S from
crude oil in a process deemed advantageous because it would
be once-through for both the gas and liquid streams. No
mention is made of the fact that the H2S will subsequently
have to be recovered before the natural gas can be used for
any other purpose, and that this requirement may make its
use as the stripping medium somewhat less attractive.
Of course, much of the basic material on the key
unit operations (adsorption, distillation, liquid-liquid extrac-
tion, crystallization, etc.) is timeless and can bear retelling
by a master in the field. Nevertheless, one may question
the value of a handbook of this type in which many of
the sections are one to two decades out of date in the first
year of its publication. 0

M 1 classroom
-- >______________



University ofBritish Columbia
Vancouver, B.C., Canada V6T 1Z4

In the first place, all men agree with the familiar maxim,
"If you don't have a thing, simulate it."
Desiderius Erasmus Il

his early reference to simulation points up a certain
ambiguity in our current use of the word. On one
hand, a simulation (especially a computer simula-
tion) tends to be regarded as a representation or prediction of
the behavior of a real thing to a considerable degree of
accuracy. On the other hand, the word can also mean making
a pretense of -so a simulation could be a counterfeit or
sham object. Error bars are usually lacking in the results of a
simulation for process design, and this may fool us into
believing that our simulation is of the first kind when it is
actually of the second kind.
To illustrate this, consider a simple separation process
based on an example in Douglas' Conceptual Design of
Chemical Processes.21 In this process, acetone is to be re-
moved from an air stream (containing -1.5 mol % acetone)
by absorption at atmospheric pressure in water at 25C. The
aqueous solution is subsequently distilled to recover the
acetone. Design specifications call for 99.5% of the incom-
ing acetone to be removed in the absorber, and 99.5% of the
acetone in the resulting feed to the distillation column to be
recovered in the top product at a composition of 99% ac-
etone. The entering air stream flows are 687 lb mol/hr air (G)
and 10.3 mol/hr acetone.
A pencil-and-paper calculation, using the absorption fac-
tor (L/mG) as a parameter, shows that when
L -1
= 1.4
the required water flow (L) is 1943 lb mol/hr and, from the
Kremser equation, 12.1 ideal stages are required in the ab-
sorber.121 But a computer calculation by the Chemos model-
ing program (developed at the University of British Colum-
bia) shows that the desired acetone removal cannot be ob-
@ Copyright ChE Division ofASEE 1994

D.W. Thompson received his BSc and PhD
degrees from the University of Birmingham
(England), and worked in industry for a num-
ber of years prior to joining the chemical engi-
neering faculty at the University of British Co-
lumbia in 1967. He uses and teaches com-
puter simulation and is greatly concerned over
the possibility of a future catastrophe occur-
ring from a design error caused by uncritical
reliance on computers.

trained at the hand-calculation design point and that much
higher water flows are needed. The difference is due to
greatly differing values for the predicted equilibrium ratio,
m, in the absorption factor, where

m = _- = Io
x P
P total pressure
P' vapor pressure of pure acetone at system temperature
S1 activity coefficient of acetone in solution
While P is specified and PO can be predicted to good accu-
racy, the value of Y1 may be in some doubt.
The value used by Douglas (y, = 6.7, giving m = 2.02) is
referenced to page 14-15 of Perry's Handbook13] where it
occurs in a table used in an example calculation (Example
2), but no source is given for the table. The value used by the
Chemos program came from a binary data bank containing
parameters for the Wilson equation obtained from Hirata.[4'
But the only entry then present in the Chemos data bank for
the acetone-water system was set 504 from Hirata, which
refers to equilibrium at 150C. The default method used by
the program for extrapolating to other temperatures assumes
that the values of
(X12 -11) and (X21 -22)
in Eqs. (9) and (10) are independent of temperature and
gives a value for the activity coefficient of acetone at infinite
dilution in water at 250C to be

y = 19.61
Alternatively, the "athermal model" could be chosen, which
Chemical Engineering Education

The minimum solvent rate from the absorber is
an important parameter for design since it sets a
lower bound for the energy requirements if this
stream is to be subsequently distilled... this
may enable a go/no-go decision to be made for
this route early in the design process.

assumes that AGE/RT is independent of T and gives (see
Eq. 7)

yT (250 C) = y (1500C) = 12.64
Obviously, either of these predicted values would require
proportionately larger liquid flows to keep the absorption
factor at a reasonable design value. Since all of this extra
water has to be distilled, the size and energy requirements of
the recovery column would also be greatly increased.
Extrapolation of activity coefficient values from one tem-
perature to another always introduces extra uncertainty.[5;p'2621
It can be avoided in this case since the Hirata data collection
contains results at 250C.[4: set 503] Using the values of the
Wilson parameters calculated by Hirata for that data gives

yT = 4.69
which would reduce the water flow to 70% of the amount
originally calculated. Gmehling's data collection161 has this
same data76, set 232] and one other at 25 C.[8:;6, se 2451 Gmehling's
fit of the Wilson equation to the Beare'71 data gives

y' = 7.00
and to the Taylor'18 data gives

y7 =6.44
Since data set 503 in Hirata is the same as data set 232 in
Gmehling, the difference between the calculated y7 values
(4.69 versus 7.00) must either be due to different assump-
tions made in data reduction or be caused by different crite-
ria for goodness of fit.

Each of the referenced authors assumes that all non-ideali-
ties may be described by the liquid-phase activity coefficient
(y,) so that the phase equilibria may be represented by

xi yi pi P1

where the mole fractions in the vapor and liquid phases (y,
and x,) and the total pressure (P) are from the experimental
data, and the pure component vapor pressure (P[o) is given
by the Antoine equation

loglo Pio) = A B (2)
lOg(P)= (t+C) (2)

The coefficients (A,B,C) used by the two authors differ
slightly: Hirata, et al., calculate the values of
Winter 1994

Paocetone = 230.05 mm Hg and Poater = 23.76 mm Hg
while Gmehling, et al., use

Pacetone = 230.91 mm Hg and Pwater = 23.69 mm Hg
at 25'C. The effect of these differences is very slight, as can
be seen in Table 1.

The Wilson equation for the activity coefficient of each
component in a binary mixture can be written

fny1 = -n(xl + A12x2) x2 ( +A 2 A21 (3)
x, + Al2x2 A21X1 +X2)
en Y2 ==-n(x2 + A21x1-xI +A2x2 A21 (4)

A nonlinear regression method must be used to find the
values of Wilson's parameters (A 12, A21) that best fit a set of
experimental data. The choice of the measure of goodness of
fit (objective function to be minimized) and of the regression
method may affect the final values of the parameters.
Hirata, et al., use a nonlinear least-squares method to
minimize the function

F = (Qexp Qcalc)2 (5)

where n is the number of experimental points and
Q = x, n y, + x2 n 2 (6)
The independent variables are the values of L12 and L21.
Since the excess free energy of mixing is given by

Experimental and Calculated Activity Coefficients for
Acetone(l) in Acetone-Water system at 25C


Hirata, et al.
Y1 exp 1 talc



Gmehling, et al.
1 exp "I calc

(6.0)* (6.998)
5.854 6.284
5.753 5.978
5.621 5.517
5.434 5.241
4.857 4.426
4.852 4.404
4.111 3.826
3.890 3.610
3.393 3.238
2.545 2.518
2.021 2.041
1.360 1.381
1.109 1.105

* Obtained by linear extrapolation for the first four data points.

AGE = RT(x, n y, + X2 n 72) (7)
Hirata's solution best fits the excess free energy for an iso-
thermal data set.
Gmehling, et al., use the non-linear simplex method to
minimize the function

F I(exp-Tcalc 2 ( Yexp-Ycalc (
.+ (8)
i=l Yexp 1 Yexp )2

and they relate the values of A,2 and A21 to the interaction
energy between components i and j (Xj) by the relations

A2 L xp[- 12 V2L e Al2
L2 V I- RT J- RT
v, V L x2, 2- X221 VIL e[ A21
21 V2L RT V2L RT

where A12 and A21 serve as the independent variables for
minimization of F. The molar volume of pure liquid i is VL.
Gmehling, et al., use the volume at 250C for all data sets,
and Chemos calculates the saturated liquid volume at system
temperature and uses this to extrapolate to other tempera-
tures, assuming that j Xi remains constant.
The minimization function used by Hirata, et al., (Eq. 5)
is recommended on thermodynamic grounds by Reid,
et al.,15:p.259] but it weights the error in the logarithm of the
activity coefficient of a component by the mole fraction of
that component. Consequently, the final values of the pa-
rameters that result from minimizing this function are less
influenced by errors in the activity coefficients of compo-
nents at high dilution than are the corresponding values for
the minimization of the function used by Gmehling (Eq. 8).
The predicted value of the activity coefficient of acetone is
less than the value obtained from the Beare data set over
almost all of the composition range when Hirata's values of
Wilson's parameters are used (see Table 1), whereas
Gmehling's values exceed the experimental ones at the low-
est concentrations.
If Wilson's equation fit the data perfectly, then the two
minimization functions would give the same value of y7.
The observed difference could be because the equation is
inappropriate, or the data is erroneous, or both are wrong.
Leaving aside the choice of equation for the moment, two
tests on thermodynamic consistency of the data are reported
in Gmehling. The "area" testl91 evaluates

1in 71)dx

which should equal zero for isothermal data. The "point"
test101 calculates y, from experimental T, P, and x, for each
point, and evaluates

y y (calc) y I(exp)
This test requires values of P / ax, which are obtained by
fitting a smooth spline curve to the P-x, data. The Beare data
set meets the criteria set by Gmehling for the area test but
fails the point test, while the Taylor data fails both tests. The
experimental acetone activity coefficient has an s-shaped
curve with mole fraction instead of increasing with increas-
ing negative slope as x, -> 0, and the experimental water
activity coefficient has a minimum of 0.916 at a mole frac-
tion of acetone of 0.147 instead of increasing monotonically
from 1.0 at x, = 0. These differences may be the result of
systematic experimental errors. In any event, they make
prediction of the activity coefficient very uncertain at low
concentrations of acetone (x, < 0.15).

Prediction methods based on ASOG or UNIFAC can be
used for this system. The ASOG method described by
Pierottil[' is summarized in Table 8-17 of Reid, et al.,[4] and
predicts y7 = 7.78 at 25C. While prediction methods can-
not be better than the data they are based on, Pierotti had
access to Shell Oil Company data book values that may not
be generally available. Using UNIFAC by running Aspen
Plus"' predicts values of the Wilson equation parameters
that give y7 = 11.49.

Table 2 lists values of y7 from various sources. Any one
of these values (except perhaps the two highest) might be
chosen as a basis for design in the absence of other informa-
tion, but the most likely values to be selected are either of the
two values from data compilations 4.69't4 and 7.00161, or the
Aspen/Unifac value of 11.49.

