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:

Full Text

chaa*emical engineeingeducatio

acknw/led a? i tluhd a ....


wiA a dowonian oaj jads.


Department of Chemical Engineering
University of Florida
Gainesville, Florida 32611
Editor: Ray Fahien (904) 392-0857
Consulting Editor: Mack Tyner
Managing Editor:
Carole C. Yocum (904) 392-0861
Publications Board and Regional
Advertising Representatives:
Gary Poehlein
Georgia Institute of Technology
Past Chairmen:
Klaus D. Timmerhaus
University of Colorado
Lee C. Eagleton
Pennsylvania State University

Richard Felder
North Carolina State University
Jack R. Hopper
Lamar University
Donald R. Paul
University of Texas
James Fair
University of Texas
J. S. Dranoff
Northwestern University
Frederick H. Shair
California Institute of Technology
Alexis T. Bell
University of California, Berkeley
Angelo J. Perna
New Jersey Institute of Technology
Stuart W. Churchill
University of Pennsylvania
Raymond Baddour
Charles Sleicher
University of Washington
Leslie W. Shemilt
McMaster University
Thomas W. Weber
State University of New York

Chemical Engineering Education


2 Virginia Polytechnic Institute & State University,
William L. Conger, Lynn Nystrom


8 Neil Berman, of Arizona State University,
Jenny Berman, V. E. Sater


18 Applications of a Microcomputer Computation
Package, Mordechai Shacham,
Michael B. Cutlip
26 A Systematic Approach to Modeling,
James B. Riggs
36 A Simple Algorithm for Calculation of
Phase Separation, Philip T. Eubank,
Maria A. Barrufet
48 A Course on Making Oral Technical Presentations,
B. S. Brewster, W. C. Hecker
52 A Simpler Way to Tame Multiple-Effect
Evaporators, Donald D. Joye,
F. William Koko, Jr.


22 An Undergraduate Experiment in Alarm System
Design, R. A. Martini, C. Sullivan, A. Cinar
42 The Large Laboratory Course: Organize it to
Parallel Industrial Process Development,
Roger E. Eckert, Robert M. Ybarra


30 Rationale for Incorporating Health and Safety into
the Curriculum, Marvin Fleischman

7, 11, 16, 17, 51 Book Reviews

11 Positions Available

CHEMICAL ENGINEERING EDUCATION (ISSN 0009-2479) is published quarterly by 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. Advertising mate-
rial may be sent directly to E. O. Painter Printing Co., P. O. Box 877, DeLeon Springs, FL 32028. Copyright
1988 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 with
120 days of publication. Write for information on subscription costs and for back copy cost and availability.
POSTMASTER: Send address changes to CEE, Chemical Engineering Department, University of Florida,
Gainesville, FL 32611.


Aerial view of the VPI campus showing the College of Engineering in the upper left portion.



Blacksburg, VA 24061-6496

versity, better known by the initials VPI and/or
the nickname "Virginia Tech," is located in
Blacksburg, Virginia, and is the state's largest univer-
sity, resting on a plateau 2,100 feet above sea level
between the Blue Ridge and the Allegheny moun-
tains. Founded in 1872 as an agricultural and mechan-
ical school and as the landgrant university of the Com-
monwealth, VPI has become a comprehensive institu-
tion with seven colleges and a professional school of
veterinary medicine. Overall, 23,000 students are cur-
renty enrolled. This includes about 19,000 under-
graduate students, of which 5,000 are students in the
College of Engineering. A Board of Visitors (ap-
pointed by the Governor), to whom the President re-
ports, governs the university.
The College of Engineering is directed by Dean
Paul E. Torgersen, a member of the National

Academy of Engineering. He altered the course of the
college in the early 1970s, allowing it to compete with
the premier engineering schools in the country
through a commitment to research, graduate educa-
tion, and scholarship. The change was gradual at first,
but has become much more dramatic in the past dec-
ade. As one result, the college was listed in the top
twenty in a 1987 U.S. News and World Report survey
of the best graduate engineering schools in America.
It also now ranks in the top ten percent of the 220
engineering institutions reporting their research ex-
penditures to ASEE. These citations have occurred
because the external support of the college for faculty
research has grown from about $6.5 million awarded
in 1978-79 to $18.6 million in 1986-87. The college pre-
sently has about 270 faculty, two-thirds of which are
engaged in research. While trying to hold the under-
graduate population relatively constant (5,325 under-
graduates in the fall of 1979, 6,280 undergraduates in
the fall of 1987), a substantial increase in the graduate
Copyright ChE Division ASEE 1988


student population has occurred (518 graduate stu-
dents in the fall of 1979, 1,167 graduate students in
the fall of 1987). These changes are mirrored in the
degrees presented by the college. In 1978-79 there
were 661 bachelor, 157 master, and 35 doctoral de-
grees awarded by the college, and in 1986-87 these
figures had increased to 954 bachelor, 296 master, and
77 doctoral degrees. The university grew significantly
in the 1960s, from about 7,000 students to its present
size. The students have a choice of studying one of
twelve different engineering undergraduate cur-
riculums in ten degree-granting departments. In addi-
tion, the freshman year is administered by the Divi-
sion of Engineering Fundamentals.
The College of Engineering and the university
maintain a rich history in both the state and the na-
tion. In the fall of 1985, the college celebrated the
100th anniversary of the granting of its first four-year
BS degree. That same fall, the Department of Chem-
ical Engineering celebrated its 50th year as a modern
department of chemical engineering at VPI.

The Chemical Engineering Department was estab-
lished in 1919, and two years later offered post-
graduate degrees. In 1929, the name was changed to
Chemistry and Chemical Engineering.
In 1935, Frank C. Vilbrant founded the modern
Department of Chemical Engineering at VPI, and he
remained as department head until 1953 when Fred
W. Bull became head. During Dr. Vilbrant's tenure,
essentially all of the students were members of the
Corps of Cadets. There was a strict code of conduct
on campus associated with the military flavor of the
institution. For example, students were not allowed
to have automobiles in Blacksburg. A student found
with an automobile was suspended. Dr. Vilbrant,
while he seems not to have liked some of the military
codes of conduct, made good use of the general atmos-
phere it created; he ran a tight ship.
After World War II and the Korean War the cli-
mate of the university changed; the military flavor of
the institution began to subside and, toward the end
of Fred Bull's headship, the civilian portion of the stu-
dent body began to dominate. The mandatory mem-
bership in the Corps of Cadets for all undergraduates
became a problem as veterans were returning to pur-
sue their education. Also, at about this same time
there was an influx of transfer students, due in par-
ticular to the new co-op program started in 1952 at
the Norfolk, Virginia, division of William and Mary
and VPI. During Fred Bull's last year as department
head (1964), the president of the university, T. Mar-

shall Hahn, made membership in the Corps of Cadets
voluntary. By then, the civilian student population
surpassed the membership of the Corps.
One other happening in those years influenced the
development of the university and thus of the depart-
ment. Until 1964, VPI had an association with the
all-female Radford College in Radford, Virginia. Rad-
ford became independent of Virginia Tech in 1964 and
began to admit men in 1972. Simultaneously, the pop-
ulation of women on the VPI campus began to grow,

The Cascades, situated a short distance from the Ap-
palachian trail and just 25 minutes from campus.

and as a result, the department began to see its first
women students during the 1970's.
Fred Bull, who was known as a strong department
head and as a forgiving and compassionate man,
turned the department over to Gerry Beyer in 1964.
Those were years of rapid growth and turmoil in the
university. The department faculty realized that the
curriculum needed modernization in order to allow the
graduates to compete better in the major graduate
programs in the country. The department had always
had a strong practical orientation and had graduated
many who became captains of industry, such as Clif
Garvin and Al Giacco, the recently retired CEOs of
Exxon and Hercules, respectively. The department
was respected by industry for the quality of its under-


graduates, but it was felt that a greater portion of the
graduates should attend graduate school.
In 1967, Nelson F. Murphy became department
head and the department entered an era of high fac-
ulty turnover. This continued through the 1970s and
into the 1980s. Henry McGee became head in 1971 and
introduced stronger ties to chemistry during his ten-
ure. Over this period, the size of the faculty increased
significantly to the present number. Dr. McGee's
headship coincided with the thrust in the college for
stronger research, graduate education, and scholar-
ship. He seized this opportunity to build strength in
these areas in the department. In 1981, Dr. McGee
resigned the headship to return to research and teach-
ing. Chester Spencer, department head of materials

S. the college was listed in the top
twenty in a 1987 U.S. News and World Report
survey on the best graduate engineering schools in
America. It also now ranks in the top 10% of
the 220 engineering institutions reporting
their expenditures to ASEE.

engineering, served as the interim head until Bill Con-
ger arrived in 1983 to assume the position.
A feeling of stability has marked the last five years
in the department. The graduate program has grown
significantly while the size of the undergraduate stu-
dent population decreased and then stabilized. Since
a number of faculty are now approaching retirement
age, major changes in the faculty are expected to
occur. Last year, Arthur Squires, (a member of the
National Academy of Engineering) formally retired,
although he remains very active on campus and re-
tains an office. In 1987, Ken Konrad left to take a
position with Exxon and Roland Mischke took a termi-
nal year leave to help with the ChE program at
Hampton University. These openings created an op-
portunity to add new input to the program at a time
in the history of our profession when we are searching
for new directions for chemical engineering.

For the fall of 1987, there were thirteen fulltime
faculty and one emeritus professor, Arthur Squires,
very active in research. The department is seeking
one, maybe two, additional faculty for the fall of 1988.
Don Baird came to VPI in 1978 from Monsanto.
Initially, he held a joint appointment between the de-
partments of engineering science and mechanics and
chemical engineering, but in 1981 he became fulltime

in chemical engineering. Although Don received his
PhD at Wisconsin, we seem to hear more about his
days at Michigan State where he earned All-Big Ten
and Academic All-American honors as a linebacker
and offensive guard on the MSU football team. Don is
interested in the application of rheology to the proces-
sing of polymeric and food materials. His research
runs from very fundamental to applied levels.
Gerry Beyer is another Wisconsin graduate and
proud of it. In 1964 he came from the University of
Missouri where he was department head, joining the
VPI faculty as department head and remaining as head
until 1967. Gerry is interested in energy production
and use, separation and purification of materials, and
the application of computers to the modeling of chem-
ical and magnetic systems. His course on the use of
the personal computer has been presented nationally
as part of the AIChE Today series. A confirmed
fisherman, he has caught several trophy-size fish, in-
cluding a striped bass which he carried fresh and drip-
ping into Bill Conger's office one afternoon.
Bill Conger was recruited to Tech as department
head from the University of Kentucky in 1983. Bill
received his degree from the University of Pennsyl-
vania. His interests have included hydrogen as an al-
ternate energy source and the thermodynamic
analysis of energy intensive processes. Presently, he
spends most of his professional time administering the
department. He continues his outdoor activities (one
of the lures of the Blacksburg, Virginia, area) includ-
ing his work with Boy Scouts, his interest in canoeing,
and even an occasional fishing trip with Gerry Beyer.
David Cox is one of the newest additions to the
faculty. After graduating with a PhD from Florida, he
spent two years at the National Bureau of Standards
as a National Research Council Postdoctoral Associate
before joining the faculty in 1986. David uses the spec-
troscopic tools of ultrahigh vacuum surface science to
study the chemistry of solid surfaces. His work is
aimed at understanding the adsorption, desorption,
and surface reaction processes important in
heterogeneous catalysis, chemical gas sensing, and
the growth of interfaces on electronic materials. When
he cannot be found in his laboratory, there is a good
chance he can be located on a golf course.
Mark Davis chose to join the department after
completing his PhD work at the University of Ken-
tucky. Named a Presidential Young Investigator sev-
eral years ago, his work under the PYI program has
been extremely successful. Mark's major area of in-
terest is the study of new molecular sieves. Dow
Chemical Company and Mark recently announced the
development of a zeolite with a 13 Angstrom opening