Value of y7 for the Acetone (1)-Water System
at 250C (Various Sources)
yj Source
4.69 Hirata, et al., 1975141 (Beare data)
6.44 Gmehling, et al., 1977'61 (Taylor data)
6.70 Perry, etal., 1973'1 (Example 14-15)
6.93 Perry, etal., (1984)"3' (Example 14-21)
7.00 Gmehling, et al., 1977'6] (Beare data)
7.78 Reid,etal., 198715 (Table 8-17)
8.96 Perry, etal., 1963[14] (Examples 14-31,14-32)*
11.49 Aspen/Unifac, 1991
12.64 Chemos athermal extrapolated from 150C
19.61 Chemos default extrapolated from 150C
y-value back-calculated from value of m used in examples

Chemical Engineering Education

Table 3 lists the results of some Aspen simulations taking
the same feed-gas flows and product specifications as were
used in the hand-calculation example.21' The absorber was
represented by the ABSBR block (the RADFRAC block
gave almost identical results), the equation-of-state option
set was SYSOP8A (Wilson/Redlich-Kwong/Henry's Law),
and N2 (used instead of air as the non-absorbed gas compo-
nent) was declared an unsymmetrical (Henry's Law) compo-
nent. A specification statement was used to adjust the liquid
flow rate to satisfy the design requirement on the acetone
leaving in the gas stream. Since the entering gas contained
no water vapor, the simulation showed some evaporation of
the water (-23 lb mol/hr). This evaporation required more
heat than was provided by the absorption of acetone so that
in all cases the water leaving the bottom of the tower was
cooler than the entering liquid-by 13F at the lowest flow
rates and by 3F at the highest.
It was not possible to obtain convergence of the simulation
model at absorption factors as low as 1.4. We used all three
data sources to model a six-stage absorber which gave a
liquid rate ranging from 1460 lb mol/hr for the lowest activ-
ity coefficient (Beare/Hirata) to 5030 lb mol/hr for the high-
est (Unifac/Aspen) (see Table 3). In a second series of tests,
we increased the number of stages to the maximum that
could be solved (minimum liquid rate that would converge).
The solutions varied from six stages at 1460 lb mol/hr liquid
(Beare/Hirata data) to ten stages at 3538 lb mol/hr (Unifac/
Aspen). Even this considerable range may underestimate
the extent of our ignorance since other equations (Van
Laar, Margules, NRTL, Uniquac) could have been fitted to
the equilibrium data. Also, we made no attempt to convert
ideal stages to real packing-this step is commonly re-
garded as being much less accurate than the prediction of
equilibrium relationships.
The minimum solvent rate from the absorber is an impor-
tant parameter for design since it sets a lower bound for the

Simulation of Absorber for Acetone Using Aspen
Inlet gas flow rate: N2 = 687 lb mol/hr
Acetone = 10.3 lb mol/hr
Design specification: 99.5% of acetone to be absorbed
Inlet temperature (both streams): 77F

Data Source Stages Water Flow A=L/KG K(acetone)
lb mol/hr (geometric mean)
1. Beare/Hirata 6 1460 2.16 1.041 0.782
2a. Beare/Gmehling 6 1738 2.16 1.212- 0.943
2b. Beare/Gmehling 8 1360 1.72 1.208 0.875
3a. Unifac/Aspen 6 5030 2.18 3.305 2.984
3b. Unifac/Aspen 10 3538 1.524 3.305 2.858

energy requirements if this stream is to be subsequently
distilled. Depending on energy costs, this may enable
a go/no-go decision to be made for this route early in
the design process. (Douglas121 considers a number of other
process alternatives.) If the go/no-go decision turns out
to be dependent on y-, then clearly it becomes essential
to get better data.
If the range of y- values all lead to feasible absorber-
stripper designs, and the designs are optimized, there is still
no escaping the necessity to circulate more water when
higher values of y- are assumed. If the process were de-
signed and constructed based on the Hirata databook value,
and the real value turned out to be that predicted by Unifac,
then the process would fail to meet the required air-pollution
discharge standards. The chemical engineer who replaced
the original designer might spend some time trying to im-
prove the packing or tray efficiency, to no avail. If the
apparently conservative Unifac value was used and the Hirata
value was correct, the process would work-but it would be
more costly than it should be. A contractor proposing this
design might find his bid rejected as being non-competitive.

If the absorbed acetone is to be recovered by distillation,
equilibrium data at constant pressure (760 mmHg) are re-
quired; it is much easier to make equilibrium measurements
under these conditions, and many more data sets are avail-
able. Whereas Gmehling, et al., only list two data sets for
acetone-water at 250C, they report ten sets for the same
system at atmospheric pressure. Table 4 lists the y- values
(predicted from the Wilson equation parameters at 100C)
for these data sets, arranged by year of publication. For four
of these sets, the y1 values are also available from data
compilations of Hirata, et al.,141 and of Oh6,1E51 who used the
same measure of fit as Hirata. In some cases the different
fitting methods have resulted in greatly different values of
y7 (34.83 compared to 11.81 for the 1952 data set, for
example), even though that data satisfies both measures of
thermodynamic consistency.
The more extensive data allow different activity coeffi-
cient equations to be compared. Hirata only fits the Wilson
equation, but Gmehling fits each data set with five equations
(Margules, Van Laar, Wilson, NRTL, and Uniquac) and
compares their goodness of fit by evaluating

Syilexp Ylcalc
Table 4 lists the equations that are reported to best fit the
data and the corresponding y7 predicted by that equation.
No clear picture emerges, since NRTL is best for five sets,
Continued on page 67.

Winter 1994

MR% classroom
^_________________. ___



University ofPittsburgh
Pittsburgh, PA 15261

O nly a foolish person ignores a warning. Julius Cae-
sar heard the words of this warning in one of
Shakespeare's great tragedies: 1
Beware the ides of March
but upon hearing them, he simply inquired as to who uttered
them. Brutus (who was one of the assasins) replied that they
were said by a soothsayer. History tells us that Caesar did
not heed the warning and reveals the consequences of his
In contemporary life, we find the following warning on
cigarette packs:

Smoking causes lung cancer, heart disease,
emphysema, and may complicate pregnancy

As I still see people smoking, it is apparent that they, like
Caesar, choose to ignore warnings.
The purpose of this paper is to direct your attention
to another warning: when too much attention is placed on
the teaching of ideal gases, troubles can result. I hope
to convince you that neglecting this warning can also be
I fully expect that you will introduce the equation of state
of an ideal gas to your students during their chemical engi-
neering education. But I encourage you to limit the discus-
sion to a much greater extent than you may have practiced in
the past. My reasons for this cautionary note are several.
I will begin my discussion by considering just what in-
fluence I may have on my students after they graduate
and leave the university setting. Since I will not always be
at their elbow, I hope I will have taught them how to
cope without faculty assistance. I want them to be able to
view subjects from a general perspective and to decide

Alan J. Brainard began his formal education at
Fenn College (now Cleveland State University),
did his graduate work at the University of Michi-
gan, and graduated with his PhD the last year the
Cleveland Browns won the National Football
League title. He married his wife, Judy, that same
S year, and they have two sons, John and Paul.
His professional interests are in thermodynam-
ics, coal conversion and utilization, and creativ-
ity. He is currently working on a text in chemical

just what skills they have that may be applicable to the
problem at hand.
One of the educational experiences the students will have
had is an introduction to the equations of state and their use,
both in the prediction of the pressure, volume, and tempera-
ture (PVT) properties of fluids and in the determination of
the values of other thermodynamic properties.While I (and
others who are conversant with the subject matter) know just
when the predictions of the ideal gas equation of state can be
used to estimate the properties of fluids, my experience has
been that students have not had sufficient exposure to the
subject to make such a judgment.
Because of the difficulties I have encountered in trying
to dissuade students from their near devotion to the use of
an ideal gas, in this paper I will limit my comments to
ideal gases. It has been my experience that once students
have been exposed to ideal gases, finding some way to
get them to unlearn this equation of state is more difficult
than was teaching them a more realistic equation of state
in the first place. I will begin by first introducing the com-
pressibility factor and then using it to define an equation of
state for a real fluid.

I will first consider systems whose variables of state, pres-
sure, volume, and temperature are sufficient to set the equi-
librium state. Since these variables occur frequently, I will
use the symbols PVT when referring to them throughout this
paper. This limits our considerations to gaseous systems

Copyright ChE Division ofASEE 1994

Chemical Engineering Education

whose only work mode is P-V work.
The defining equation of the compressibility factor of any
real gas can be written as

RT -(1)
R universal constant known as the gas constant
V molar volume
z the compressibility factor

R is not dimensionless but has units which depend upon the
choices of units for the variables P, V, n, and T. This can be
seen most clearly by inspection of Eq. (2):

R nTPV (2)

While Eq. (1) serves as the defining equation for the
compressibility factor, it also represents the equation of state
for any real gas. I feel that students should be introduced to
the general subject of equations of state for gases by means
of the compressibility factor and Eq. (1).
The compressibility factor is important for at least two
1. It is dimensionless. Although that point may seem
simplistic and not worthy of note, I believe it is a mistake
to overlook its importance since dimensionless quantities
are used a great deal in chemical engineering, both in
empirical representations of experimental data and in
theoretical considerations. This fact should be pointed
out to the students-the compressibility factor may be
the first dimensionless variable they encounter in their
2. The variable, z, is subject to direct experimental
measurement. I stress this point since I believe that
measurement is a topic essential for all engineering. I
point out that a series of experimental measurements of
the PVT behavior of a known quantity of gas can be
made, and that we can thus determine a numerical value
for z directly from its definition given by Eq. (1). Once
the z values have been determined, they can be presented
as tables or plots.

The factor z is a variable when we obtain values of it for
real gases by direct experimental measurement. This means
that z is not something which has a constant value.
When we consider gases which follow the ideal gas equa-
tion of state, we see that the compressibility factor is a
constant identically equal to unity. Unfortunately, no known
gas follows the ideal gas equation of state over all of its
equilibrium states. Its very name should make it completely
clear that it is an idealization, and thus an ideal gas serves as
a limit for the behavior of real gases under certain special-
ized conditions. It is imperative that students make note of
the above italicized words. Real gases follow ideal gas be-
Winter 1994

havior only as limiting cases.
An ideal gas is any substance whose pressure, volume, and
temperature behavior can be represented by
PV = nRT (3)
where R is the gas constant (the other variables were defined
in Eq. 2).
The ideal gas "law" is given considerable attention in
many texts and by many of my colleagues in teaching chem-
istry, physics, and chemical engineering, but I do not join
their ranks. The time I spend in teaching an ideal gas is very
limited. While I can understand the importance of an ideal
gas in the historical development of an equation of state, I
feel that too much attention is given to it in most instances.
What is particularly troubling to me is that I have seen far
too much evidence that all this attention does not lead to
student understanding of equations of state or their applica-
tion. Rather, it has been my experience that students use an
ideal gas under all circumstances-whether or not it is appli-
My thoughts concerning an ideal gas are summarized in
the following paragraphs.

Please note that I do not refer to it as a "law." A law
is something which should have near-universal validity,
and the equation of state for an ideal gas does not meet
that criteria.

It is a qualitative equation of state, at best, and its quanti-
tative predictions can be very far from reality.

If we divide both sides of Eq. (3) by nRT, we find
PV 1 (4)
nRT -
Comparison of Eq. (4) to Eq. (1), which is the defining
equation for the compressibility factor, indicates that the
ideal gas equation of state predicts that z is a constant identi-
cally equal to 1. But inspection of Figures 1 and 2 (next
page) reveals that this is not true.
While the value of the compressibility factor depends on
the magnitude of the reduced temperature and the reduced
pressure, it has a value of unity only in the following excep-
tional cases:

All isotherms in the limit as P -> 0
Isotherms with T, values which are ~<2.5. The compress-
ibility factor is unity twice for these isotherms: once at the
zero pressure limit and once at the point where it crosses the
line z=1.