Mark Davis holds a model of VPI-5 the new molecular
sieve with a 13 Angstrom pore that he and DOW have

which could have a major impact on such areas as
petroleum catalytic cracking and chromatography.
Mark has also won the ASEE Dow Young Faculty
Award. He is an accomplished scuba diver and under-
water photographer. Examples of his work can be
seen on display in some of the Bahamaian resort
Y.A. Liu, a graduate of Princeton, came to the
department in 1982 from Auburn University. Y.A. is
the Frank C. Vilbrant Professor of Chemical En-
gineering and has received several prestigious
awards, including the 1984 Western Electric Award
of ASEE and the 1986 Catalyst Award of the Chemi-
cal Manufacturers Association. He is currently in-
terested in computer-aided design and process synthe-
sis, fluidization processes, and magnetochemical en-
gineering. He is active as an advisor to the Student
Christian Fellowship at VPI and has led Bible studies
for international students for years.
Don Michelsen, a Cornell graduate, arrived in
Blacksburg in 1966 after securing industrial experi-
ence with Dow Chemical Company. He spent a period
away from the department as Director of Career Ad-
vising and Director of the Graduate Engineering Pro-
gram in Northern Virginia. Don is interested in indus-
trial waste treatment, passive solar energy design,
and treatment of hazardous wastes. He enjoys sailing
and this past summer started sprouting a beard as
"every old salt has to have a beard."
Henry McGee, a Georgia Tech graduate, left his
alma mater to direct the department. Henry promotes
his interest in closer ties between chemistry and
chemical engineering through writing papers on the
subject and is also available as a speaker on the topic.

He has always had several irons in the fire at the
same time and is present developing a large property
on nearby Claytor Lake, where he keeps his cabin
cruiser and enjoys water-related activities most of the
year. Henry is interested in chemically pumped lasers
as well as the application of laser irradiation in chem-
ical reactions and in separations.
Peter Rony, a graduate of Berkeley, came to the
department from Exxon in 1971. He is noted for his
textbooks and his work with laboratories. He has
earned several awards, including the Dreyfus Teacher
Scholar Prize in 1973 and the DELOS Award in 1984.
He is interested in the creation of novel measurement
devices and instrumentation. Peter likes to travel and
he delights in water sports. It is no surprise that his
favorite places to visit include Florida, Cancun, and
Southern California.
Felix Sebba, a very active senior citizen of our
department (age 75), came in 1979 from the Univer-
sity of Withwatersrand in Johannesburg, South Af-
rica. He has degrees from the University of Capetown
and the University of London. Felix became a U.S.
citizen in the fall of 1987. He is interested in the phys-
ical and chemical behavior at interfaces. He has de-
veloped separation processes using colloidal gas aph-
rons and liquid core aphrons, both of which he in-
Bill Velander is the newest member of our facul-
ty, having arrived from Penn State in January of 1987.
Prior to his work at State, he was employed for sev-
eral years at Merck. He is interested in cell-to-cell
adhesion and synthetic vesicle membrane structure as
well as the separation of proteins from biological
fluids. Bill enjoys softball, basketball, golf, and (since
coming to VPI) volleyball.
George Wills, another Wisconsin graduate, came
to VPI in 1964. He is known for his excellence in
teaching and has won the Wine Award, a VPI award
for such excellence. He was instrumental in develop-
ing the summer unit operations laboratory program in
England in which seven to ten of our students partici-
pate each year. George likes to travel, especially in
England and Europe. He takes trips abroad as often
as he is able.
Garth Wilkes, the Fred Bull Professor of Chemi-
cal Engineering, earned his degree at the University
of Massachusetts and came to VPI from Princeton in
1978. Garth is co-director of the Polymer Materials
and Interfaces Laboratory with James McGrath of
chemistry. PMIL is an extremely successful interdis-
ciplinary program with fourteen faculty members
from six different participating departments. A total
of more than 150 people work with the faculty. His


Four years ago, the College of Engineering made owning an IBM PC (or compatible) a
requirement for all entering students. Now there are over 6,000 PCs in the college. The chemical
engineering students use them extensively in their classes and many software packages are provided. The
most difficult part of introducing these machines . was requiring the faculty to become proficient.

own primary interest is in the structure/property re-
lationships of polymeric solids, with an emphasis on
the molecular approach. Garth is an avid out-
doorsman, an accomplished hunter and fisherman.
Also, he has played guitar professionally with several
rock groups.

The undergraduate students come mainly from
Virginia and a strip of states along the Eastern Sea-
board. In the most recent sophomore class, one out of
every ten students was valedictorian of his/her high
school class, and well over 60% were in the top 10%
of their high school class. Average SAT score was
about 1240.
In fall 1988 the university will convert from the
quarter system to the semester system. This conver-
sion provided the opportunity for an in-depth analysis
of the curriculum. The faculty decided to seize upon
the flexibility offered in the new ABET and AIChE
statement on the chemistry requirement. In the new
curriculum, only the first semesters of organic and
physical chemistry are required. In place of the second
semesters of these courses, the student selects from
a list of chemical science electives which includes the
second semester of organic and of physical chemistry,
and also includes advanced courses in inorganic
chemistry, solid state chemistry, biology and
biochemistry, polymer chemistry, etc. This allows the
student to acquire a wider background than was pre-
viously possible at VPI. However, the advising func-
tion of the faculty is now much more important since
the faculty member must take more time to discuss
with the student the various options that are available
and what they mean.
The undergraduate program is known for its
strong control and design components. This will not
change under the semester system. Two required
courses and one elective course will remain in the un-
dergraduate control sequence. In the last five years
the students have won two third places and one second
place in the AIChE Process Design Contest. Students
have access to Aspen Plus and other software pack-
ages useful in the design course.
Probably the summer Unit Operations Labora-
tory, supported by many industrial friends, is the one
aspect of the undergraduate program most remem-

The summer Unit Operations Laboratory-one aspect of
the undergraduate program most remembered by VPI

bered by the graduates. Roland Mischke called it a
"slice of life." Students work in teams on experiments
all day, every day, for five weeks. It is an intense
experience. Each group is assigned a classroom in
Randolph Hall and they essentially move into the
building. Refrigerators, easy chairs, computers, etc.
appear in the classrooms as these will be the students'
"homes" for the period of the lab. The routine is bro-
ken about halfway through the session with a cookout
and picnic for all. Students complain about the pres-
sure while taking the course, but years after gradua-
tion, they recall the coursework as one of the most
rewarding experiences of their undergraduate pro-
Four years ago, the College of Engineering made
owning an IBM PC (or compatible) a requirement for
all entering students. Now there are over 6,000 PCs
in the college. The chemical engineering students use
them extensively in their classes and many software
packages are provided. The most difficult part of in-
troducing these machines into the program was re-
quiring the faculty to become proficient. Almost every
course now has significant use of the machines.

The department maintains a population of about
fifty graduate students. These are currently about
equally distributed between the MS and PhD degree


programs. Many of the MS students intend to continue
for the PhD. There is a core curriculum of four
courses: advanced thermodynamics, advanced kine-
tics, and two transport phenomena courses. Beyond
these requirements, the students are free to design
their course of study. Under the semester system, 30
semester credits (10 of which can be research) are
required beyond the baccalaureate for the MS degree,
and a total of 90 semester credits (60 of which can be
research) are required beyond the baccalaureate for
the PhD. Doctoral students must pass a Preliminary
Examination generally within the first two years of
study. An oral examination on the proposal for their
PhD research (Qualifying Exam) is also required.
A wide variety of research areas exist within the
department, as mentioned in the section on faculty,
but the majority of the graduate students and funding
is in five areas: polymers, catalysts and solid state
surface science, environmental engineering and sur-
face chemistry, fluidization engineering, and bioen-
gineering. In particular, the polymers (Polymer Mate-
rials and Interfaces Laboratory, PMIL) and the
catalyst programs have international reputations and,
as a result, are very much in demand by prospective
graduate students.
As mentioned previously, the department is in the
process of adding more faculty. We intend to continue
strengthening the research effort in the department
while maintaining the traditionally strong teaching
commitment. Ground will be broken soon for an addi-
tion to Randolph Hall. It will include modern facilities
for the undergraduate Measurements and Control
Laboratory and Unit Operations Laboratory. Initial
planning is beginning for a new engineering building
which would house the department and its research
laboratories. Present space in Randolph Hall would
then be turned over to Aerospace and Ocean En-
gineering and to Mechanical Engineering for a much
needed expansion of their facilities.
As the change to the semester system occurs,
there will be opportunities for the curriculum to
evolve to meet the needs of the changing profession.
Changes to the new approved semester curriculum
are under consideration which, if implemented, will
give the student even greater freedom of choice.
The department looks forward to new facilities, an
even stronger faculty interested in research and
teaching, a more flexible curriculum, and increased
emphasis on the quality of the graduate program. The
department is in a position of strength, facing a future
that is bright for the faculty, the students, and the
profession. [E

JIN book reviews

by H. S. Fogler
Prentice-Hall, Inc., Englewood Cliffs, NJ 07632
(1987) $49.95
Reviewed by
John L. Falconer
University of Colorado
This is an excellent text for an undergraduate reac-
tor design course. Mole balances and reactor staging
are introduced first. Then rate laws, stoichiometry,
and isothermal reactor design are discussed. The
order of presentation, though somewhat different
from other texts, is good. The next chapter, on
analysis of rate data, also discusses laboratory reac-
tors. Catalysis and homogeneous kinetics are then
presented before nonisothermal reactor design and
multiple reactions. The text concludes with two chap-
ters on external diffusion and diffusion in porous
catalysts, one on multiphase reactors, and two on res-
idence time distribution and nonideal reactors. The
Appendices present common integrals used in reactor
design, numerical techniques for integration and dif-
ferentiation, and a series of guided design problems.
Both SI and English units are used in the text. A
complete Solutions Manual is available.
The text is well written, the order of presentation
and the approach are very good, the notation is clear,
the figures and print type are excellent, and the
number of errors is small. A number of desirable fea-
tures separate it from other reactor design texts. The
text emphasizes problem solving and reasoning rather
than memorization. Thus, at the end of most of the
chapters, a few pages are devoted to techniques used
in problem solving. This is valuable, though it could
be improved upon by showing the application of these
techniques to reactor design problems. Important
equations are boxed in, and the margins are used to
emphasize and summarize important points. Volume
changes due to gas-phase mole changes or phase
changes are covered more completely than in most
texts. A summary of important points is presented at
the end of each chapter. Numerical techniques for
solving differential equations are presented.
A strong point of the text is the inclusion of numer-
ous example problems and homework problems for
real reactions in each chapter. A variety of challenging
problems are given, such as those taken from the

Continued on page 41.


S educator

Neil, above, circa 1967, with his "homemade" equipment and today, at right, in a
more modern lab setting.


of Arizona State University

Gutierrez-Palmenberg, Inc.
Phoenix. AZ 85017
Arizona State University
Tempe, AZ 85287-6606

When Neil and Sarah Berman first arrived in
Phoenix in September of 1964, in a car with no air
conditioning, they wondered how anyone could live
there. Neil joined the chemical engineering faculty at
Arizona State University as planned, however, and
four years later they bought their first air conditioned
car in order to take their two month old son Dan on a
trip to Texas.
Chemical engineering at ASU was just beginning
when Neil joined the faculty. The College of Engineer-
ing was established in 1956 and the first chemical en-
gineering professor, Cas Reiser, arrived in 1958, fol-

lowed by Sam Craig in 1960. Gene Sater and the first
graduates came in 1962. When Neil Berman was hired
in 1964, the department had enough faculty to become
accredited and to establish a chapter of AIChE.
Teaching loads were high (three or more courses each
semester), lab space for both teaching and research
was scarce, and graduate students were difficult to
Since then, the department has expanded to in-
clude biomedical and materials engineering and has
twenty-three full-time faculty with a well-respected,
established graduate program.
Neil is a native of Milwaukee, Wisconsin (born in
1933). He attended what was then the University of
Wisconsin in Milwaukee (Extension) for two years and
then completed his BS in chemical engineering at
U.W. Madison. There was a three-year gap in his uni-
c Copyright ChE Division ASEE 1988


Neil continues to have an interest in air pollution and works with both students and government
on the current problems in Arizona. The scientific problems concern the extremes of mixing that occur in the
desert, from shallow ground-based temperature inversions at night to vigorous instability in [midday].

versity education while he worked for Standard Oil
Company in California and subsequently served as a
sanitary engineering officer in the U.S. Army in
Puerto Rico. As the army staff at Rodriguez Army
Hospital in San Juan was reduced he took on jobs
ranging from restaurant inspector to company com-
mander. Neil, however, rarely talks about anything
in Puerto Rico except playing tennis with Charley
Paserell. Charley has probably forgotten all about it,
since he was only fourteen at the time.
After the army, Neil attended the University of
Texas at Austin where he earned his PhD working
with John McKetta. He also received a MS in chemical
engineering and an MA in math from the University
of Texas. While in Austin, he met his wife, Sarah, and
they were married on June 3, 1962-Jefferson Davis'
birthday-the day after Neil's PhD commencement.
After completing his PhD, Neil went to work at du
Pont Central Research in Wilmington, Delaware.
They had last hired a chemical engineer about fifteen
years before and didn't really know what one was good
for, so they presented him with problems that the
chemists and physicists considered impossible. Neil
wasn't as successful as the other chemical engineers
at solving such problems until they asked him to try
to make Cr02 at conditions less severe than 350 atm
in a sealed platinum tube. Not knowing any better, he
proceeded to put the precursors in an open steel cup
and insert the cup in a closed pressure vessel. He
found that the reaction ended when the pressure
reached 15 atm and the products had the desired fer-
romagnetic properties. He left du Pont for ASU be-
fore his original experiment was commercialized as
magnetic tape.
While at du Pont, Neil noticed the first report of
velocity measurement by Laser Doppler Velocimetry,
and after he settled at ASU, he decided to build a
similar instrument. Fortunately, the Army had sent
a group of officers to ASU to obtain masters degrees
in mechanical engineering, and several of them helped
build the first instrument. It was extremely difficult
to align the optical system, so to test the system he
searched for a flow experiment which would stay in
alignment throughout the experiment. The ideal ex-
periment was entrance flow. One experiment led to
another, and polymer flows were also analyzed. Since
Dan Jankowski, a faculty member at ASU, was in-
terested in stability, Neil's next research topic was

Neil and his daughter, Jenny, take time out to share a
little lab talk and experimentation.

pipe flow stability. Neil broadened his scope to turbu-
lent flow during two summers at NASA Lewis
working with John Dunning.
Although Neil gives the appearance of working by
himself, in real life he does best in collaboration with
others. He has a long list of associations starting with
the Army officers who built his equipment and fol-
lowed by Gene Cooper who came from NASA to ASU,
John Dunning at NASA, Bill George then at Penn
State, H. Usui of Japan, H. Tan and Guou of China,
and currently H. W. Bewersdorff of West Germany.
He feels fortunate to have had the right colleagues
and students on hand at the right times.
At NASA Lewis, John Dunning and Neil mea-
sured the velocity fluctuations in turbulent flow to test
the applications of LDV. Several limits in laminar flow
had been previously studied by John. In turbulent
flow, it was clear that turbulence represented a fre-
quency modulated (FM) signal, but there was a noise
component present in addition to the turbulence. This
noise component had been extensively studied at Bell
Labs many years before, and a similar analysis for
LDV turbulent signals was made by Bill George and
John Lumley. There was a difference between these
two analyses that Neil could not figure out, so he
asked Bill George if he could come to Penn State on
his sabbatical to continue work on the problem. Bill
agreed. The discrepency was quickly resolved and the
NASA experiments finally interpreted, but Bill did
not have a functioning Laser Doppler instrument since
homemade ones continued to have alignment difficul-


ties. Neil turned to the other current research topic
of interest in the aerospace department at Penn
State-drag reduction by polymers.
Several important experiments in this area were
developed through discussions (sometimes called
shouting matches by observers) between Bill George
and Neil, and the experiments were carried out with
the aid of personnel in the Applied Research Labora-
tory at Penn State. Neil and Bill proved, at least to
their satisfaction, that the time scale was the impor-
tant parameter in the interaction between high
molecular weight polymer molecules and turbulence.
The clever experiment developed required that the
direction of change be one way for a time scale and
the other way for a length scale and did not depend
on the absolute scale of the experimental numbers.
After returning to ASU with a much greater foun-
dation in turbulence and signal analysis, Neil began
work on two aspects of turbulent mixing-air pollu-
tion and drag reduction. Air quality in the Phoenix
area had been perceived to be declining, but no one
was studying the problem. Some undergraduate stu-
dents were interested, however, and agreed to locate
the data and run the computer programs. This air pol-
lution research has had very little financial support
over the past fifteen years, but has involved many
students who have gone on to careers in the environ-
mental area.
Neil's involvement with air pollution research in
the middle 1970's led to some publicity in the local
newspapers (partly due to his friendship with a local
reporter) and to his subsequent appointment to the
governor's commission on Air Quality. He pointed out
to local authorities that the wind speed instruments
originally installed at the monitoring stations had too
high a starting speed to be of any use in Phoenix. He
also suggested that wintertime daylight savings time
had advantages in reducing pollution.
Neil continues to have an interest in air pollution
and works with both students and government on the
current problems in Arizona. The scientific problems
concern the extremes of mixing that occur in the des-
ert, from shallow ground-based temperature inver-
sions at night to vigorous instability in the middle of
the day. Critical aspects of the air pollution problem
include how the inversion sets in during the evening
rush-hour in winter, creating high levels of CO, and
how the inversion breaks up in the morning in sum-
mer, diluting the reactants which produce photochem-
ical oxidants.
In his drag reduction research after his return
from Penn State, Neil studied the role of molecular
weight distribution. The experiment was to separate

Hoppalong Catsidy Berman, always ready to investigate
running water and its vortices.

molecular weights using gel chromatography, to
analyze the fractions, and to measure the drag reduc-
tion ability. The analysis and drag reduction measure-
ments had to be done simultaneously and soon after
collection from the column, to avoid degradation.
Nothing was automated, so the group consisted of
Neil, two graduate students (John Yuen and Sharam
Elihu), and Neil's children (Jenny and Dan). This dif-
ficult experiment required density, viscosity, and con-
centration measurements in addition to the pipe flow
Dr. Usui came from Japan to show Neil how to
make a pipe test section from mylar film. A few years
later, Hung Tan spent a year measuring velocities in
a submerged jet and developing a new experimental
technique for laser velocity measurements. It seems
that a year's collaboration results in at least three pa-
pers, but it takes a while to write them. Neil has per-
sonally written most of his nearly 100 papers begin-
ning with the first draft. Only about 10% were jointly
written, although joint authors contributed in other
ways. Sarah, Jenny, and Dan have all helped with the
proofreading, and Jenny co-authored one of his pa-
In addition to his research work, Neil has taught
most of the undergraduate courses at ASU at one time
or another. He has also taught graduate courses. Cur-
rently he is teaching transport phenomena and special
courses in air pollution, math, turbulent transport and
mass transfer. Some of his former students accuse him
of not being able to speak without at least three
blackboards to fill with equations. He is currently


Graduate Studies Coordinator for the department an(
has been active in university and community service
as well. He has served three terms on the faculty sen
ate where he chaired the Personnel Committee and
was a member of the Executive Committee and the
Academic Affairs Committee. He has also served on
numerous special committees and search committees.
He is a recipient of the Distinguished Research Award
from ASU. This award recognized Neil's excellent re-
search program by funding a visiting scholar (Bob
Kabel of Penn State) to take over Neil's teaching as-
signments for a year, allowing him to devote all of his
time to research and to working with graduate stu-
Neil has always tried to involve everyone around
him in his work. His children grew up helping him in
his lab. Sarah has, on more than one occasion, been
offered offices in the local section of AIChE for her
dedication in accompanying him to meetings for so
many years. Now, he finally has a pet who shares his
interest in turbulence; Hoppalong Catsidy, the newest
addition to the family, is always ready to investigate
running water and the vortices formed when it flows
down the drain. O

W 16 3 book reviews

National Academy Press, 2101 Constitution Avenue,
NW, Washington, DC 20418
Reviewed by
Bryce Andersen
Southeastern Massachusetts University
At the request of the National Science Foundation,
the National Research Council embarked on a major
study of engineering education and practice in the
United States. A distinguished 26-member committee
drawn from education, business, and industry was ap-
pointed in 1982. Fifty additional persons on nine
panels prepared background papers for the commit-
tee. In 1985 the National Academy Press began pub-
lishing a nine-volume paperback series of reports from
the committee and its panels. The general report of
the committee, Engineering Education and Practice
in the United States: Foundations of our Techno-
Economic Future, was reviewed in the Spring 1986
issue of Chemical Engineering Education. Sub-
sequent volumes were published on undergraduate
and graduate engineering education. This review
Continued on page 29.

Use CEE's reasonable rates to advertise. Minimum rate
1/s page $60; each additional column inch $25.

Department of Chemical Engineering

Full-time, tenure track position at the Assistant or Associate Pro-
fessor level available commencing September 1988. Candidates
must have a PhD in Chemical Engineering by September 1988;
industrial experience is desirable. Teaching and research in-
terests are not restricted to any particular areas. Villanova Uni-
versity is a fully-accredited institution which emphasizes under-
graduate and graduate teaching. The Chemical Engineering de-
partment offers programs leading to the B.Ch.E. and M.Ch.E.
degrees. The successful candidate will be expected to develop an
active research program compatible with the faculty's other obli-
gations to the University. Applicants should submit a curriculum
vita, a discussion of teaching and research interests, and a list
of three references before March 15, 1988, to Professor Vito L.
Punzi, Search Committee Chairman, Department of Chemical En-
gineering, Villanova University, Villanova, PA 19085. Villanova
University is an Equal Opportunity/Affirmative Action Employer.
Women and minorities are especially encouraged to apply.

NJIT seeks applications for chairperson of chemical engineering,
chemistry and environmental science. There are 35 tenure-track
faculty in the department which offers accredited undergraduate
degrees in chemical engineering and applied chemistry, and grad-
uate degrees in chemical engineering, chemistry and environ-
mental science. Research funding in the current year is
approximately $1.3 million with broad support of local petroleum,
pharmaceutical and food industries. Candidates must qualify for
a tenured position as full professor and have demonstrated lead-
ership and administrative skills. Qualifications: earned doctorate in
chemical engineering, chemistry or related field; a record of
achievement in research; commitment to excellence in teaching.
Inquiries regarding this position may be addressed to Dr. R.
Parker, chairman, Search Committee, (201) 596-3588.
NJIT is the comprehensive technological university of New Jersey
with 7,500 students enrolled in baccalaureate through doctoral
programs within three colleges: Newark College of Engineering,
the School of Architecture, and the College of Science and
Liberal Arts.
NJIT does not discriminate on the basis of sex, race, color, handicap,
national or ethnic origin, or age in employment.
Send resume including publication record and names of three
references by February 12, 1988: Personnel Box CCCE.