I do not believe that these cases are sufficient to allow us to
use the ideal gas equation of state to predict the phase behav-
ior of most gaseous materials.
I can almost see a glow in the eyes of some of you upon
reading the words above, and I can hear your rebuttal:
"What you say is true, but there are an infinity of points on
a straight line, and thus the ideal gas must predict the correct
value of z for an infinity of
equilibrium states. Aren't you
satisfied with something which .1 o --o -
works correctly infinitely of- 1
ten?" I I I

This is music to my ears,
and I respond:
"I was hoping you would
ask that,"
... and I explain why...
There are an infinite num-
ber of points on a line, and if
we use a notation introduced
by Georg Cantor, we can say
that this infinity is represented
by a Hebrew character, aleph
1. Infinities do not stop with
aleph 1, however, and the in-
finity of possible geometric
curves known as aleph 2
(which represents the iso-
therms for real gases) is of a
greater infinity than aleph 1.
Aleph 2 is infinitely larger than
aleph 1. Thus, I can say that
ideal gas behavior is followed
in only a miniscule number of
cases. It is too simplistic to
represent the PVT behavior of
real gases. More information
concerning Cantor's classifica-
tion of infinities can be found
in references 2-4.
The following example
serves to illustrate just how
much error can be introduced
into predictions made by as-
suming that gases are ideal
rather than real.

Propane is stored at 1000 psia
and 300'F in an insulated ves-

sel. Use the generalized compressibility chart to determine V
under these conditions.
Calculating values for the reduced pressure and the reduced
temperature, I read a value of z = 0.55 from Figure 1, the
generalized compressibility chart. This allows me to calcu-
late V

Reduced pressure P.

Figure 1. Generalized compressibility factor for the low pressure range.

Pr Reduced Pressure
Figure 2. Generalized compressibilityfactorfor the high pressure range.

Chemical Engineering Education

(0.55)(10.73)(760) = 4.52(ft3/ lb mole)

This result can be compared with the value of 8.22 ft3/
lb-mole which is the value predicted using an ideal gas.
Thus, the ideal gas equation of state predicts that a volume
approximately twice the actual size would be necessary to
contain the propane. I do not consider this to be a particu-
larly accurate prediction and feel you would agree with me,
especially if I asked you to pay for the difference necessary
to construct and install a storage tank which was nearly
100% oversized!
This example does not represent the worst possible case. If
I had attempted to make a prediction of the volume of a gas
present at its critical point, the ideal gas equation of state
would have failed to an even greater degree.

It fails to predict the existence of the liquid state.

The ideal gas predicts that all isotherms must have the
mathematical behavior shown in the following equation:
PV = Constant (5)
This is the equation of a rectangular hyperbola and its
predictions are known as Boyle's "law." The equation does
have some validity, and the early investigators of the critical
point noted the approach of the critical point by observing
the breakdown of this equation. I wish to note an interesting
characteristic which is relative to this discussion-the slopes
of all isotherms which satisfy Eq. (5) are everywhere nega-
tive. This is all well and good if we want the material to
satisfy a condition of stability and remain in a single phase.
Real substances do not satisfy the stability criteria for all
states, however, and other phase(s) appear. If you have any
doubt concerning this, consider that the surface of the earth
is approximately 80% water; if you have further difficulty
with it, I suggest you try to go for more than one day
consuming only gas and not any liquid. While you may
survive the experiment, I am confident you will not reject an
opportunity to partake of liquids when they are presented to
you following the exercise.

It Violates the Third Law of Thermodynamics.

This is a most serious crime. The laws of thermodynamics
are so universal and pure that it is unthinkable to consider
ways that fail to follow their dictates. If the ideal gas is
used to predict the behavior of entropy as the absolute tem-
perature approaches zero, it predicts a value of This
is certainly specious, for while there is some disagree-
ment about the numerical value of this limit, the third
law of thermodynamics informs us that it approaches a

finite limit and this limit is not -.
In one sense, it is foolish to even attempt to use the ideal
gas equation to predict any behavior as the absolute tempera-
ture approaches zero. The ideal gas equation is valid only for
substances which are at elevated temperatures, and zero is
hardly an elevated temperature.
I summarize my case thusly:

The ideal gas is too simplistic to represent the be-
havior of real gases and fluids over the vast majority
of cases of interest to chemical engineering. Thus, I
should like to have it delivered to Walt Disney World
where it can join with other fairy-tale characters and
live happily ever after. It would quickly become a
great tourist attraction and with the passage of time
could become as celebrated as Donald Duck and
Mickey Mouse.

I am aware that there are those who will attempt to defend
an ideal gas as an equation of state that provides an introduc-
tion to the general topic and serves as a starting point for any
subsequent discussion of a real gas. But I do not accept this
as a valid defense. It has been my experience that once a
student learns the equation of an ideal gas, it is only this
equation of state that is retained in his memory, and it is used
under all circumstances, whether or not it is applicable.
A prime example of this is a recent experience I had when
grading a problem I had included on a PhD qualifying ex-
amination. I had given a quantity of steam at a temperature
greater than its critical temperature and had asked the stu-
dents to determine the energy interaction necessary to bring
the steam to its critical state at constant volume. Five out of
the eight students used the ideal gas law in attempting to
solve this problem. Since an ideal gas never has a critical
temperature, it was, of course, impossible for them to locate
the critical point and thus to determine the magnitude of the
energy interaction. One student even went so far as to state
that the gas was incompressible.
A gas is incompressible? This certainly was news to me!
There are no gases which are incompressible!
I have no idea where such ideas find their origin. I expect
that it was just a reflection of the fact that the student felt a
great deal of pressure at the time of the examination, and he
was merely grasping at straws. If so, he certainly selected a
most inappropriate one.
I am very much of the opinion that it is more difficult to
convince a student to "unlearn" something than it is to teach
the correct approach in the first place. While it may take a
little longer for the student to see the general relationship
among the PVT variables when real gases are considered, the
time will be well spent. It is certainly worthwhile to invest a
little additional time in introducing gases by using the com-

Winter 1994

pressibility factor. Ideal gases simply present too many prob-
lems. The following is another case in point.

The compressibility factor can be written in the virial form
as a power series in reciprocal specific volume, and if we do
this and perform some algebraic operations on the resulting
equation, we find

RT B(T)RT C(T)RT (6)
P PV p2 (6)

where B(T) and C(T) are the second and third virial coeffi-
cients of the substance of interest.
We introduce a new symbol, a, to represent the expression
which appears on the left-hand side of this equation. We
now want to pass to the limit at P -> 0. Whenever P 0,
V V' -> 0, and PV -> RT. Thus, we can write

lim ( B(T)RT C(T)RT (7
lim a = lim P PV ... (7)
P-O p-o Pv pV2

The limit of the right-hand side reduces to -B(T). This is a
completely general result and all equations of state must
satisfy this condition.
I next ask a student to evaluate the limit of a as P -> 0. The
student reasons that real gases must approach ideal gas be-
havior as P -> 0, and he introduces the expression PV = RT
into the defining equation, only to find that it becomes
identically equal to zero! This result is in conflict with the
result given by Eq. (7).
Q. How do you explain this apparent paradox?
A. When P -4 0, aP -> 0, and PV -> RT even though
a 0. Thus, the result the student obtained resulted
from his being led into a trap. An ideal gas can do
many nasty things to you. It is best to avoid traps.
Dodge has commented on this paradox, and he concludes[51
In explanation of this we may say merely that there is no
ideal gas. It is an imaginary gas introducedfor
convenience and one must not expect an actual gas to
have the behavior postulated for an ideal gas even as P
approaches zero.
Those words were published nearly fifty years ago. I can't
help but wish that they had been read by textbook authors
and others who present an ideal gas as an acceptable equa-
tion of state. It simply isn't.
Could the student have avoided the trap? I don't know. I
recall being present at an oral exam a number of years ago
and seeing a graduate student led into the same trap. He
became unnerved when he recognized his error and was
unable to resolve the paradox at that time.
The best advice is to avoid traps whenever possible. Help
your students learn to build a better mouse trap rather than
be caught by one.

Having considered some of the difficulties resulting from
the use of an ideal gas on the macroscopic level, I would
now like to examine some of the problems encountered
when we consider an ideal gas at the microscopic level. I
think it will become quite clear that an ideal gas can lead to
some conclusions that don't make a great deal of sense.
Both real and ideal gases exert pressure on the walls of
those vessels used to contain them. This pressure certainly
results from the momentum exchange between the gas mol-
ecules and the confining walls. The kinetic theory of gases
assumes that the gas is composed of molecules which are in
motion and which experience collisions with the surround-
ing walls and with one another. There is no term in the ideal
gas equation of state which even attempts to account for the
volume occupied by the molecules themselves. (You might
note that the parameter b in the van der Waals' equation of
state does take this into account.) Thus, the molecules of an
ideal gas are considered to occupy zero volume.
If the molecules have zero volume, this introduces the
following difficulties:

The molecules exert a pressure so they must have a mass.
If their volume is zero, their density must be infinite.
The molecules have a viscosity which is proportional to
the square root of the temperature.

If the gas molecules have no volume, how is it possible for
them to have a viscosity?
I suggest that you return to the words of Dodge. An ideal
gas is an imaginary gas which exists only in textbooks. It can
lead one astray even under those conditions where it is
supposed to be valid.

Do not teach it as your introduction to
equations of state!

I expect that some discussion of an ideal gas is necessary,
but I suggest that you point out some of the limitations I
have discussed in this paper and that you make minimal
references to ideal gases.

1. Shakespeare, William, The Complete Works of William
Shakespeare, Chatam River Press, New York, 814 (1975)
2. Gamow, George, One Two Three... Infinity, a Mentor Book,
published by New American Library, New York, 25-34 (1947)
3. Asimov, Isaac, Asimov on Numbers, Pocket Books, New
York, 68-83 (1977)
4. Freund, John E., A Modern Introduction to Mathematics,
Prentice-Hall, Inc., Englewood Cliffs, NJ, 399-414 (1956)
5. Dodge, Barnett F., Chemical Engineering Thermodynamics,
McGraw-Hill Book Company, New York, 297 (1944) 0
Chemical Engineering Education


Continued from page 61.

Margules for two, Wilson for two, and Uniquac for one.
Restricting the comparison to the six sets of data that satisfy
both of the thermodynamic consistency tests gives a score of
three for NRTL, two for Wilson, and one for Unifac.
Based only on the "good" data, only on the Gmehling
measure of fit, and only on the Wilson equation, this subset
still yields values of y7 ranging from 7.41 to 11.81. If the
Hirata measure is also included, the range is from 7.41 to
34.83. For many distillation calculations the different data
sets would yield similar results, but in this example, if the
bottom product from the tower has to have a very low
acetone concentration to meet discharge standards, or to be
recycled to the absorber, then y, approaches y7 and the
scatter is important.
At first sight it appears that the more recent sets of data
show less variation, but note that three of these sets are
smoothed data, at equal x increments. Some of these may
have been obtained by interpolating smoothed bubble-point
and dew-point curves, with no simultaneous measurements
of x and y. Applying the point consistency test to these data
may be tautologous.

The error bars in chemical engineering design may be
much wider than is commonly believed. This is especially
true for non-ideal systems at limiting concentration-a re-
gion of operation approached by many absorbers designed to
meet discharge specifications for air pollution abatement.
There are warnings in the literature:
Finally, it must be emphasized that the biggest limitation to
further developing the models is not theoretical insight or
computational shortcoming, but the lack of good, relevant,
experimental phase equilibrium data.