Neaw Jersay
Inlstitutie o Tcohnology
Newa, NMw Jerwmy 010a


[mJi views and opinions



Massachusetts Institute of Technology
Cambridge, MA 02139

W E ARE IN THE middle of an extraordinary time
of change, both in the explosive growth of new
opportunities and in the restructuring of older indus-
tries. How should we educate our students so they
will be effective in solving the most important prob-
lems of tomorrow?
Perhaps the number-one problem in America
today is how to keep our industries competitive in the
world. We have seen our commanding lead in cutting
edge technologies vanish in industries such as textiles,
steel, consumer electronics, and automobiles, and
weaken in industries such as computer chips and high
purity chemicals. Currently, America has only two in-
dustries that enjoy a healthy surplus of export over
import: the manufacturing of commercial airplanes
and of chemicals. We have a current export surplus in
chemicals of about eight billion dollars, so we are still

James Wei was named Warren K. Lewis Professor and Head of the
Department of Chemical Engineering at the Massachusetts Institute of
Technology in 1977. Prior to assuming this post, he served as Allan P.
Colburn Professor of Chemical Engineering at the University of Dela-
ware for six years. A graduate of Georgia Institute of Technology and
MIT, Dr. Wei spent the first fifteen years of his career conducting re-
search studies in catalysis and reaction engineering for Mobil Oil Re-
search, and performing long-range forecasts on the supply and de-
mand for energy for Mobil Oil Corporation. He has written 70 technical
publications, and contributed as a writer or editor to four textbooks.
Copyright ChE Division ASEE 1988

The second chemical engineering paradigm
of transport phenomena arrived in 1960 in a
textbook by Bird, Stewart, and Lightfoot. This
approach declared that the proper study of chemical
engineers is the molecular phenomena that
are fundamental to the understanding and
performance of chemical equipment.

above water, but how long can we keep this lead?
Why are Boeing and the chemical companies better
than the rest? It may be that these companies see
themselves as technology companies, try to gain com-
petitive advantages through research and innovation,
and are managed by people committed to technological
Many reasons are given for the decline of once
mighty American industrial power:
Corporate capital is in the hands of funds managers and
corporate raiders who are interested in quick profit rather
than long term strength.
Corporate managers lack long-range vision and commit-
ment in investing in cutting edge technologies.
There is government interference by EPA, OSHA, FDA,
and the anti-trust division of the Justice Department.
The labor force is lazy, overpaid, and lacks commitment
and work ethics.
It is interesting that there is little or no complaint
that America's decline is due to the inadequate educa-
tion of engineers. However, there is no reason to be
complacent, and we should continue to expand the
knowledge and tools that our graduates need to be
armed with as they formulate and solve problems.

Since the role of chemical engineers is changing
fast, it would be well to spend a moment in discussing
who they are. Chemical engineers are defined by what
they know and what they do. Their knowledge is
gained through the undergraduate curriculum and
through their work experience in the chemical proces-
sing industries. Chemical engineers make things
through chemistry. They also build plants, make
equipment, teach students, provide government ser-



vices, clean up the environment, study physiology,
and make artificial organs.
The principal homes of the chemical engineers are
the chemical and petroleum refining industries, shown
in Table 1. They can also be found in the more ex-
tended chemical processing industries, which employ
27% of all manufacturing labor and provide 37% of all
manufacturing value added. They share their profes-
sional work with scientists and other engineers, but
without their contributions the US economy would

The U.S. Chemical Processing Industries, 1984


Stone, Clay
Primary Metal
Total Manufacturing
Chem. & Pet. as %
CPI as %






Chemicals and Allied Products
Petroleum, Coal, Gases
Primary Metals
Process Equipment
Paper and Forest Products
Stone, Clay, Glass
Other Manufacturing

Engineering Design, Construction, Consulting
Nuclear Energy
Public Utility



The word "paradigm" was coined
by Thomas Kuhn to mean a constellation
of characteristics that set a profession apart
from other professions.

To get a clearer picture of the current employment
of chemical engineers, let us turn to a study done by
the AIChE in 1986 and shown in Table 2. Three quar-
ters of AIChE members are employed in manufactur-
ing, and nearly half of the members are employed in
the chemical and petroleum industries. Electronics is
the fastest growing field, outstriping the mighty food
industry. One quarter of the members are in services.
It may be a surprise that government employment is
greater than education.
The National Research Council study on "Chemical
Engineering Frontiers: Research Needs and Oppor-
tunities," chaired by Professor Neal R. Amundson of
the University of Houston, became available in
November 1987. It declared that there are four major
areas of opportunity:

(1) The development of new high tech industries that are driven
by scientific breakthroughs
electronic, photonic and recording materials and devices
advanced materials
(2) The rejuvenation of traditional technologies
raw materials
(3) The protection of health, safety, and environment
(4) New engineering sciences, concepts, and tools, such as
computers, artificial intelligence
surfaces and interfaces

In response to these new opportunities, chemical
engineers need to revise their bag of tools and know-
ledge. The word "paradigm" was coined by Thomas
Kuhn to mean a constellation of characteristics that
set a profession apart from other professions, such as:

a characteristic set of problems that the profession deals
systematic knowledge and methods that are effective in
solving these problems
75% 0 a set of classical cases of successful solutions that are ad-
mired and studied
stable textbooks, curriculum and accreditation, handbooks
exchange of ideas through meetings and journals
professional societies for the members to get together to
govern themselves and to set common goals

Source: AIChE Economic Survey 1986
Note: "Chemicals" include: industrial chemicals, agricultural chemi-
cals, paints, petrochemicals, pharmaceuticals, soaps, plastics.

When the first chemical engineering curriculum in
the US was established by Lewis M. Norton at MIT
in 1888, the curriculum consisted of little but indus-
trial chemistry courses that were specific to each sec-


Value Added = Wages and salaries of employees, plus interest and div-
idends to capital, plus depreciation, plus corporate in-
come taxes (but does not include purchases of goods and
services from other firms)
Source: U.S. Statistical Abstract, 1987.

Current Employment of AIChE Members

-- *U /

tor, such as soap, turpentine, and salt. The first chem-
ical engineering paradigm was unit operations, which
was embodied in the texbook Principles of Chemical
Engineering by Walker, Lewis, and McAdams and
published in 1923. It declared that any chemical pro-
cess can be analyzed as a collection of unit operations
such as distillation and heat transfer. Thus the proper
study of chemical engineers is the study of these indi-
vidual units, which can be assembled to form any pro-
cess. This fiat turned chemical engineering into a sys-
temic profession where the students have powerful
tools to solve manufacturing problems. It also codified
how knowledge should be organized, and how re-
search should be conducted, and of course, how
courses should be taught.
Unit operations provided many valuable research
problems for chemical engineers for four decades,
while rapid advances in knowledge and tools helped in
the birth of the modern petroleum refining and the
commodity chemical industries. Many chemical en-
gineers became so successful in solving crucial prob-
lems in their industries that they were elected to be
presidents and chairmen of the mightiest companies
in the world. The gradual exhaustion of important and
scientifically accessible problems after the second
World War did not mean that unit operations could
not be revived by the application of new and more
powerful tools. The handling of solids and granular
material remain a disgrace in our inability to predict
and to scale-up, and separation is enjoying a re-
surgence once again.
The second chemical engineering paradigm of
transport phenomena arrived in 1960 in a textbook by
Bird, Stewart and Lightfoot. This approach declared
that the proper study of chemical engineers is the
molecular phenomena that are fundamental to the un-
derstanding and performance of chemical equipment.
The kinetic theory of gases and Newtonian fluid me-
chanics particularly play central roles in this ap-
proach. Once again, many chemical engineering re-
searchers are fully occupied in solving problems in this
arena. This approach is so successful that we often
look upon prior achievements as largely empirical and
lacking in mechanistic understanding.
We are ready now for the birth of the third chem-
ical engineering paradigm, a new way to state what
are the most important problems that we should
study, and what tools should be used to study them.
We deal with problems and solve them at three vastly
different scales:

(1) The microscale of molecular and aggregate level. Subjects
in this scale include thermodynamics, transport phenomena
and kinetics. These subjects describe the most fundamental

chemical properties of matter, and they form the core of the
scientific basis of chemical engineering.
(2) The mesoscale of process equipment. We have courses in
unit operations and in reaction engineering. In this scale,
we synthesize many elements from the microscale to design
and operate process equipment efficiently and reliably.
(3) The macroscale of plants and systems. Here we are con-
cerned with plant design and economics, product properties
and market needs, safety and environment, productivity,
and world competition. These are very important topics, but
so far we have few courses that will give students a head
start in dealing with them.

Our colleagues, the chemists, deal mainly in the
microscale, and the mechanical engineers deal mainly
with the mesoscale. But practicing chemical engineers
must deal with all three levels. The first paradigm of
unit operations is in the mesoscale, and the second
paradigm of transport phenomena is in the microscale.
We need seminal courses in the macroscale, so that
our students will have competitive advantages in deal-
ing with the bigger picture. A new paradigm is wait-
ing to be born.


As a consequence of new trends, the problems that
face future chemical engineers will be different from
today's problems. In addition to the classical problems
of commodity chemical manufacturing, they will face
a host of new problems. We must revise the cur-
riculum so that they will have the necessary concepts
and tools. I have listed the enduring and new problem

Problem Areas for Chemical Engineers


* Small Molecules, Gases and
Homogeneous Liquids
* Inorganic and Organic
* Inexpensive and Large Volume
Commodity Chemicals
* Undifferentiated Products with
Long Life Cycles
* Dedicated Plants with
Continuous Processing
* Innovations Dominated by
Process Improvements to Save
* Mesoscale of Process

* Science and Technology


* Large Molecules, Complex
Liquids and Solids
* Biochemistry, Material Science,
Condensed State Physics
* High Value Added and Small
Volume Specialty Chemicals
* Proprietary and Differentiated
Products with Short Life Cycles
* Flexible Plants with Batch
* Innovations Dominated by
Product Development for
Improved Product Performance
* Microscale of Molecular
Aggregates, Macroscale of Plant
Productivity, World Competition,
Safety and Environment
* Environmental and Society
Concerns, Ethics, Humanities


areas in Table 3. What is important in petroleum refin-
ing and petrochemicals will still be important in the
decades ahead, but we must not neglect the new
emerging problem areas.
How do we deal with all these new forces? We may
revise existing courses, we may develop new courses,
and we may change the degree program. Much of the
reform can be accomplished by changing the content
of the existing courses. In the traditional courses, we
illustrate the power of chemical engineering concepts
by solving manufacturing problems. In the 1923 text
of Principles of Chemical Engineering, the examples
are largely drawn from the utilization of coal as raw
material. The textbooks in the 60's and 70's are heavy
with the solution of manufacturing problems in pet-
roleum refining and in commodity petrochemicals.
These chemical engineering principles can also be il-
lustrated by manufacturing problems from the emerg-
ing technologies. There are several efforts underway
to develop such course materials for the classroom use
of teachers who are familiar with the principles but
not with the applications in these new industries. A
set of suggested directions of existing courses is listed
in Table 4.
We also need new courses to broaden the horizons

Direction of Existing Courses


Small Molecules
Gases and Simple Liq
Kinetic Theory of Gas

Gaseous Diffusion
Knudsen Diffusion
Newtonian Fluid Mec

Homogeneous Gas Re
Catalytic Reactions



Large Molecules
uids Complex Liquids and Solids
es Statistical Mechanics
Configurational Diffusion
Liquid and Solid Diffusion
hanics Non-Newtonian and Polymeric
Fluid Mechanics
actions Solid-Fluid Reactions
Biochemical Reactions

Distillation of Oil Protein Separation
Ultra High Purity
Dedicated Continuous Reactors Flexible Batch Reactors
Oil and Petrochemicals Polymers, Biotechnology,
Maximize Yield and Product Minimize Hazards and Harmful
Selectivity Emissions

Economic Efficiency

Product Quality
Safety, Environment

of our students. It would be desirable to have senior
electives in the emerging technologies for students
who want to go in these directions, such as in biotech-
nology, microelectronics, and advanced materials. It
is increasingly difficult for chemical engineers to solve
manufacturing problems if they are not familiar with
surface and colloid chemistry and with polymer sci-
ence and engineering. So many of our problems are
concerned with surface and interface forces, with com-
plex fluids stabilized by surfactants or polymers, with
emulsions and micelles, and with the properties of
polymers, that one is tempted to say that such courses
should be required.
We also need courses in the macroscale, so that
students will be familiar with the "Big Picture" of how
chemical engineers can fill the needs and expectations
of society. The ultimate justification of chemical en-
gineers is their contribution to society: they can
develop products that society wants to buy at desira-
ble quality and price, they can improve productivity
and help their companies compete in the world, they
can maintain high professional ethics, and they can
protect fellow workers and the public from toxic sub-
stances and accidents. An increase in the humanities
and social sciences would sensitize the students to the
needs of the world.
One of the most urgent tasks is the development
of courses in "Product Engineering." Alone among the
engineering fields, the heart of chemical engineering
education is concerned with processing, and we wait
for someone else to develop a product and for the sum-
mons to scale-up the process. A product must have
customers, and marketing is the tool to find out what
customers want. It is the task of product engineers to
design or to discover products with the desired
characteristics. In other disciplines, aeronautical en-
gineers design airplanes and space ships, electrical en-
gineers design transformers and computers, civil en-
gineers design bridges and structures, and mechanical
engineers design compressors. In recent years, many
chemical engineers have become increasingly involved
in the design of catalysts, polymers, and numerous
other products, but this has not yet been crystallized
in one or more seminal courses in product design. As
product cycles become shorter, many products be-
come obsolete so quickly that a next generation pro-
cess is not needed. Product engineering may be where
future action will be, and our students need the train-
ing to make them effective in this important arena.
One may argue that there is little in common be-
tween the product chemistry and engineering of lub-
ricating oil and cosmetic soap, so that one is reduced
to teaching industrial chemistry in the old boring way.


The key to product engineering must be a set of prin-
ciples that would be common to all product develop-
ment and design. Perhaps the principles involve the
relation between molecular and aggregate structure
on one hand, and product properties and performance
on the other hand, as well as how processing affects
these structures and properties.

It is easy to agree that chemical engineers will
have to know much more to solve problems in the
future. How can we squeeze more content into the
same four-year bachelor's degree program? In fact,
the four-year BS is fast becoming fiction on many cam-
puses. It is not at all uncommon, in going over trans-
cripts, to find that the average number of years from
freshman to graduation is close to five years in many
colleges. The schedule is so rigid that anyone who
wants to have a little fun in college, to join the ROTC,
to play in the band, or to change majors, will have to
pay the price of five years.
We have the traditional assumption that the first
professional degree is the four-year BS. Students
learn the fundamentals in universities in four years,
then work in industry as apprentices to experienced
senior engineers for a few years to gain the necessary
knowledge to become full-fledged independent en-
gineers. This assumption is also running into prob-
lems, as traditional industries are reducing senior
staff and emerging industries do not have experienced
senior staff to teach the new graduates. We count on
about a quarter of our BS students going to graduate
school for advanced training, but a shrinking percent
of top U. S. students are choosing this path and we
are becoming heavily dependent on immigrants for
teaching and research. Are we following in the path
of the Boston Symphony Orchestra, where an Amer-
ican-born is an oddity? The graduate schools of busi-
ness, law, and medicine are taking an increasingly
heavy toll from the top of our graduating class. Is this
due to the two track system in industry where the
fast track leads to management and the slow track
leads to senior engineers? How can we deal with these
Let me propose a realistic scheme which may be
the basis for a new consensus. The first "professional"
degree should remain the four-year BS which provides
general education suitable for a variety of career
paths. Without further specialization and schooling,
such a degree would be sufficient for light technical
work such as marketing, personnel, administration,
production, and planning. To be effective in heavy

technical work such as design, process development,
and construction, the graduate needs either an MS
degree or considerable apprenticeship with senior en-
gineers and continuing education. For a career in
teaching and research, a doctoral degree would be
needed. Let us give up the pretense that we can
adequately pack enough education into four years
(given the high school graduates that we get) to train
a student for heavy technical work immediately after
the BS degree. Let us consider the four-year BS as
general preparation for a number of exciting careers
and further education in chemical engineering. []

n book reviews

by D. M. Ruthven
Wiley-Interscience, Somerset, NJ 08873; 1984.
433 pages, $49.50
Reviewed by
Ralph T. Yang
State University of New York, Buffalo
Adsorption has become a key separations tool and
an important unit operation in the chemical industry.
Yet, with the exception of a very brief (but excellent)
coverage on adsorption in King's Separation Proces-
ses and in the recent new edition of Unit Operations
of Chemical Engineering by McCabe, Smith, and
Harriott, this important process has been largely
overlooked in our undergraduate texts and cur-
riculum. Ruthven's book is a timely addition to the
chemical engineering library and, hopefully, will
stimulate interest in both teaching and research on
this subject.
Many books exist on the physical chemistry of ad-
sorption. This is the first book bridging the gap be-
tween chemistry of adsorption and its engineering ap-
plications in separation and purification. It is not sur-
prising that Ruthven wrote this book because he is
one of very few who has done research on all aspects
of adsorption, ranging from thermodynamics and
transport processes to adsorber design and cyclic ad-
sorption processes.
The book gives a complete (although not even)
coverage of all aspects of adsorption: adsorbents, sor-
bent characterization, thermodynamics and energetic
of adsorption, pore diffusion, transport processes in
fixed-bed adsorbers, adsorber dynamics and cyclic ad-
sorption processes. The author speaks with authority


in the areas in which he has done important research:
simplified statistical thermodynamic model for mixed-
gas adsorption, diffusion in zeolite crystals, adsorber
dynamics and the Sorbex system. The two chapters
on adsorber dynamics are probably the most readable
and informative single review available. However, as
stated in the preface of the book, emphasis is placed
on the subjects with which the author is most familiar.
Thus, zeolite is emphasized throughout the book, on
all above-mentioned subjects. In a few instances, the
author's opinions are a bit strong. A most notable
example is in the use of Darken's relation to interpret
the dependence of diffusivity in zeolite on sorbate con-
Areas that deserve more coverage are: other sor-
bents, especially activated carbon, mixed-gas adsorp-
tion and cyclic adsorption processes. The potential
theory for mixed-gas adsorption, which is important
for carbon, is not covered. Cyclic processes, especially
the pressure swing adsorption cycles, can use more
coverage. After all, it is the pressure swing adsorption
process that has promoted adsorption to a key separa-
tions tool in industry, and in which many future inno-
vations are likely to be made.
On balance, this is a superbly written book. It can
be used as an introduction or as a textbook. It also
goes into enough depth for use as an excellent refer-
ence by researchers as well as by adsorber users and
designers. FO

Edited by C. E. Grant and Patrick Pagni
Hemisphere Publishing Corp., 79 Madison Ave.,
New York, 10016; 1226 pages $135 (1986)
Reviewed by
G. K. Patterson
University of Arizona
Fire has been used by man as a tool from the begin-
nings of his rise as the dominant species on earth. The
power of fire, however, causes it to become a source
of destruction and accidental death when it or other
concentrated sources of energy are misused. The
study of accidental fire in enclosed spaces where
people live and work is an area rapidly becoming a
distinct discipline because of its great economic impor-
tance. The science of safety from accidental fires
draws together investigators from many different dis-
ciplines, mainly mechanical engineering and structural
engineering, and to a lesser degree architecture,
mathematics, chemistry, and chemical engineering.
Fire Safety Science is the collection of papers pre-

Scented at the First International Symposium on Fire
Safety Science. It also records the formation of the
International Association for Fire Safety Science.
The book is very well edited and produced. Al-
though camera-ready manuscripts produced on vari-
ous typewriters and word processors were used, the
print quality in all but one case (too small and dark)
was very good, making the book easily readable. The
editors exercised good control over manuscripts from
authors in non-English speaking countries, because all
of them seem to be in very good, highly-understand-
able English. The various sections of the book are well
segmented and logically placed.
The book is a highly comprehensive collection of
topics on accidental fire, ranging from basic theory of
fire physics and chemistry through detection, risk
analysis, and smoke toxicity to fire suppression. The
book is a good blend of theoretical and experimental
research topics and practical, design-oriented papers.
There are ten main divisions in the book: Fire
Physics, Structural Behavior, Fire Chemistry,
People-Fire Interactions, Translation of Research into
Practice, Detection, Specialized Fire Problems,
Statistics Risk and System Analysis, Smoke Toxicity
and Toxic Hazard, and Suppression. Although there
is a section specifically labeled Smoke Toxicity and
Toxic Hazard, there is substantial coverage of this
critical hazard of fires in several other sections as well.
There seems to be a good balance between fire physics
and smoke physics.
The section on Fire Physics is written such that
chemical engineers with good training in fluid
dynamics and turbulent flow simulation can under-
stand it. The sections on Fire Chemistry, Detection,
Specialized Fire Problems, and Smoke Toxicity and
Toxic Hazard are heavy in chemistry and are thus
easily accessible to chemical engineers and chemists.
Sections on Structural Behavior, People-Fire Interac-
tions, Statistics Risk and System Analysis, and Sup-
pression, tend to be much more specialized but under-
standable to any well-trained engineer.
Interestingly, even though the fundamentals of
fire physics and chemistry represent just another
class of chemical reaction engineering, very few of the
paper authors were clearly chemists or chemical en-
gineers. Fire physics and chemistry is an area in
which chemists and chemical engineers could probably
make rapid contributions both in research and applica-
tion to fire detection and suppression programs.
Hopefully, this well-edited, comprehensive book on all
aspects of fire research and application will lead some
chemical engineers to contribute to this very impor-
tant field. O


m P classroom



Ben Gurion University of the Negev
Beer Sheva, 84120 Israel
University of Connecticut
Storrs, CT 06268

T HERE IS GENERAL agreement among engineering
educators that the use of computers in education
should result in the students spending more time on
problem definition/analysis/interpretation of results
and progressively less time on the technical details of
problem solution. The role of the the computer in en-
gineering education has been well articulated in a
study at the University of Texas [6] which concludes
the following:

Course and teaching changes should result in more effec-
tive use of student time. Time-consuming manual calcula-
tions should be supplanted by efficient software designed for
ease of learning and rapid development of ideas.
The computer should become the dominant calculational
tool early in the curriculum so that the ... problems as-
signed can be more realistic without requiring excessive stu-
dent time outside of the classroom.
Major emphasis should be on how to analyze and solve
problems with existing software including that for simula-
tion to evaluate and check such software with thoroughness
and precision.

While the goals and objectives of appropriate edu-
cational computing are often discussed, very few de-
tails have been reported about practical implementa-
tion of problem solving with personal computers
within engineering education. This paper presents
three specific examples of chemical engineering prob-
lems which can be interactively solved on personal
computers with commercially available software. Our
purpose is to illustrate the potential impact of interac-
tive problem solving on chemical engineering educa-
tion by discussing how this has been accomplished at
Ben Gurion University (BGU) and the University of
Connecticut (UConn).