Most chemical engineering textbooks, however, are silent
on this and treat experimental data as given, with no error
bars. To avoid being fooled, or fooling others, data sources
must be properly referenced. This may require finding
out where the numbers in a computer data bank came
from. When possible, the range of uncertainty of the input
data should be reported and the consequences of that range
carried through to the final design. At that point it may
be necessary to put the simulation aside and go back to
the laboratory.

1. Erasmus, D., Moriae Encomium or The Praise of Folly, first
printed 1511, translated by Hoyt Hopewell Hudson, Ran-
dom House (1941)
2. Douglas, J.M., Conceptual Design of Chemical Processes,
McGraw-Hill, New York, NY (1988)

Winter 1994


Value of y" for Acetone(1)-Water System at 760 mm Hg,
100C from Experimental Data

Wilson's Wilson's
Year Data Equation Fit by Equation Fit by
Published (Hirata, '75)/(Oh, '89) (Gmehling, '77) Notes
DataSet y1 Dataset Y1 S' A' P' E* 0 (Best)'

1942 251 5.546 4 N 5.39
1943 234 9.268 4 N 8.90
1945 241 9.273 M 8.64
1947 496(H) 20.100 244 6.731 N 6.56
1950 235 7.411 M 8.96
1952 497(H) 34.833 242 11.810 4 W 11.81
1953 249 6.820 '1 '1 U 7.56
1958 236 10.434 4 N 10.09
1968 495(H) 10.070 237 10.172 4 1 W 10.17
1973 510(0) 11.277 250 9.961 4 4 N 7.85

Points in experimental data set were smoothed
2 Data met "area" thermodynamic consistency test
3 Data met "point" thermodynamic consistency test
4 Equation that best fit the data, based on smallest I lYexp Ycalc I
from Gmehling: N-NRTL; M-Margules; W-Wilson; U-Uniquac
Value predicted by equation indicated in Column E, as fit by

3. Perry, R.H., and C.H. Chilton, Chemical Engineers' Hand-
book, 5th ed., McGraw-Hill, New York, NY (1973)
4. Hirata, M., S. Oh6, and K. Nagahama, Computer Aided
Data Book of Vapor-Liquid Equilibrium, Kodansha, Tokyo,
Japan (1975)
5. Reid, R.C., J.M. Prausnitz, and B.E. Poling, The Properties
of Gases and Liquids, 4th ed, McGraw-Hill, New York, NY,
6. Gmehling, J., U. Ouken, and W. Arlt, Vapor-Liquid Equi-
librium Data Collection, Vol. 1, DECHEMA, Frankfurt/
Main (1977)
7. Beare, W.G., G.R. McVicar, and J.B. Fergusson, J. Phys.
Chem., 34, 1310 (1930)
8. Taylor, J. Phys. Chem., 4, 290 (1900)
9. Redlich, 0., and A.T. Kister, Ind. Eng. Chem., 40, 345
10. Van Ness, H.C., S.M. Byer, and R.E. Gibbs, AIChE J., 19,
238 (1973)
11. Pierotti, G.J., C.H. Deal, and E.L. Derr, Ind. Eng. Chem.,
51, 95 (1959
12. Aspen Plus, Aspen Technology Inc., 251 Vassar St., Cam-
bridge, MA 02139; Version DOS-386, Release 8.4-1 (1990)
13. Perry, R.H., D.W. Green, and J.O. Maloney, Perry's Chemi-
cal Engineers' Handbook, 6th ed., McGraw-Hill, New York,
NY (1984)
14. Perry, R.H., C.H. Chilton, and S.D. Kirkpatrick, Chemical
Engineers' Handbook, 4th ed., McGraw-Hill, New York,
NY (1963)
15. Oh6, S., Vapor-Liquid Equilibria Data: Physical Sciences
Data, 37, Kodansha, Tokyo, Japan (1989)
16. Fredenslund, A., Fluid Phase Equilib., 52, 135 (1989) 0

Me classroom
---- ----- s.___________________________________Io_



University ofNew Brunswick
Fredericton, N.B., Canada E3B 5A3

S tepping from the industrial world into the academic
world can be a daunting, but challenging, experi-
ence. After almost twenty years in industry, includ-
ing some ten years in project management, I returned to
academia and found myself facing the question of how I
could best apply my industrial experience to at least partially
compensate for my lack of teaching experience. I felt that
several aspects needed to be addressed: first, I needed an
overview of the position, including all of its duties and
responsibilities; second, I needed to understand the role that
lectures and exams play, and their interrelationship, in addi-
tion to an understanding of the grading system; and third,
consideration had to be accorded the students and their abili-
ties and aspirations. I concluded that some of the fundamen-
tal philosophies of project management that I had practiced
in industry could be applied to teaching, and that a number
of analogues existed between the two professions.

In constructional project management, the role of the project
manager is to build something-anything from a house, say,
to a nuclear power plant. Overseeing the construction in-
volves meeting deadlines and specifications. In most cases,
however, both time and specifications are subject to change,
and this in turn affects costs-so the project manager often
has to strike the best balance he can between time, specifica-
tions, and costs. In addition, he stands between the contrac-
tor and the client, and, whether or not he is employed by
either of them, he must satisfy both of them. Contractual

Robin A. Chaplin received his BSc (1965) and
MSc (1968) from the University of Cape Town,
another MSc from the University of London
(1972), and a PhD from Queen's University
(1986), all in engineering. He worked fora power
company prior to obtaining his PhD and has
since been teaching in the field of Power Plant
Copyright ChE Division ofASEE 1994

obligations must be met and personal differences must be
overcome if they exist.
The professor is similarly positioned between the require-
ments of the university and the needs of the students. The
university requires a certain academic standard (the specifi-
cation) in upholding the reputation of the institution, while
the students seek an education, within a certain time frame,
that will qualify them for a professional position in society.
Cost is also a factor, just as it is in project management.
Additional fees are incurred when a student, who does not
meet the academic specification, fails and graduation (and
earning power) is delayed. It is idealistic to hope that all
students will pass, but students who fail should serve as an
incentive for reviewing and improving teaching standards.
Another aspect of cost is the value of the service provided.
While a project manager is well aware of the costs involved
when a contractor stands idle for an hour, is the professor
fully aware of the costs involved when he cancels a class? A
significant number of dollars can be involved if you consider
his salary, his time commitment to a particular course, the
students' fees for the course, and the number of students
enrolled in the class (not to mention government subsidies).
Professors should feel an obligation to make the best use of
the allocated time and to provide a service that is commensu-
rate with the overall cost.

Lectures have been defined as the transfer of knowledge
from the professor's notebook to the students' notebooks
without that knowledge having passed through the mind of
either of them. If this is true, why not just hand out copies of
the notes at the beginning of the term? If we presume that
students absorb only a small amount of knowledge through
the writing process, then we must assume that the bulk of
their knowledge is obtained outside the classroom. Consider
the opposite extreme-a highly gifted lecturer who can hold
the class spellbound for an hour at a time. Can we assume
that those students will automatically achieve top grades? Of
course not. There is obviously still a need for acquisition of
knowledge outside even the best lectures. An integral part of

Chemical Engineering Education

the learning process is usually some form of self-study-
most people maintain that what they know best is what they
learned on their own, through their own efforts. It seems,
therefore, that the most efficient method of teaching would
be to create a self-learning environment, with the direction
and inspiration coming from the lectures.
From the point-of-view of a project manager, the end
result is what really counts-and for students the end result
(the final examination) is also what counts. The normal
progression of giving lectures and then giving a test on the
lecture material can be reversed with good results. The tech-
nical content and the required standard for the final exam-
ination can be set in advance, just as specifications for
engineering construction are pre-set. Setting such instruc-
tional objectives allows the professor to adopt a manage-
ment-by-objectives approach to teaching and gives the
students direction. Further guidance can be provided
through regular assignments that have a similar content as
the final examination, with the lectures providing guidance
on how to execute the assignments.
Students need other skills in addition to technical skills,
and some of them (particularly written communication and
mathematical computation) are demonstrated in the exami-
nations. Assignments throughout the course help to develop
these skills, and the final examination is to some extent a
measure of the development of these skills during the course.
There is an analogue with athletes or musicians who practice
avidly for the big event-students need to practice their
skills before attempting the final examination; such practice
comes through regular execution of assignments.

The obvious end result of a student's aspirations is a de-
gree. That is the goal of all students, and each course can be
viewed as one step along the way. In general, the more
demanding the steps, the more valuable the degree. The
aspirations of the students are no different from those of the
university-to achieve a high academic standard. Human
nature, however, provides some inertia to the process and
fosters a tendency to take the line of least resistance, creating
short-term goals of merely surviving from one examination
to the next. It is therefore important that the students have an
overall vision of the program.
In construction, it is obviously foolish and expensive to
build something that cannot be used, or to have to rebuild
something that was built wrongly. It is also important for the
sequence of construction to be correct and logical. The same
principle applies to a student's educational program and even
to the individual courses within that program. Expertise
should be developed in a logical and constructive manner.
Students accept hard work when they can see a tangible
return for the effort they put forth. They find a measure of
satisfaction if their examination grade reflects that effort,
Winter 1994

and, if they can find some satisfaction in each of the approxi-
mately fifty courses they take in the engineering curriculum,
they will graduate with a sense of achievement.
There is a before-and-after perception of a degree, just as
there is in any project, where the completed project never
seems quite as good as its initial visualization. Figure 1
shows a mind's-eye conception of a new house before its
construction and the actual view of that same house after
completion. It is obvious from the illustration that some time
must be allocated to reap the full benefit of good landscap-
ing. The same principle applies to a degree: often, a number
of years can pass before a satisfactory job related to an
individual's expertise is secured. It should also be accepted
that the degree may have a few poor grades in its construc-
tion, just as a house may have some bad workmanship. This
is part of the student's overall experience, and ultimately, a
good engineer is an engineer who has the foundation of a
solid basic education around which to frame a career.

To structure a course from a project-management perspec-
tive, definite goals and certain specified levels of acceptance
are required. The course should be monitored with respect to
time and performance so that the required standard is achieved
on the due date, precise specifications for the structure of the
final examination should be set in advance, and the steps
toward the final examination and final grade should be clearly
indicated. As an example, when the final examination em-
Continued on page 73.

17 /- .7\


Figure 1


W- k




Loughborough University of Technology
Loughborough, Leicestershire, England LE11 3TU

p presentations are used extensively for communica-
tion and persuasion in almost all professions. In
particular, engineers must be able to present well
since often their technical expertise alone will not get the job
done-their ideas must be "sold" through verbal persuasion
in order to be implemented. "If engineers cannot inform
others of what they have done, they might as well not have
done it.""' In the Department of Chemical Engineering at
Loughborough we recognize this fact and as a result have
increased the use of presentations in our undergraduate
courses. We also believe that presentations can be efficient
learning experiences because presenters must understand the
material they are presenting.
In the past, the method of assessing a presentation was
left to each individual staff member, and it usually con-
sisted of assigning a mark based on the assessor's im-
pression of the presentation. We felt that better guidelines
were needed since
The proportion of total marksfor presentations in our
courses is expanding.
The grader's impression of the presentation can be
subjective and we wanted to eliminate any potential for bias.
We wanted to be able to show our students the basis of the
assessment and to identify their strengths and weaknesses
for them so they could both improve and consolidate their
Students have in the past criticized presentation assess-
ments as being subjective and of variable quality, and as a

David W. Edwards has been a lecturer in
chemical engineering at Loughborough Univer-
sity since 1990. He teaches first-year mass
and energy balances, second-year laboratory,
and second- and final-year design. His research
interests are in feasibility assessment and de-
sign and are focused on inherent safety, cost
estimation, and computerized integration of the
design process.