Mordechai Shacham is an associate professor and head of the
chemical engineering department at the Ben Gurion University of the
Negev in Beer-Sheva, Israel. He has a DSc in chemical engineering
from the Technion, Israel Institute of Technology. In addition to his
interest in computer based instruction, he is involved in research in
chemical process simulation, design and synthesis and numerical
methods. (L)
Michael B. Cutlip is professor and head of the chemical engineering
department at the University of Connecticut. He has a PhD in chemical
engineering from the University of Colorado. He has additional in-
terests in the use of personal computers in engineering education and
computer-based instruction. He is active in research in chemical and
electrochemical reaction engineering. (R)

The general numerical analysis package which has
been utilized is commercially available from Control
Data Corporation under the name of POLYMATH.
The choice of this package was not entirely unbiased;
the authors of this paper also designed and developed
POLYMATH, and thus it was tailored for general
educational use with particular application to chemical
engineering. The package runs on an IBM PC/XT and
most compatibles, and a minimum of 128K of memory
and a color graphics board are required.
The package currently consists of the following
0 Copyright ChE Division ASEE 1988


Matrix Manipulator evaluates matrix expressions containing
all basic operations, solves linear equations, and determines
Nonlinear Equation Solver handles up to twelve simultaneous
linear and nonlinear algebraic equations.
Differential Equations Simulator solves up to twelve simul-
taneous first-order ordinary differential equations.
Curve Fitting Program fits polynomials up to the 5th order
and cubic splines directly to entered data of y versus x. User-
provided functions can be invoked to transform the data before
Multiple Regression Program fits mathematical models con-
taining up to five linear parameters to data consisting of one
dependent and up to four independent variables.

These programs are equation and data oriented in
that all the equations can be typed in their regular
algebraic form with the user's own notation. The best
solution algorithms are automatically determined, and
solutions are presented in numerical, tabular, or
graphical form. Matrices are entered and edited in a
format similar to that of spread sheets. All of these
programs except the Matrix Manipulator have been
described in considerable detail [4].
There are other programs/packages which are
commercially available and have some of the
capabilities described above. Notable examples are
TK!Solver for solving algebraic equations and TUT-
SIM for dynamic simulations. POLYMATH has the
following benefits that are valued relative to other
programs and packages:

It is specifically tailored to engineering education.
It is menu driven for easy use by students and faculty.
It provides all important numerical analysis capabilities
for undergraduate engineering coursework.

Solution of More Realistic Problems
The detailed modeling of chemical processes often
leads to systems of nonlinear algebraic or differential
equations. The current practice is to modify these
equations by manipulation or simplification so that an
analytical solution can be found. Often the simplified
models no longer accurately represent the original
model. Use of the calculational package often requires
no simplifications to the original model, and this allows
for a more accurate solution which retains a clear cor-
respondence between the model and the physical/
chemical phenomena of the process.
As Example 1 of the calculational package use,
consider a typical packed bed reactor design problem
found in a textbook by Fogler [1]. This problem con-

This paper presents three specific examples
of chemical engineering problems which can be
interactively solved on personal computers
with commercially available software.

siders an isothermal plug flow reactor for the hydro-
dealkylation of mesitylene over a Detol catalyst where
the reactions are

M + H + X + CH
X + H2 T + CH

Reaction 1
Reaction 2

where M is the reactant mesitylene, X is the desired
metaxylene, and T is the undesired toluene. The pro-
duction of the metaxylene is to be maximized. The
respective rate expressions for the above reactions

rT = k XC 0.5
r k C 0l- 5

The differential equations (material balances) for this
plug flow reactor are

dCM/dV = rM/v

dCH/dV = (rM r)/v
dCx/dV = (-rM rT)/v

This system of equations and the numerical con-
stants can be easily entered into the Differential
Equation Simulator program as shown in Figure 1.
The computer solution is accomplished in less than a

The equations:
d (c)/d (vol):rm/v
d (cx) (-r*-rt)/u
k2=39, 16
Initial values: voyl 8.8, cr: 8.0186, chgo 8.8213, cxg 8.8
Final value: volf= 158. 9

Vi-F7 to solve the problem.
*f- 4J to alter he problem. F6 for helpful information,
S-F8 new problem or library, ft -F16 for the MIII MEIU.

FIGURE 1. Equations and numerical constants for Exam-
ple 1.



1 CM
2 ch
3 ex


.0 6i.99 9 .9el2aiee5.ei5

t. Results in TABULAR form, 9. Results in GRAPHICAL for,
o. Other variable(s). p. Partial Results table.
-F8 new problem or library to make chan es.

FIGURE 2. Concentrations of mesitylene (cm), hydrogen
(ch) and metaxylene (cx) in reactor for Example 1.

minute, and the graphical output of the concentration
profiles is presented in Figure 2. Thus, in this example
the problem solution is efficiently obtained from the
equations which describe the reactor. This is in con-
trast to the transformation of variables (typically to
conversions) and the analytical solution approach
which cannot lead to the concentration profiles in this
case as discussed by Fogler [1]. Consideration of tem-
perature effect is also easily accomplished with the
differential equation simulator while this complication
usually makes the traditional analytical approach un-

Illustration of Modern Process Design Techniques

Modern process design utilizes complicated tech-
niques such as multistage, multicomponent distillation


Feed r z Flash
.R. ..ER+EL kl=2.5258
7r k2=1.56998
F- 309 lb/hr k3=0.0329
[391.9 .19 0 [03030 0
ER= 398.8 E2= 0.5 0 = 0.2943 0.4529 0
L209.3 0 0. 52] 0.-1526 0.2971 0.75]

z = Estimated value of the tear streams
r = Calculated value of the tear streams
Molecular weigth for all of the components is 90 lb/mole.

FIGURE 3. Simplified flowsheet for a chemical process,
from Henley [2].

calculations and multiunit, multicomponent, nonlinear
recycle calculations. It is beneficial for students to
write computer programs which utilize these tech-
niques, but this is very time consuming. Also, it is
possible to give them a computer program which will
provide a "black box" solution. The computational
package allows an effective solution to this educational
dilemma by allowing the efficient execution of the
steps of a design algorithm, one by one. This is illus-
trated in Example 2 in which the methods for acceler-
ation of convergence in a recycle stream are com-
Example 2 is taken from Henley [2] in which a
chemical process consists of two reactors, three simple
separators, and one isothermal flash unit. This process
contains several recycle streams, and the optimal cal-
culation sequence requires iteration on a single (tear)
stream. In this example, different solution methods
(Quasi-Newton, Newton-Raphson, Direct Iteration,
etc.) are to be compared.
In this system, all of the process units are linear
except the isothermal flash. Thus all the calculations
involving the "linear" units can be combined by using
the "transformation matrix" approach discussed by
Shacham [5]. The simplified process is shown in Fi-
gure 3. The calculation of the recycle system involves
estimation of the flow rates in the tear stream (z),
solution of the isothermal flash problem for the vapor
(V) and liquid (L) flowrates, and calculation of the
values for the tear stream from

r = FR + EV + E2L

Iterations are continued according to the particular
solution method until convergence.
For illustration, the Quasi-Newton method of solu-
tion requires updating the H matrix which is an ap-
proximation to the inverse of the matrix of partial
derivatives. Therefore

H1 =Ho -[(Hy P )PTH]/Po TH yo (6
P0 =Z Zo

Yo = f(z) f(Zo)
f(z) = z r

and the subscript represents the iteration number.
The solution of this example requires the Matrix
Manipulator and the Nonlinear Equations Solver
programs from the computational package. The calcu-
lations for H, are shown in Figure 4 where the compli-
cated matrix operations can be entered into the Ma-


1 :21(3x1)
2: 1 (3x1i
4: 2(3x1)
6: F8(3xl
7:FI (3xl)

) H- (H8xYB-PO)xtran (PO)xH8/(tran (PS)xHOxYB)

Enter your algebraic expression at the arrow. Use the +J key
to have the expression computed, Use the F3, F5, t A -- keys
to edit the expression, Press F6 for a discussion of the
operators and functions available at the arrow.

FIGURE 4. Defined matrices and desired matrix expres-
sion for Example 2.

trix Manipulator in a form almost identical to Eq. (6).
The corresponding nonlinear equations for this exam-
ple are shown in Figure 5 as they have been entered
into the Nonlinear Equations Solver.
Thus a comparison of the several methods for this
recycle stream calculation can be obtained in a period
of several hours of a student's time with the use of the
calculational package. This provides a valuable educa-
tional experience for the student without the burden
of writing/debugging a source code computer program
which would take significantly more student time.
This approach requires that the student have a funda-
mental appreciation of the various methods available.

Analysis and Simulation of Laboratory Experiments

The educational value of laboratory experimenta-
tion can be significantly enhanced by efficient data
reduction and simulation of the experiment with the
computational package. In such an exercise, the stu-
dent must come prepared to 1) conduct the experi-
ment, 2) treat the data to develop a mathematical
model and possibly evaluate model parameters, and 3)

Data for Example 3 (from Hill [3])

Data Point


Total Pressure

The equations:
f a)=-1. 5258xzl/ (sunx (1+axl. 5258))-. 57xz2/ (suNx (l+a+.57))+b
f (z3)=8.1526xv1+. 2971xv2+8.75xv3+0.52x13+289.3-z3
b=z34. 9671/(sunx (-a x, 9671))
11 (l-a)xzl/(1+1.5258xa)
12= (1-a)xz2/(l+.57xa)
13= (-a)xz3/(1-8,9671xa)
Initial values: a= 8.5890, zlg= 392.88, z2g- 298.98
z30- 289.88

A. F7 to solve this problem.
*t-4-J to alter the problem 40 F8 new problem or library.
F18 for the M~IN HENU. F6 for helpful information

FIGURE 5. Nonlinear equations utilized for Example 2.

simulate the experiment to verify the analysis. The
computational package is typically useful in evaluating
models and determining model parameter values, and
additionally it is quite helpful in simulating the exper-
iment to verify model suitability.
Example 3 uses the "student" data shown in Table
1 for the gas phase dimerization of trifluorochloro-
ethylene as reported by Hill [3]. This reaction is sim-
ply written as 2A -> C, which occurs in a constant
volume batch reactor at 440 C. The reactor pressure
change was measured as shown in Table 1. In this
example, the reaction rate coefficient and overall
order of the reaction are to be determined, and the
experiment is to be simulated to confirm model valid-
The differential equation in terms of conversion
for this batch reactor has the form

dXA/ dt= kCA0 (a-) (1 XA)a

where the reaction rate constant, k, and the order of
reaction, a, are to be determined from the data of
Table 1. The relationship between conversion and
total pressure is given by

XA = 2(1.0 P/Po)

where Po is the initial pressure, 101.32 kP.
The differential analysis approach for this example
requires that Eq. (8) be linearized by taking the
logarithm of each side of the equation. This results in

PLn(dXA/dt) = (n(kCA0a-1) + a kn(1 XA) (10)

In order to use this linear relationship to deter-
mine k and a, it is first necessary to obtain dXA/dt
Continued on page 34.


WM laboratory



Illinois Institute of Technology
Chicago, IL 60616

AN UNDERGRADUATE experiment involving data
acquisition by a computer, digital signal trans-
mission from the computer to a digital logic circuit
and signal interpretation by this circuit is being used
at the Illinois Institute of Technology (IIT) to illus-
trate alarm system design and implementation. The
signal generated by an analog source is read by the
computer and interpreted in a FORTRAN program
using prespecified alarm levels. Based on the result of
this assessment, digital signals are sent from the com-
puter to a digital logic circuit which interprets the
signal and activates the appropriate alarm lights.
Before starting the experiment, elementary con-
cepts in combinational logic design, as well as funda-
mentals of analog and digital signals and analog to
digital conversion, are presented. The students are
responsible for the design and construction of the logic

7420 --

7400 2-input NAND
7402 2-input NOR
7404 NOT
7420 4-input NAND

7402 alar

FIGURE 1. Logic diagram of alarm system.