Copyright ChE Division ofASEE 1994

possible solution to the problem they have suggested that
additional staff members also moderate and grade the pre-
sentation. But observing and assessing presentations is time-
consuming, and with the recent worsening of staff-student
ratios it is simply not possible to assign additional staff to
assess and moderate all student presentations. Therefore, in
order to eliminate inconsistencies, without increasing staff
involvement, a more formal method of assessment has been
devised and is being presented in this paper.
Assessment is only one facet of effective teaching, how-
ever. Instruction and feedback are also important, so some
suggestions relating to these components are also included
in this paper. But since a basis for assessment must be in
place before methods for instruction and feedback can be
established, this paper will concentrate on objective assess-
ment as a first step toward improving the teaching of presen-
tations. Hanzevack and McKean'" deal in greater depth with
preparing students for their presentations.

One essential and inescapable difficulty with assessing a
presentation is that it must be done in "real time" and the
assessment process itself interferes with observing the
presentation. It is possible to use video to record and replay
the presentation, thereby separating data-gathering and
assessment (and it is also a powerful tool for showing stu-
dents their mistakes), but that method conflicts with the
scarcity of time already mentioned. Also, videotaping
requires expensive and complicated equipment and extra
personnel to operate it.
Real-time assessment must be simple and should not dis-
tract from the observation. This can be achieved by using a
printed form with pre-defined headings relating to the differ-
ent aspects of a presentation. Marks and comments (for later
student feed-back) are recorded under the different headings
during the course of the presentation. The form used by the
author is shown in Figure 1.
The information at the top of the form (Order: ... of:...)
records the position in the running order and the total num-
ber of presentations. The other information blanks in the
heading are self-explanatory. The remainder of the form is
Chemical Engineering Education


divided into three sections, which are described in the fol-
lowing paragraphs.
The presenter must be satisfactory in each of these six
categories for the presentation to be a success. The grade for
each category is, therefore, a simple yes or no. For example,
the presenter is either audible or not; the visual aids are
either readable or not; etc.
Personal and affiliation details indicates the speaker's
name, department, course, etc. Most students assume that
their listeners know who they are and even what they are
going to talk about. The speaker must state these details,
however, even when talking to friends or colleagues. We are
training them for real-life presentations where the audience


Assessment is only one facet of
effective teaching, however. Instruction
and feedback are also important, so some
suggestions relating to these components
are also included in this paper.

will, in general, not be known to them.
It is also important to state the topic and aim of the
presentation; that is, what they intend to achieve by making
the presentation. Audiences need as much help as possible in
how to listen to a presentation, so clearly stating its aim is
important. Is the presentation a sales pitch, or a funny story?
The speaker must clarify the aim of the presentation. It is
generally helpful for the assessor to record the stated topic at
this point and then refer back to it at the
end of the presentation.

Occasion: Start: : Finish : Date:
Name: Mark: out of: Order: of:

KEY Any presentation must be satisfactory in these key areas
Item Y/N Comments
Readable visual aids
Stated personal and affiliation details
Stated topic and aim
Used the available time
Made the points)

Item Y/N Comments
Connected them to the topic and aim
Enjoyment 1 2 3
Understanding 1 2 3
Respect/sensitivity presenter/audience 1 2 3

Item Scale Comments
Content and relevance 1 2 3
Detail and logical structure of material 1 2 3
Use/lack of prompts, signposting 1 2 3
Quality and use of visual aids 1 2 3
Summary 1 2 3
Question handling 1 2 3

Delivery/posture/mannerisms/etc comments:

Give 3 marks for a "Yes" and 0 marks for a "No"; for the categories with a scale response, the point
on the scale is the mark; you may also give 0 marks in these categories. There are two extra marks
for general impression. The total possible is 50.
Figure 1. Presentation Assessment Form
Winter 1994

One of the most effective ways for a
presenter to alienate an audience is to
miscalculate the length of the presenta-
tion and either run under or over the
time allotted for it. In the first case the
listeners may be irritated if they allot-
ted too much time for the presentation
and could have been doing other things.
They may also perceive an unstated
message that the topic is not as im-
portant as claimed. Most presenta-
tions are intended to "sell" something
(such as a product, an idea, a design),
and during the presentation the speaker
usually has the undivided attention of
the persons) who will make the deci-
sion to "buy." Obviously, the speaker
should use the available time (but no
more) in order to make the most effec-
tive case possible.
On the other hand, when a presenta-
tion runs over the allotted time, the
speaker is probably keeping the listen-
ers from other tasks which they ex-
pected to accomplish. They are most
often distracted and annoyed by this
usurping of their time, and the impact
of the presentation is thus diluted. In
the worst case, of course, the audience
will walk out before the point of the
presentation has been made.
The assessor should record the time
taken by the presentation in order to
gauge the degree of under- or overrun.
Students often assume that doing more
than is required will result in a higher

mark, and they should be made aware of the fact that this is
not the case with presentations.
The last category, made the pointss, concerns the overall
effectiveness of the presentation. The assessor should ask
the questions: What was the main point? Would I buy it? Am
I convinced? Referring to the topic and aim that were noted
at the beginning of the presentation is helpful in determining
if the presenter accomplished those aims.

"Connected them to the topic and aim" means explain-
ing the relevance of the topic to the audience. For example,
"enzymes are important because ...." The next two catego-
ries, enjoyment and understanding are self-explanatory.
In the last category in this section the grader looks for a
mutual respect and sensitivity between the presenter and
the audience. For example, was the audience bored or talk-
ing among themselves while the speaker continued, bliss-
fully unaware? Some other things to look for and include in
this category are if the style of the presentation was suited to
the type and size of the room it was given in, and did the
presenter correctly judge the audience's previous knowledge
of the subject, altering his or her presentation accordingly?
For example, a sensitive speaker would not explain some-
thing that had already been explained by a colleague in a
session of presentations; it would be sufficient to say, "as so-
and-so has already mentioned."

The categories in this section are for grading the mechan-
ics of the presentation. Content and relevance is an assess-
ment of whether too little, sufficient, or too much material
was presented and whether or not it was pertinent to the
topic and aim of the presentation. The arrangement of the
material and the quality of the argument's development is
scored under detail and logical structure.
Most speakers need prompts to remind them of the im-
portant points they want to present. Bad presenters read the
entire presentation, putting their audience to sleep, but a
good speaker appears to know the subject well and delivers
the material in an interesting and engaging manner, using
such elements as visual aids as prompts.
Signposting indicates whether or not the speaker has ex-
plained the structure and charted the current position of the
presentation as well as where it is going. Examples of
signposting are, "I shall begin by talking about...," "Then I
will...," and "We have now reached the last section of ...."
Signposting is quite helpful for the audience members.
The quality and use of visual aids category is for
gra-ding the quality and appearance, as well as facility
of use, of the visual aids. It is different from the read-
able visual aids category in the first section which is a

simple test of readability.
The other headings in this section are self-explanatory,
and the last section is for recording general impressions and
any comments that do not fit into any of the above catego-
ries, such as excessive "uhms."

I have found that for short presentations the above head-
ings are sufficient for grading purposes. Occasionally, how-
ever, the headings could and should be expanded. For ex-
ample, if the speaker is presenting the results of experimen-
tal work, a category could be added to indicate if a diagram
was used to explain the experimental rig, or whether data
was correctly presented on graphs.

Bearing in mind that presentation assessments are made in
"real time," the scoring must be done at the time of the
presentation or the information must be noted on the check-
list so that the scoring can be done at a later date. It would be
ludicrous to be too precise. My method is to give 3 points
for a "yes" and 0 points for a "no" in the Y/N column, and a
3 for "good," a 2 for "average," and a 1 for "bad" in the
scale category. This gives a maximum of 48 points. To make
it a nice round 50 points I often add an extra 2 "discre-
tionary" points.
Some readers may feel that a scale with only three points
is too coarse, but I feel that in the majority of cases it is
difficult to be any more accurate. It is still possible, for
example, to give 0 points in a category if the presenter was
appalling and 4 points for a performance that was exception-
ally good (although this last score should be used sparingly
because it changes the total marks).
As the students progress through their courses and become
more proficient, they should satisfy the key requirements;
therefore, the weighting given to these categories should be
progressively reduced. The mark for a "yes" could be cut to
2, and then to 1, and perhaps a negative mark could be given
for a "no" if they have had sufficient training and practice to
know better. After all, experienced presenters should be
audible and should produce readable visual aids. The
number of points on the scale could also be increased to
five (0,1,2,3,4). It is not possible to score to a finer pre-
cision than five points.
In scoring the use-of-time category, marks should be sub-
tracted for serious overruns as well as underruns (unless the
presenter has given a good explanation for not using all of
the available time). Overrunning the allowed time can never
be justified.
I normally give the first presenter in the session a few
extra marks since it is the most difficult slot for both the
presenter and the assessor.

Chemical Engineering Education

It is easy to set up a session of presentations. All that is
needed is a room, some chairs, an overhead projector, and a
screen. The audience can consist of the group of presenting
students, and in that case the session can also become an
effective teaching situation since students tend to listen to
their peers, especially when the work presented is similar to
their own. The assessor should sit as far from the presenter
as possible in order to check the speaker's audibility and the
visibility of the visual aids. Keeping in mind that it is diffi-
cult for an audience to concentrate for more than an hour at a
time, breaks should be scheduled during the series of presen-
tations. We have found that eight ten-minute presentations
during a 2 2.5-hour period is a workable session.
It is helpful to spend some time preparing students before
they make their presentations. Hanzevack and McKeanm11
recommend a brief lecture in addition to written guidelines,
and they also provide a useful summary for distribution to
the students. The most important point to be stressed is that
the audience wants the presenter to do well. Most students
are very nervous about giving a presentation, but their fears
can be somewhat allayed by convincing them that the audi-
ence is on their side.
Students should also be shown the assessment form before
they give their presentation so they will be aware of the
aspects that are being observed and graded.
The best presentations are always given without notes, so
students should be encouraged from the very beginning to
present without using them. Suggest that they use their over-
heads as prompts. It is also important to stress how much
time they have for the presentation and that they will be
penalized for running beyond that limit. Since presenters
often try to include too much material, they should be made
aware of the importance of tailoring their presentation to the
available time. A good guideline to give them is, "If in
doubt, leave it out."
The audience should be encouraged to ask questions since
one of the best ways to determine if the speaker understands
the material is from the way he or she handles questions.
When the audience is made up of peer presenters, a scheme
can be devised to provide a material reward for question-
ing-that is, marks can also be given for questions. For
example, each student can be allowed a certain number of
questions, with a mark given for each of them (provided the
question is relevant and hasn't already been asked!).
Most presentations discuss work done by the student, but
the mark given for the presentation should be only for its
effectiveness. Marks for the technical content should be
assessed separately, although the impression formed from
the presentation may also have some impact on the other
mark. For example, we run second-year laboratories where
the total mark of 100 is split 30 for the experimental work,

40 for the written report, and 30 for the presentation. The
quality of the experimental work is marked in the first two of
these categories, but if the presentation indicates a lack of
understanding by the student, the mark achieved in the first
two categories can be affected.
The completed assessment form, with both the good and
bad points highlighted, should be shown to the presenter.
At the end of the session the assessor should comment on
some general points, remembering to find something posi-
tive to say about each presentation, no matter how dire it
was. Constructive criticism goes a long way in helping to
correct a presenter's deficiencies. Good presentations are
given by confident presenters, so it is beneficial to build up
the student's confidence.

1. Hanzevack, E.L., and R.A. McKean, "Teaching Effective
Oral Presentations as Part of the Senior Design Course,"
Chem. Eng. Ed., 24(1) (1990) 0

Continued from page 69.
braces problems that have a time limit for solution, the
student should have had experience (practice) with similar
problems on prior assignments and tests throughout the
course. Feedback from those problems can be quite valuable
in monitoring the student's progress through the steps lead-
ing to the final examination.
Lectures are a means of presenting basic material and
guiding students to problem solutions, and they should be
structured in that fashion. They should be well-directed to-
ward an end condition, with the professor delivering the
necessary information as efficiently as possible. The prob-
lems should be sufficiently varied to test all of the student's
problem-solving skills.
Students who are motivated to learn through good self-
study habits should be recognized for their efforts and be
awarded grades that reflect their initiative as well as their
skills. Grades should not be scaled to some inflexible scale
dictated by policy or statistics. Those students who get an
"A" should have earned it.

It is possible to meet both the requirements of a university
system and the aspirations of its students by applying con-
ventional project-management philisophies to teaching. This
method of teaching also provides appropriate direction for
the student, making his education meaningful and produc-
tive, and leaving him well-prepared for a career in industry
where project-management philosophies are applied daily.

Winter 1994




University of Dayton
Dayton, OH45469-0246

Advising undergraduate students is an important task

for engineering faculty. Not only must it be done
well, but it must also be done efficiently. Advising
first-year students is particularly challenging because at that
point the students are involved in only a few engineering
activities and have only a vague idea of what engineering
actually is. Most of their time and effort is focused on simply
surviving their mathematics, chemistry, physics, and general
education courses. Typical engineering curricula do not of-
fer the first-year student an opportunity to learn what an
engineering career really involves and whether or not it is
the correct course of study for them.
In an effort to address and overcome this problem, many
departments have added engineering courses to their first-
year curricula.1 1 Some universities have added design
courses,[2'31 while others have added introductory engineer-
ing courses.4'51 These courses typically use various combina-
tions of videos, lectures, plant trips, guest speakers, and
faculty and student presentations. Students are usually graded
on their performance on homework assignments, written and
oral reports, and examinations.
The University of Dayton previously required all first-
year engineering students to complete a two-credit hour
course, "Introduction to Engineering," and it was attended
by students from all of the engineering majors. Faculty mem-
bers from each engineering department taught sections of
the course, but this led to substantial variation in the course
emphasis and quality. Attempts were made to coordinate
and improve the course, but in 1987 it was dropped in a
move to control the rising number of credit hours required
for graduation.161 Since then some of the engineering depart-
ments at the University of Dayton have added required, no-
credit introductory seminars for their first-year students.

The seminar course "Introduction to Chemical Engineer-
ing" is offered to incoming chemical engineering students
Copyright ChE Division of ASEE 1994

and is described in the university bulletin as an "introduction
to the chemical engineering faculty, facilities, and curricu-
lum, including a survey of career opportunities in chemical
engineering." It was originally conceived through our efforts
to improve the effectiveness of advising. The program (out-
lined in Table 1) was developed to improve communication
between the students and their advisors.
The seminar provides an opportunity for the faculty to
share information with the students and offers the students
an opportunity to ask questions which they may otherwise
hesitate to ask during a one-on-one visit with their advisor. It
also attempts to involve the first-year students in departmen-
tal activities and to demonstrate that there is fun, challenge,
and reward in the work of the department and in the field of
chemical engineering.
The primary reason for teaching the course on a no-credit
basis is that most of our students would be subject to a
tuition surcharge for exceeding the university's maximum

Kevin Myers is Associate Professor in the De-
partment of Chemical and Materials Engineer-
ing at the University of Dayton. He received his
BChE from the University of Dayton and his
DSChE from Washington University in St. Louis.
His research interests are in the fields of multi-
phase agitation and chemical reactors.

Lawrence Flach, Associate Professor of Chemi-
cal Engineering at the University of Dayton, has
bachelors and masters degrees from the Univer-
sity of Cape Town, South Africa, and a PhD from
the University of Colorado, all in chemical engi-
neering. He conducts research in the areas of
process modeling, dynamics, and control.

Amy Grosjean received her BChE from the
University of Dayton in the spring of 1993. While
an undergraduate she worked at General Mo-
tors and Procter and Gamble, and during her
senior year she served as the student coordi-
nator for the first-year seminar program.

Chemical Engineering Education


Up to this point in the program the students have only been told about chemical engineering. The fifth
meeting consists of a tour of our departmental laboratories and includes demonstrations of some
of the equipment. The tour is conducted by upper-level students who describe the work they are
performing in our transport phenomena, unit operations, and process control laboratories.

Schedule of Class Meetings

Departmental welcome and faculty introduction; announcement of
student AIChE kick-off meeting
Discussion of chemical engineering, highlighting career opportuni-
ties. Video presentation
Job-experience presentations by senior students
Job-experience presentations by practicing engineers
Tour of chemical engineering laboratories
Computer/word processing tutorial
Joint meeting with junior/senior seminar
Registration advising

number of credit hours for full-time students (students are
charged a flat tuition for taking from twelve to seventeen
credit hours). Since it is a no-credit course and we are not
really interested in increasing the students' workload, grad-
ing is on a pass/fail basis. Attendance at all class meetings is
the sole criterion for determining whether a student passes or
fails the course. Any student who misses a class meeting is
required to attend a faculty-approved student AIChE meet-
ing to make up for their absence. This attendance policy is
clearly explained during the first class meeting, and since the
course does show up on the students' transcripts, very few
students miss class. During the four years that it has been
taught, no one has failed the course.
The course meets once a week for eight weeks, and covers
the topics listed in Table 1. Originally we met every two to
three weeks in an attempt to distribute the course throughout
the semester, but the result was poor attendance since many
students forgot about scheduled class meetings.
Involving the student AIChE chapter in the seminar was
designed to provide a service activity for that organization as
well as to alleviate the faculty workload associated with the
course. This involvement, coordinated by a student assistant,
has been surprisingly effective. The upper-level students are
very enthusiastic because they believe that they have impor-
tant information to share with the first-year students, and the
first-year students are often more at ease asking questions of
upper-level students than of faculty. An added bonus is that
attendance of first-year students at AIChE activities has also
increased since the introduction of this course. This is prob-
ably due to improved communication between upper-level
and first-year students.
A student-faculty social is held after one of the seminar
meetings, and the first-year students are encouraged to at-
Winter 1994

tend so that they can meet faculty and upper-level students in
an informal atmosphere.

At the first meeting, the departmental faculty is introduced
and the structure of the course is explained. The student
AIChE officers are also introduced and given the opportu-
nity to discuss the purpose and activities of AIChE as well as
to announce the first meeting of the year.
The second meeting is designed to give the first-year
students a broad overview of chemical engineering as a
profession, including its history, future prospects, skills re-
quired, industries involved, and job opportunities. This is
accomplished through a presentation by the faculty coordi-
nator, followed by the showing of an AIChE video, "Fron-
tiers in Chemical Engineering."
The third and fourth meetings expose the first-year stu-
dents to actual chemical engineering job experiences, first
by senior students and then by practicing engineers. In both
instances we attempt to cover the spectrum of chemical
engineering industries and job types. To illustrate, at the
most recent presentations, the senior students discussed work-
ing with environmental, consumer products, automotive, and
petroleum companies, while the practicing engineers dis-
cussed their work in research, technical sales, and manufac-
turing. The first-year students' enthusiasm for these presen-
tations is indicated by the large number of questions they
ask. They are particularly interested in the internship and
cooperative education programs.
Up to this point in the program the students have only been
told about chemical engineering. The fifth meeting consists
of a tour of our departmental laboratories and includes dem-
onstrations of some of the equipment. The tour is conducted
by upper-level students who describe the work they are
performing in our transport phenomena, unit operations, and
process control laboratories. For many of the first-year stu-
dents this is their first visual exposure to chemical engineer-
ing, and they are usually quite impressed by the work of the
upper-level students. Hearing an upper-level student say
things like "We're studying the recycling of plastics in this
injection molding device" or "I'm working on the redesign
and instrumentation of this heat exchanger" is probably
equivalent to an entire semester of simply discussing chemi-
cal engineering in a classroom.
The sixth meeting is used to introduce the departmental
personal computer facilities that are available to the stu-
dents. Since they are all required to write papers in their

courses, the first-year students are usually pleased to learn
that word processing facilities are available to them. Upper-
level students present a tutorial on the computers, while the
department provides each first-year student with a computer
disk. We have found that use of the computer facilities is a
convenient way to get our first-year students involved in
department activities.
At this point the first-year students attend an upper-level
seminar given by a guest speaker. This joint seminar not
only shows the first-year students the type of seminar they
will be attending in later years but it also provides additional
interaction with upper-level students.
The final meeting of the course occurs near the mid-
point of the semester and is devoted to a discussion of
midterm grade reports and registration for the next term.
Since the first-year students have never registered on
campus (their first-term registration is handled through the
mail), it is important to instruct them in the logistics of
registration and to ensure that they register for the correct
courses. After a short introduction by the faculty, this
meeting is turned over to upper-level students, thereby giv-
ing the first-year students both the faculty and student view-
points concerning registration.

During the last class meeting we ask the students to com-
plete a course evaluation that poses three questions:

1. What do you believe was the purpose of this course ?
2. Do you feel that the course accomplished this purpose?
3. What would you do to improve this course?

Most of the students correctly identify that the purpose of
the course is to initiate communication between the faculty
and the new students and to provide them with information
about chemical engineering. They also recognize that the
course offers an opportunity to get to know one another. The
students feel that the course accomplishes these goals. Typi-
cal comments include: "I feel more comfortable about what I
am trying to become. Now I know what it is."; "I found it
[the course] very helpful."; and "It didn't waste any time and
accomplished its purpose nicely." On occasion, students found
that chemical engineering was not what they were interested
in, and we were able to effectively counsel them concerning
alternate programs of study.
We were surprised when the students suggested that the
course could be improved by expanding its scope (the course
met only five times the first year it was taught). The students
requested that we add the meeting with practicing chemical
engineers, increase their interaction with upper-level stu-
dents, and provide more information concerning student chap-
ter AIChE activities.
The upper-level students have always been happy to assist

us with this course. They feel that they have useful informa-
tion and experiences to share with the first-year students.
They do a good job, although they sometimes say things that
make the faculty cringe, such as "I don't really use chemical
engineering in my job."
We have been pleasantly surprised by the success of this
course. We were worried that the students might see it as a
waste of time since it is a no-credit course that requires only
their attendance. Based on the evaluations, however, this
does not appear to be the case. The current format appears
to be optimal; there is a substantial amount of contact
time, and the first-year students are exposed to chemical
engineering in a variety of contexts. The course ends at the
middle of the term, before the students lose interest, and
allows them to subsequently focus all of their attention on
the more rigorous courses. Also, involvement of the first-
year students in departmental activities, particularly in stu-
dent AIChE activities, has increased since the course was
added to the curriculum.
Since we see all the students on a regular basis and have
an opportunity to share information with them and answer
their questions, we feel our advisory role has been improved.
We have also noticed that the students know what to expect
and are better prepared for their individual advising meet-
ings. We feel that the small amount of time and effort re-
quired to conduct this course is a good investment in our
first-year students.