'Current address: Dow Chemical Company, Midland, Michigan

The design and construction of the circuit and
the implementation of the alarm system is completed
in a three-hour session. Besides illustrating alarm
system design and implementation, the experiment
introduces analog to digital conversion
and digital signal transmission . .

circuit. After completing the construction of the cir-
cuit and debugging it, they set various alarm levels
and test the alarm system using an adjustable analog
source. The design and construction of the circuit and
the implementation of the alarm system is completed
in a three-hour session. Besides illustrating alarm sys-
tem design and implementation, the experiment intro-
duces analog to digital conversion and digital signal
transmission which are crucial in understanding real-
time computer operation and in interfacing the com-
puter to a process.
The experiment is used in the reaction engineering
and process control laboratory course. The course is
a senior level technical elective with two lectures and
one three-hour laboratory per week. Digital logic de-
sign concepts and fundamental issues in process com-
puter interface are presented in lectures before begin-
ning the experiments. The course is offered in the
same semester with the second process design course.
Since process safety and reliability assessment are
presented in the design course, this experiment pro-
vides a good illustration of process safety concepts to
interested students.

Digital System Design
The first step in designing a digital logic system is
to understand the task and the properties of the sys-
tem that will be designed. This includes identifying
the inputs and outputs of the system and making a
thorough description of the logic circuit (an alarm sys-
tem for a 'process' in this case). Next, the description
of the circuit is reformulated using logical 'connec-
tives' [1, 2] such as AND, OR, NOT. Truth tables for
0 Copyright ChE Division ASEE 1988


these connectives and the symbolic representations of
the corresponding logic gates are given in the Appen-
dix. Although the reformulation is nothing more than
expressing the descriptive information in another
form, this new form makes the design of the logic
diagram straightforward. Then the logic diagram
(Figure 1) is designed and all inputs and outputs are
noted on it. The next step is the drawing of the circuit
diagram (Figure 2). The summary layout of various
electronic chips that contain the logic gates are used
in the circuit diagram. The circuit diagram is com-
pleted by assigning the connectives in the logic dia-
gram to the gates in the circuit diagram. The mapping
of each connective is indicated both on the logic dia-
gram and on the circuit diagram by writing down the
assignments made. This is crucial for minimizing the
time and effort that will be spent for debugging the
When the circuit diagram is completed the circuit
is constructed. For the construction of prototypes it
is much better to use breadboards which permit the
alteration of the connections merely by unplugging
and inserting the wire lead to another location. This
minimizes labor while testing the circuit.

Analog/Digital Conversion
For data collection from a process, computers are
interfaced to chemical processes through analog to di-
gital (A/D) conversion systems [1]. Signals sent by
transducers which carry information generated by the
sensors are in general in 'analog' form. The informa-
tion is expressed by the magnitude of the signal. Then
the 'analog' signal is converted to a 'digital' signal
using the A/D converter and is sent to the computer.

Robert A. Martini is a re-
search engineer with Dow
Chemical in the Computer Ser-
vices group of the Michigan
Applied Science & Technology
Laboratories. He received a BS
in chemical engineering from
the Illinois Institute of Technol-
ogy in 1983 and continued to
do graduate work in advanced
process control under Ali Cinar,
completing his Masters in
chemical engineering in 1985.
Cindy Sullivan is currently a
graduate student at the University of Wisconsin, Madison. She has a
BS degree (1987) in chemical engineering from the Illinois Institute of
Technology. (C)
Ali Cinar is an assistant professor of chemical engineering at the
Illinois Institute of Technology. He has a BS degree from Robert College,

FIGURE 2. Circuit diagram of the alarm system.

Most A/D converters accept voltage signals as input.
The span of these input signals is usually 0 to 10, 0 to
5, -5 to + 5, or -10 to + 10 volts. The A/D converters
have resolutions of 8-bit, 12-bit or 16-bit, with the
12-bit being the most common resolution.
The A/D converter systems are available as stand
alone units or plug-in cards that are mounted in the
extra slots of the computer backplane. Often a multi-
plexing system is integrated to the data conversion
system on the card and many input channels are avail-
able to the user on one card.

Istanbul, and a PhD from Texas A&M University. His research interests
are in control of tubular autothermal chemical reactors and in stabili-
zation and yield improvements in chemical reactors by forced periodic
operation. (R)


Digital Input/Output
Process information which has only two possible
states, or command signals with two possible states,
can be communicated in digital form [1]. The mag-
nitude of a voltage signal is divided into two ranges,
and each range is assigned one state. Traditionally 0-5
V signals are used; 0-0.8 V range corresponds to one
of the states and 2.4-5 V corresponds to the other
state. A signal to change the state of an on/off valve
is an example of a digital signal. Digital input/output
(I/O) cards are used for transmitting digital informa-
tion between the process and the computer.

Real-time Operation
Since computerized data acquisition and control
has as one of its aims the automation of data collection
and process control, an interval timer [1] which acti-
vates the data acquisition procedure at prescribed
times is needed. This timer may be available as a sepa-
rate card or may be mounted on the data communica-
tion card.
Special programs are needed to set the timer, to
coordinate the sampling and conversion of analog
input signals, to transmit output signals, and to carry
out digital I/O operations. Such computer programs
can be developed in lower level languages such as as-
sembly language, or can be purchased from most data
communication card manufacturers for their specific

In the alarm system design experiment, the de-
sign, implementation, and evaluation of a process
alarm system is sought. As a typical process a fluid
storage tank is considered. If the tank is holding fluid
which is fed to a pump, as long as the level is less than
a maximum (hence the probability of tank overflow is

Alarm Levels Corresponding to Voltages


0.0 < V < 0.5 2 RED

0.5 < V < 2.5 i YELLOW

2.5 < V < 7.5 0 GREEN

7.5 < V < 9.5 4 YELLOW

9.5 < V < 10.0 8 RED

FIGURE 3. Information flow diagram in the alarm exper-

reduced) or greater than a minimum value (hence the
probability of running the pump dry is reduced) there
is no cause for alarm. It would be desirable to issue a
warning alarm if the level crosses a preselected
minimum or maximum threshold value which indicates
a potential danger. If the level continues to drift in
the dangerous direction, a hazard alarm is issued
when the specified maximum or minimum level is
reached. In this experiment the level measurement
signal is simulated by a variable voltage source. Warn-
ing alarms are indicated by YELLOW and hazard
alarms are indicated by RED.
The alarm system consists of a real-time computer
and a digital circuit. An analog signal generated by
the 'process' is sampled and digitized. The latest value
of the signal is compared to preselected threshold lim-
its and if its magnitude is beyond these limits the cor-
responding alarm signal is issued. The alarm signals
are transmitted to the digital circuit which interprets
these signals and activates the corresponding lights.
The alarm levels corresponding to the signal mag-
nitude are given in Table 1. There are two alarm
levels, indicated by yellow and red. Green corres-
ponds to a signal magnitude inside safety threshold
limits. The signal is generated by an adjustable resis-
tor connected to a 10 V voltage supply and has a range
of 0 to 10 volts. The magnitude of the input voltage is
displayed on a digital voltmeter and on the computer


ALARM : Alarm status
ASAMP Analog sampling rate ( 1 sample / ASAMP sec)
COUNT : Time elapsed since last digital transfer to parallel
line interface
DIGOUT: Digital output to parallel line interface
DITRA : Digital transfer rate to parallel line interface


ALARM ; Alarm status
ASAMP Analog sampling rate ( 1 sample / ASAMP sec)
COUNT : Time elapsed since last digital transfer to parallel
line interface
DIGOUT: Digital output to parallel line interface
DITRA : Digital transfer rate to parallel line interface
AVOLT : Analog voltage (signal from sensor) magnitude

FIGURE 4. Flow chart of the computer program: (a) main
program (b) external control program.

terminal. Four digital outputs are sent from the com-
puter to the circuit through digital I/O lines. Bit 1 cor-
responds to YELLOW LOW, bit 2 to RED LOW, bit
4 to YELLOW HIGH and bit 8 to RED HIGH. A
typical logic diagram designed by students is given in
Figure 1 and the corresponding circuit diagram is
shown in Figure 2. Three process status lights are
connected to the outputs of the circuit.
The information flow diagram (Figure 3) illustrates
the inter-connections among the components of the
alarm system. The data acquisition and signal genera-
tion program (Figure 4) is provided to the students.
The construction and testing of the digital circuit is
the responsibility of the students conducting the ex-
periment. A breadboard with built-in voltage supplies
and lamps, various chips and their diagrams are given
to the students.
A DEC 11/23 microcomputer with a DRV11-J
parallel line interface card, a KWV11-C real-time
clock card and a AXV11-C analog input/output board
with eight A/D and two D/A channels is used. The
computer program is written in FORTRAN and data
transmission programs developed by Data Translation
are used. Implementation of the experiment on other
computers is straightforward.
The alarm system can be implemented without the
digital logic circuit, by using the logical statements in
the computer. In that case the digital outputs can be
used to turn the alarm lights on and off. Alternately,
the computer screen may be used for displaying the
alarm status. Such a display is included in the com-
puter program to enable the students to debug their
circuit more easily.
Both the alarm scenario and the alarm circuit are
kept simple. Further refinements are possible by issu-
ing different alarms for HIGH and LOW deviations.
Also, elimination of redundancies in the circuit may
be requested. In the design shown in Figure 1 for the
yellow alarm, the alarm is on when either bit 1 or bit
4, or both, are on. Physically the magnitude of the
signal from the process can not be less than 2.5 V and
greater than 7.5 V simultaneously. Hence a condition
resulting in having both bit 1 and bit 4 on is not possi-
ble. Inclusion of this constraint in the circuit may be
requested. Since the handling of such "don't care" and
"can't happen" states requires further theoretical
studies, such refinements were not requested from
the students. For this particular case they do not have
an effect on the status of the alarm light.


1. Mellichamp, D. A. (Editor), Real Time Computing with Appli-
Continued on page 47.




Texas Tech University
Lubbock, TX 79409

Input Output
Process --
X1 Yl

IT IS INTERESTING to note that much of the chem-
ical engineering curriculum is devoted to develop-
ing a quantitative description of physico-chemical sys-
tems, yet there is little attention given to the subject
of modeling. There is usually some discussion of
dynamic modeling as an introduction in process con-
trol courses, and some departments offer elective
courses in process modeling; but, in general, the
chemical engineering graduate does not have a good
foundation in the fundamentals of modeling. This
paper is designed to present a framework for model
development that, when used, will help the student
(or professor) avoid the major pitfalls associated with
modeling: i.e., not properly identifying the controlling
factors, lack of model validation, developing a model
that is incompatible with its end use, etc. A model is
defined by The Random House Dictionary as

MOD EL . -n. 1. a standard or example for imitation
or comparison. 2. a representation, generally in minia-
ture, to show the structure or serve as a copy of some-
thing. 3....

With regard to chemical engineering applications,
models are used to approximate certain characteris-

James B. Riggs received his BS (1969) and MS (1972) from the
University of Texas at Austin, and his PhD (1977) from the University
of California, Berkeley. He is a registered professional engineer (State
of Texas) and has over four years of industrial experience. He taught
at West Virginia University for five years and has been at Texas tech
since 1983.