Both faculty and students are pleased with this introduc-
tory seminar. It is an effective and efficient means of initiat-
ing communication between the department and its first-year
students, and the increased contact between faculty and
students enhances our advising program. The seminar also
encourages the first-year students to participate in depart-
mental activities and is a good service project for the student
AIChE chapter.

1. Landis, R.B., "National Survey on 'Introduction to Engi-
neering' Courses," California State University, Los Angeles,
CA (1992)
2. McNeill, B.W., D.L. Evans, D.H. Bowers, L. Bellamy, and
G.C. Beakley, "Beginning Design Education with Fresh-
men," Eng. Ed., 80(5), 548 (1990)
3. Olds, B.M., M.J. Pavelich, and F.R. Yeatts, "Teaching the
Design Process to Freshmen and Sophomores," Eng. Ed.,
80(5), 554 (1990)
4. Miller, W.M., and M.A. Petrich, "A Novel Freshman Class
to Introduce ChE Concepts and Opportunities, Chem. Eng.
Ed., 25(3), 134 (1991)
5. Sproull, R.D., "An Introduction to Chemical Engineering: A
First-Term Freshman Course at OSU," presented at the
1987 ASEE Summer School for Chemical Engineering Fac-
ulty, North Dartmouth, MA
6. Swaim, R.L., and P.M. Moretti, "The Case for the 120-Hour
B.S.," Eng. Ed., 81(5), 463 (1991) 0
Chemical Engineering Education

Continued from page 23.

octants, denoted by numerals I through VIII. We make the
following observations:

1. cE is usually larger in absolute value than is sE; for
the systems shown in Figure 5, I1 I > Is I about 85% of
the time.
2. Octants IV, V and VI are very sparsely occupied.
Thus the following behaviors are unusual:
sE) with cTE
sEQ with cE@ and c' <- sE
Other octants have reasonable representation.
3. c < -1 is an unusually large negative c; positive
values of c can, however, be considerably larger
than unity.
The above observations lead to a few generalizations.
Mixtures with negative hE and sE usually have positive cp;
mixtures with negative cE usually have positive sE and hE.
For NP/NP mixtures, cE is usually negative, though excep-
tions are observed (e.g., for mixtures of CC14 with an aro-
matic hydrocarbon). For NA/NP mixtures, cE is often nega-
tive, though exceptions occur (e.g., when the polar species is
a ketone). For all other mixtures, positive cE is the norm;
exceptions obtain, e.g., for quasi-ideal mixtures of compo-
nents with similar effective polarity.

To stimulate interest in this method, a card game has been
devised for use in class. Each card has the name of a com-
mon chemical written on one side, and the other side is
blank. There are 52 cards in the deck, with representations of
possible mixture types approximately as indicated above.
The cards are shuffled and someone pulls out two cards
without looking at the compounds. The signs on gE and hE for
the mixture are then guessed. (The best guesses are @.)
Then, an "entropy coin" is taken out to determine the sign on
sE. (Since the chances are about even, a coin flip is as good as
any other guess.)
At this point the compounds are revealed and the mixture
behavior predicted based on the probabilities and connec-
tions given in the tables. Finally, if the Appendix is available
and it has the system (or a related one), it is possible to see
how good the predictions are; if not, one of the group contri-
bution methods for activity coefficients could be used to
predict the sign on gE.
This is one of those exercises where quick students can
often surpass the instructor in accuracy, though good fun is
almost always had by all involved.

Winter 1994

The k vs. h, the h vs. s, and the c vs. 9 diagrams are
effective props for displaying and categorizing the excess-
property behavior of binary liquid mixtures. The six-type
mixture-classification scheme, when used in conjunction with
the diagrams, allows one to make some broad generaliza-
tions about liquid-mixture behavior. Thermodynamic argu-
ments are few and classical in nature.
It would be foolish to ignore molecular concepts as aids
for further organizing and explaining the results presented in
this paper. In fact, the required level of molecular argument
seems to be relatively modest, and precedents exist. We are
preparing a "molecular exegesis" of this Field Guide which
will be a second paper in this study.
The Appendix, which is available to interested readers by
writing the senior authors, contains an enormous amount of
information, which we use in various ways. For example,
one can employ selected data in conjunction with the k vs. h
or the h vs. s diagrams to illustrate and explain trends in
families of binary mixtures containing a common compo-
nent with a series of homologs. Educators will have no
difficulty devising their own examples with this data collec-
tion and finding variations on the card game. There are a
thousand stories here!

This paper was dedicated to Hendrick C. Van Ness on the
occasion of the celebration of this 65h birthday. Support for
developing much of the material was provided to M.M.A. in
the form of a Rensselaer Distinguished Teaching Fellow-
ship, while his preparation of the manuscript was supported
by the donors of the Proteus F. Trope Fund.

1. Malesinski, W., Azeotropy and Other Theoretical Problems
of Vapor-Liquid Equilibrium, Ch. III, Wiley-Interscience,
London (1965)
2. Kauer, E., H.-J. Bittrich, and K Krug, "Eine Systematik
der Mischungen von Nichtelektrolyten," Wiss. Z. Tech.
Hochsch. "Carl Schorlemmer" Leuna-Merseburg, 8, 139
3. Gaube, J., and E. Koenen, "Temperaturabhiingigkeit des
Aktivittskoeffizienten," Chem.-Ing.-Tech., 51, 496 (1979)
4. Kohler, F., and J. Gaube, "Temperature Dependence of Ex-
cess Thermodynamic Properties of Mixtures and Intermo-
lecular Interaction," Pol. J. Chem., 54, 1987 (1980)
5. Koenen, H.-E., and J. Gaube, "Temperature Dependence of
Excess Thermodynamic Properties of Binary Mixtures of
Organic Compounds," Ber. Bunsenges, Phys. Chem., 86, 31
6. Shukla, K.P., A.A. Chialvo, and J.M. Haile, "Thermody-
namic Excess Properties in Binary Fluid Mixtures," IEC
Res., 27, 664 (1988)
7. Gmehling, J., and B. Kolbe, "Limitations of Modern Expres-
sions for the Excess Gibbs Energy," Fluid Phase Equil., 13,
227 (1983)

classroom )




Universidade Federal de Uberlandia
38,400 Uberldndia, MG, Brazil

he average professor must evaluate, or enlist the
help of others in evaluating, a great number of report
assignments written by students. How can that pro-
fessor arrive at a grade rapidly, objectively, and reproduc-
ibly? The following offers a pragmatic and proven solution
to this problem.
Evaluating technical texts is a dynamic task and no single
method for doing so can be perfect. All professors challenge
students in their own particular and personalized way, and
through the years they find that their grading methods change
and improve as new procedures appear in the literature and
become known. In order to become involved in the quest for
improved grading methods, some crucial questions must
first be answered:
Is report writing considered an essential part of en-
gineering apprenticeship?
Is enough time taken in teaching students how to
write a report and check the feedback, or is the
teaching assistant left to struggle along on his own?
How objective is the professor when evaluating a
Can the same mark for the same report be arrived at
both today and a month from now?
Is flawless composition required, or are the calcula-
tions the only thing that matter?
Arriving at a report grade can be as fascinating as a chess

My basic premise is that the grading system
must be corrective, instructive, and simple. I have
extracted ideas from all of the known methods
and have added other, specific, items that
apply to the learning process ...

game. The grader tests the writer as much as the writer
probes the grader. The various methods used to evaluate
technical writing have been explored by Plung,m' but
they are not suitable for the case at hand. They evaluate
only continuous text and require rewriting to arrive at a
score, whereas the reports we are dealing with contain text,
figures, calculations, appendices, and nomenclature, and the
number of assignments as well as the time pressure do not
allow for rewriting.
My basic premise is that the grading system must be
corrective, instructive, and simple. I have extracted ideas
from all of the known methods and have added other, spe-
cific, items that apply to the learning process, with the result
that over the years a pragmatic scoresheet has evolved which
is effective in generating good-quality reports.

A basic scoring guide is presented in Table 1. It addresses
all the basic items that contribute to the quality of a
report: efficiency or timeliness, overall presentation,
quality of editing, technical level, and quality of calcula-
tions, tables, and figures.
It is essential that the students become familiar with the
scoring guide before writing their reports. This is the instruc-
tive aspect of the method since the students can calculate
their own grades and improve them at will before submitting
the report. The method's corrective side comes from the fact
Copyright ChE Division ofASEE 1993
Chemical Engineering Education

Manfred Fehr is a professor of chemical engi-
neering at the Universidade Federal de
Uberlandia (Brazil). He holds a PhD from Laval
University (Canada), has been professionally
active in twelve countries, speaks five lan-
guages, and has published more than fifty tech-
nical papers.

C "hE

that the basic items failed or violated are clearly stated,
and the students are challenged to refine and improve their

The grader assigns either a "1" or a "0" to each item on the
scoring guide, according to whether or not it has been satis-
fied. Items that are not applicable receive a "-" and are not
entered on the score calculation. This is the simplicity of
the method. After reading the report the professor can arrive
at a precise and reproducible score in only a matter of min-
utes. The score is arrived at by dividing the total of numbers

"1" by the total number of items judged. Table 2 shows a
typical score sheet.

The system is also quite flexible and can be adapted to
any number of situations. Items can be added to the list
(or deleted), as necessary, to suit the type of assignment
being graded.


I have used this system for several years now, with excel-
lent results. The students are usually surprised and intrigued

Scoring Guide for Written Technical Communications

1. Work Efficiency
0 1.1 Submitted on first requested date
0 1.2 Submitted on second requested date
0 1.3 Submitted with one granted extension of deadline
0 1.4 Submitted with repeated extensions
O 1.5 Submitted on an arbitrary date
[ 1.6 First or second version accepted
O 1.7 Third or later version accepted

2. Overall Quality ofPresentation
0 2.1 Absence of scratches and visible corrections
E 2.2 Required margin is observed
E 2.3 Divisions of subject or text are clearly visible
O 2.4 Correct linkage to prior sections or parts of work
0 2.5 Figures and tables are on separate sheets
O 2.6 Writing is easily readable
O 2.7 Writing is readable with some effort
0 2.8 Requested format on paper and ink is used
0 2.9 Lines are double spaced
E 2.10 Item 2 is accepted in its first version

3. Quality of Editing
E 3.1 Existence of 100% of required text
0 3.2 Existence of 80% of required text
[ 3.3 Existence of 60% of required text
O 3.4 Mean number of words per sentence is less than 16
O 3.5 Absence of grammar infractions
O 3.6 Percent of sentences containing grammar infractions is less
than 5
0 3.7 Percent of sentences containing grammar infractions is less
than 10
l 3.8 Absence of superfluous words or expressions
E 3.9 Percent of words with more than 4 syllables is less than 8
O 3.10 Percent of sentences containing style infractions (syntax) is
less than 10
0 3.11 Absence of sentences without meaning
D 3.12 Absence of incomplete sentences
E 3.13 Existence of continuous and easily readable text
E 3.14 Text conforms to the required number of words
O 3.15 Item 3 is accepted in its first version

4. Technical Level
[ 4.1 Existence of all items requested
E 4.2 Appropriate use of subtitles

E 4.3
0 4.4
E 4.5
[ 4.6
[ 4.7
E 4.8
O 4.9
[ 4.10
E 4.11

0 4.12
0 4.13
[ 4.14

Coverage of subject matter is above 90%
Coverage of subject matter is above 60%
Existence of correct reasoning for all items covered
Existence of correct reasoning for part of the items covered
Number of original ideas is above zero
Number of original ideas is above three
Absence of nonsense
Absence of meaningless, superfluous, or false arguments
Number of ideas expressed divided by number of sentences
used is above 0.90
Sufficient information for reader to understand everything
Correct reference to attachments and literature
Item 4 is accepted in its first version

5. Quality of Calculations

E 5.1
O 5.2
E 5.3
E 5.4
E 5.5
E 5.6
El 5.7
D 5.8
E 5.9
E 5.10
E 5.11

E 5.12
0 5.13
D 5.14

Existence of titles and subtitles in the sample calculation
Elegant layout on the sheet
Origin of the data is identified
Symbols are defined
Sequence of presentation is clear
Equations are identified
Results are highlighted
SI units are used
Absence of arithmetic errors
Complete results are presented
Sufficient information for reader to follow and understand
the calculations
Presence of 90% of sample calculation required by subject
Presence of 60% of sample calculation required by subject
Item 5 is accepted in its first version

6. Quality of Drawings, Graphs, and Tables
E 6.1 Identification by number and title
E 6.2 Correct layout on the sheet
E 6.3 Requested margin on the left side or top of page
E 6.4 Coordinates are identified
[ 6.5 Scales are stated
E 6.6 Drawing aids used to draw lines and curves
E 6.7 Good quality lettering
0 6.8 Suitable for double reduction on a copying machine
E 6.9 Presence of 90% of information required by subject matter
E 6.10 Presence of 60% of information required by subject matter
E 6.11 Item 6 is accepted in its first version

Winter 1994

by the guide when they first receive it, and they confidently
set out to write their first report. It is a common occurrence
for their first report to be returned to them for rewriting, but
within a short time (usually within two or three assignments)
they have familiarized themselves with the system and they
begin to produce excellent reports.
The scoring guide is not limited in any sense of the word-
not to a specific language or to a specific environment. It is
applicable in any language of instruction and may even be
taken along as a guide for a junior engineer on the job. It has
been called "mechanistic," which in fact it is. It has no
pretensions of being profoundly philosophic. It guides stu-
dents (and possibly engineers) in a very simple and modest
way in the preparation of reports. It challenges them to
define and attain their own degree of perfection. In this sense
it represents a general philosophy of writing instruction that
can be applied to any type of writing. Project reports, review
reports, laboratory reports, and most technical assignments
can all be easily handled with the scoring guide (tailored, as
necessary, to the specific subject).
Different weights can be assigned to different items, or a
total score can be used instead of averaging the six partial
scores as shown in Table 2. In fact, I have tried these differ-
ent approaches in the past, but for various reasons I have
later dismissed them. The score sheet shown in Table 2 has
emerged over the years as the most effective, expedient, and
simplest method of scoring. But opinions differ, and the
beauty of this system lies in the fact that it can accommodate
any number of adaptations.
Various items on the guide require a personal judgment on
the part of the grader. This is unavoidable. As a corrective
instrument, the guide warns writers that these judgments
will be made, and it allows the writers to stack the odds in
their favor by paying special attention to those various re-
quirements. The originality of the whole idea, in fact, lies in
this corrective aspect of the guide.

No grade is assigned to the work until it is judged accept-
able, but unfortunately, acceptability in itself is a subjective
notion. ( For example, how often does peer review of techni-
cal papers produce unanimous results?) The professor will
establish his or her own definition of acceptability by an-
swering and weighting the questions posed at the beginning
of this article, which bear repeating here:
Is report writing considered an essential part of en-
gineering apprenticeship?
Is enough time taken in teaching students how to
write a report and check the feedback, or is the
teaching assistant left to struggle along on his own?
How objective is the professor when evaluating a
Can the same mark for the same report be arrived at
both today and a month from now?
Is flawless composition required, or are the calcula-
tions the only thing that matter?
Although each of the items on the scoring guide calls for a
personal judgment on the part of the grader, that judgment
will (hopefully) remain the same in all cases and not vary
from one paper to the next. For example, an "0" for items
2.7, 3.3, 4.4, or 5.9 on the guide will always render a report
unacceptable to me, and students are so informed at the very
beginning of the course.

No writing guide can be perfect or final, and this one is no
exception. But if the philosophy behind it is accepted and it
is adapted to improve your present grading system, my ob-
jective in writing this paper has been realized.

1. Plung, D.L., "Evaluate Your Technical Writing," Hydrocar-
bon Proc., p. 195, July (1981) 0

Chemical Engineering Education

Scoresheet for Written Technical Communications

Item Evaluated Score
1. Work efficiency 0 1 1 1 75%
2. Overall quality of presentation 1 1 0 0 1 1 1 0 1 67%
3. Quality of editing 0 0 1 1 0 0 1 1 0 1 1 1 1 1 1 67%
4. Technical level 1 0 1 1 0 1 1 0 1 0 0 1 1 0 57%
5. Quality of calculations 1 1 0 0 1 1 0 1 1 1 1 0 1 1 71%
6. Quality of drawing, graphs, tables 1 0 0 1 1 1 0 0 1 0 50%

TOTAL SCORE (Average) 65%


This guide is offered to aid authors in preparing manuscripts for Chemical Engineering Education
(CEE), a quarterly journal published by the Chemical Engineering Division of the American Society
for Engineering Education (ASEE).
CEE publishes papers in the broad field of chemical engineering education. Papers generally
describe a course, a laboratory, a ChE department, a ChE educator, a ChE curriculum, research
program, machine computation, special instructional programs, or give views and opinions on
various topics of interest to the profession.

(* Specific suggestions on preparing papers *

TITLE Use specific and informative titles. They should be as brief as possible, consistent with the
need for defining the subject area covered by the paper.

AUTHORSHIP Be consistent in authorship designation. Use first name, second initial, and
surname. Give complete mailing address of place where work was conducted. If current address is
different, include it in a footnote on title page.

TEXT We request that manuscripts not exceed twelve double-spaced typewritten pages in
length. Longer manuscripts may be returned to the authors) for revision/shortening before being
reviewed. Assume your reader is not a novice in the field. Include only as much history as is needed
to provide background for the particular material covered in your paper. Sectionalize the article and
insert brief appropriate headings.

TABLES Avoid tables and graphs which involve duplication or superfluous data. If you can use a
graph, do not include a table. If the reader needs the table, omit the graph. Substitute a few typical
results for lengthy tables when practical. Avoid computer printouts.

NOMENCLATURE Follow nomenclature style of Chemical Abstracts; avoid trivial names. If
trade names are used, define at point of first use. Trade names should carry an initial capital only,
with no accompanying footnote. Use consistent units of measurement and give dimensions for all
terms. Write all equations and formulas clearly, and number important equations consecutively.

ACKNOWLEDGMENT Include in acknowledgment only such credits as are essential.

LITERATURE CITED References should be numbered and listed on a separate sheet in the
order occurring in the text.

COPY REQUIREMENTS Send two legible copies of the typed (double-spaced) manuscript on
standard letter-size paper. Submit original drawings (or clear prints) of graphs and diagrams on
separate sheets of paper, and include clear glossy prints of any photographs that will be used. Choose
graph papers with blue cross-sectional lines; other colors interfere with good reproduction. Label
ordinates and abscissas of graphs along the axes and outside the graph proper. Figure captions and
legends will be set in type and need not be lettered on the drawings. Number all illustrations
consecutively. Supply all captions and legends typed on a separate page. State in cover letter if
drawings or photographs are to be returned. Authors should also include brief biographical sketches
and recent photographs with the manuscript.



The following 151 departments contribute to the support of CEE with bulk subscriptions.

If your department is not a contributor, write to
c/o Chemical Engineering Department University of Florida Gainesville, FL 32611-2022
for information on bulk subscriptions

University of Akron
University of Alabama
University of Alberta
University of Arizona
Arizona State University
University of Arkansas
Auburn University
Brigham Young University
University of British Columbia
Brown University
Bucknell University
University of Calgary
University of California, Berkeley
University of California, Davis
University of California, Irvine
University of California, Los Angeles
University of California, San Diego
University of California, Santa Barabara
California Institute of Technology
California State Poly Institute
California State University
Carnegie-Mellon University
Case Western Reserve University
University of Cincinnati
Clarkson University
Clemson University
Cleveland State University
University of Colorado
Colorado School of Mines
Colorado State University
Columbia University
University of Connecticut
Cooper Union
Cornell University
University of Dayton
University of Delaware
Drexel University
University of Edinburgh
University of Florida
Florida Institute of Technology
Florida State/Florida A&M University
Georgia Institute of Technology
University of Houston
Howard University
University of Idaho
University of Illinois, Chicago
University of Illinois, Urbana
Illinois Institute of Technology
University of Iowa
Iowa State University
Johns Hopkins University

University of Kansas
Kansas State University
University of Kentucky
Lafayette College
Lakehead University
Lamar University
Laval University
Lehigh University
Loughborough University
Louisiana State University
Louisiana Technical University
University of Louisville
Lowell University
Manhattan College
University of Maryland
University of Maryland, Baltimore County
University of Massachusetts
McGill University
McMaster University
McNeese State University
University of Michigan
Michigan State University
Michigan Technical University
University of Minnesota
University of Minnesota, Duluth
University of Mississippi
Mississippi State University
University of Missouri, Columbia
University of Missouri, Rolla
Montana State University
University of Nebraska
University of New Hampshire
University of New Haven
New Jersey Institute of Technology
University of New Mexico
New Mexico State University
North Carolina A & T University
North Carolina State University
University of North Dakota
Northeastern University
Northwestern University
University of Notre Dame
Technical University of Nova Scotia
Ohio State University
Ohio University
University of Oklahoma
Oklahoma State University
Oregon State University
University of Ottawa
University of Pennsylvania

Pennsylvania State University
University of Pittsburgh
Polytechnic Institute of New York
Princeton University
Purdue University
Queen's University
Rensselaer Polytechnic Institute
University of Rhode Island
Rice University
University of Rochester
Rose-Hulman Institute of Technology
Rutgers, The State University
University of Sherbrooke
University of South Alabama
University of South Carolina
South Dakota School of Mines
University of South Florida
University of Southern California
University of Southwestern Louisiana
State University of New York, Buffalo
Stevens Institute of Technology
University of Sydney
University of Syracuse
University of Tennessee
Tennessee Technological University
Texas A & M University
Texas Tech University
University of Toledo
Tri-State University
Tufts University
University of Tulsa
Tuskegee Institute
University of Utah
Vanderbilt University
Villanova University
University of Virginia
Virginia Polytechnic Institute
University of Washington
Washington State University
Washington University
University of Waterloo
Wayne State University
West Virginia Graduate College
West Virginia Institute of Technology
West Virginia University
Widener University
University of Wisconsin
Worcester Polytechnic Institute
University of Wyoming
Yale University
Youngstown State University

Full Text