Input Mod



FIGURE 1. Comparison of input/output for a process and
its model. Note that n>m and k>t.

tics of a process given the input to the process (see
Figure 1). Also, remember that a process can range
from an entire oil refinery to a single drop falling
through a gas. A model can never be a "true" or an
exact representation of a process because it would
have to be the same process, or an exact replica, in
order to accomplish that.
A useful model provides reliable information about
a process from the operating conditions of the process
using a "relatively easy" procedure. Models are used
for the following applications:
Process control
Process design (scale-up)
Process optimization
In addition, process models and the development
of process models can lead to an overall understanding
of the process; i.e., an understanding of the complex
interactions within a process.
Models can be categorized into one of the following
Empirical models
Scale models
Analog models
Phenomenological models

0 Copyright ChE Division ASEE 1988


An empirical model assumes the form of the func-
tional relationship between the input and output vari-
ables of a process. Then, using data from the process,
parameters or constants in the functional relationship
are determined. The usefulness of an empirical model
depends upon a judicious choice for the assumed func-
tional relationship and upon whether the model is to
be used outside of the range of data upon which it is
based. That is, empirical models are best when used
in an interpolative manner, but are dangerously unre-
liable when used for extrapolation. The application of
transfer functions in process control represents a com-
monly used empirical model.
Scale models are smaller-scale versions of a system
which is usually designed to study one factor. Scale
models have the same geometric proportions as the
full-scale system but on a smaller scale. A classical
example is a wind tunnel in which aircraft designers
can analyze the drag of a particular aircraft design
using a scale version (model) of the aircraft under spe-
cific conditions. The conditions used are such that the
dimensionless flow equations are the same for both
the scale model and the full-scale aircraft. In this man-
ner, a variety of aircraft designs can be analyzed with-
out having to construct and test the full-scale aircraft.
Other applications of scale models involve flow model-
ing and include reservoir modeling, pilot-scale reac-
tors, small-scale distillation columns, etc.
When a physical system is used to predict the be-
havior of the system of interest, it is referred to as an
analog model. For example, an electrical circuit can
be used as an analog model of a mass-spring-dashpot
system. In this analog, an electrical resistor repre-
sents the resistance to displacement, an inductor rep-
resents the inertia of the mass, and a capacitor repre-
sents the storage of potential energy. In this manner,
the behavior of the electrical analog can be used to
predict the behavior of the mechanical system. This
approach is similar to that used by an analog computer
in which the solution of sets of linear or nearly linear
differential equations is obtained by constructing elec-
trical circuits that are equivalent to the differential
equations and then monitoring the time responses of
the appropriate voltages and currents.
Phenomenological models apply conservation
equations (mass, energy, and momentum) in order to
develop relationships between process input and out-
put variables. In order to apply the conservation equa-
tions, a number of constitutive relations may be re-
quired; e.g., equations of state, chemical and phase
equilibrium relationships, and chemical kinetic ex-
pressions. Phenomenological models range from
microscopic models (distributed parameter models) to

This paper presents a framework for model
development that, when used, will help the student
S. avoid the major pitfalls associated with modeling;
i.e., not properly identifying the controlling
factors, lack of model validation, developing
a model that is incompatible with its end use, . .

macroscopic models (lumped parameter models).
Phenomenological models are the most commonly
used models by chemical engineers, and the remainder
of this discussion is primarily aimed at this approach
to modeling.
The recommended procedure for model develop-
ment is
Define the problem
Identify the controlling factors
Evaluate the data for the problem
Develop a set of model equations
Implement a solution procedure
Validate the model
Define the problem. Here you determine what you
want to predict from what input data. In addition, you
must also determine how the results of the model are
to be used.
Identify the controlling factors. In order to ac-
complish this task, you must develop a physical under-
standing of how the process works; i.e., what factors
control the behavior of the process. This is obviously
one of the most important steps in the model develop-
ment process. It is usually helpful to develop a physi-
cal description of how the process operates. Then, by
reviewing the physical description, you can identify
the controlling factors of the process.
Evaluate the data for the problem. There are two
types of data that you must analyze for a modeling
problem: parameter data and process data. Parame-
ters such as dispersion coefficients, diffusion coeffi-
cients, and thermal conductivities are typically re-
quired in the development of models. The uncertainty
associated with parameters used by a model should
also be estimated because this will have a direct effect
upon overall reliability of the model. In some cases,
you may be unable to obtain independently measured
values for parameters you need. In that case, you
must either use a predictive technique for obtaining
the parameter or you must use process data for that
purpose. When process data is used, you must recog-
nize that you are moving away from a phenomenolog-
ical model toward an empirical model. Even if you do
not use empirically determined parameters, you may


need to estimate the uncertainty associated with the
process data if it is to be used in the model validation
Develop a set of model equations. The first step is
to explicitly define the location and type of system
boundaries. That is, you must determine what the sys-
tem you are modeling is and what are its boundaries.
Then equations can be developed which describe the
key variables within the defined boundaries-these
equations are the model equations.
A major problem with developing the model equa-
tions is to determine what degree of detail you should
include in order to meet your objectives. Usually, the
more detail you include the more complicated the
model equations become and the more difficult they
are to solve. In addition, when more detail is used,
more parameters are required. The flow characteris-
tics of the process usually determine the degree of
detail required. For example, if a vessel is well mixed,
a macroscopic model will usually provide a good ap-
proximation of the process. On the other hand, if the
vessel is not well mixed, a more detailed model would
be required. In cases where the properties of the pro-
cess change spatially throughout the system, a micro-
scopic model is usually required. There are models
that are intermediate to the macroscopic and micro-
scopic models: models based upon plug flow of the fluid
or models based upon plug flow with dispersion. In
fact, you can consider that there is a continuum of
models of the real process ranging from submolecular
to macroscopic, depending upon the simplifying as-
sumptions used.
In choosing the degree of detail to use, you will
make simplifying assumptions that must be verified
during the validation procedure. The simplifying as-
sumptions come from knowledge of the controlling fac-
tors and definition of the problem.
Usually, chemical engineering model equations
come from the application of material, momentum,
and energy balances to the process with some type of
simplifying assumptions associated with the flow be-
havior. In addition, equations of state, chemical and
phase equilibrium, and reaction kinetics are at times
combined with the material and energy balance equa-
tions in the model development process.

Implement the solution procedure. Not only
should the numerical solution procedure reliably solve
the equations to yield the desired output, but it should
do so in accordance with the overall objectives of the
problem. For example, if the model is to be used for
control purposes, the solution procedure must be fast

enough for the control system to respond quickly
enough to control the process. Also, the numerical sol-
ution procedure must be selected to operate with the
storage and software capabilities of the computer that
will be used.
Validate the model. You can never completely val-
idate a model since you can only check your model
with a finite number of tests. That is, just because
your model passes certain tests does not guarantee
that it is correct. Therefore, the trick is to devise a
set of tests that provides the best chance to identify
either logic or application errors within the context of
the end-use of the model. Following is a list of ap-
proaches that are useful in the search for modeling

Verify simplifying assumptions
Check that the general model behavior is in accordance
with the process behavior
Develop analytical solutions for simplified cases and com-
Compare with other models using a common problem
Perform a sensitivity analysis to evaluate the effects of
parameter uncertainty
Compare the model directly with process data

Verification of the simplifying assumptions involves
checking your assumptions using the results of the
model, or perhaps even testing the process to examine
the accuracy of the assumptions used to formulate the
model equations. For example, if you were to assume
that a particular side reaction was insignificant, you
should take the results of your model and determine
the magnitude of the side reaction throughout the pro-
cess. If the reaction rate is sufficiently low, your as-
sumption is acceptable. As another example, consider
the assumption of plug flow through a reactor. This
assumption can be checked by measuring the outlet
concentration profile to an injected slug of tracer for
the actual process.
The model should also be checked against the gen-
eral behavior of the process. For example, if the con-
version in a reactor increases as the feed rate to the
reactor is decreased, the model should show the same
A very useful means of model validation or debug-
ging is to apply your model to a limiting case for which
you can obtain an analytical solution. For example, if
you developed a model for the non-isothermal effec-
tiveness factors in a spherical catalyst particle, it
could be checked against the analytical solution for an
isothermal effectiveness factor for a spherical catalyst


particle by simply using the model with the heat of
reaction set equal to zero. This validation procedure
allows you to check for programming errors and unit
conversion errors, as well as the overall physics of
your model equations.
Similar to the last approach, you can apply your
model to a problem for which another model (well es-
tablished and verified) can also be applied. For exam-
ple, if you had developed a two-dimensional model for
a fixed bed reactor, you could compare it with results
for a one-dimensional model of a fixed bed reactor by
making the appropriate modifications to the input data
for your model.
Next, you should perform a sensitivity study using
your model. That is, you should vary each parameter
you used over its range of uncertainty and observe
the resulting effect upon the model predictions. It is
very important to know if your model is especially
sensitive to one or more parameters.
Finally, you should compare your model with pro-
cess data whenever possible. This is always the best
test of any model. Unfortunately, process data may
not be available; e.g., the process has not yet been
constructed or you may be unable to directly measure
all the output variables of the process. In the latter
case, you should verify your model with process data
that are available.

(The material presented here will appear in An Introduction to
Numerical Methods for Chemical Engineers, by James B. Riggs,
which will be published in early 1988.

1. Aris, R., Mathematical Modeling Techniques, Pittman, 1976.
2. Denn, M. M., Process Modeling, Longman, 1986.
3. Himmelblau, D. M., "Mathematical Modeling," Chapter 2 in
Scaleup of Chemical Processes, Ed.: A. Bisio and R. L. Kabel,
Wiley-Interscience, 1985. O

REVIEW: Engineering Education
Continued from page 11.
covers four further volumes which may be of interest
to those who were stimulated by the initial report.
Engineering in Society is a thoughtful and in-
teresting overview of the evolution of American en-
gineering and its role in society today. The panelists
conclude that the engineering profession has re-
sponded well to changing societal demand, although
they are less confident that the profession can be suf-
ficiently adaptive to rapid changes in the future. They
argue for educational programs that are broad and
balanced, rather than highly specialized. They suggest

that the faculty shortage could be alleviated by a
greater use of educational technology in developing
alternative methods of instruction and they ask that
engineering students be sensitized to the role of en-
gineering in society.
Continuing Education of Engineers focuses on
formal, non-credit education of employed engineers.
The panel reviews patterns of continuing education in
industry, academic and the professional societies. Pro-
fessional societies are assuming an increasing role in
continuing education. Most universities, preoccupied
with undergraduate and graduate education and re-
search, have little interest in continuing education.
The panel examines reasons for engineers participat-
ing (or not participating) in continuing education. The
most startling finding is that while most people con-
sider continuing education a good idea, there is little
scientific evidence that it actually benefits the par-
ticipant. Nevertheless, the panel sees an increasing
need for continuing education of engineers as the pace
of technological change accelerates.
Engineering Employment Characteristics pro-
vides considerable data on the engineering work force
and its utilization. It draws no conclusions and makes
no recommendations. There are few surprises. The
numbers of women and minorities are increasing, al-
though the percentage in the engineering work force
is still woefully small. Japan graduates more en-
gineers than the United States, with a substantially
smaller population. Many of Japan's engineering
graduates enter government service. A statistic of in-
terest to academics is that it takes an engineer who
continued to the PhD about 20 years to catch up in
total earnings with his classmate who went to work
after receiving a BS degree.
Engineering Technology Education is probably of
limited interest to chemical engineers because there
are only a few associate degree programs in chemical
engineering technology and no baccalaureate pro-
grams. This brief report reviews the history and cur-
rent status of engineering technology education and
employment. It compares and contrasts engineering
and engineering technology, as well as industrial
technology and engineering technology. Engineering
technology enrollments have not grown as fast as en-
gineering enrollments, so the issues in technology
education are not so sharply focused. The doctorate
isn't typically required for faculty appointment, and
industrial experience is highly valued. Hence there
aren't the extreme faculty shortages that exist in
some engineering fields. On the other hand, the need
for state-of-the-art laboratory equipment may be
more pronounced in technology programs. D


L 3 curriculum



University of Louisville
Louisville, KY 40292

TO THE GENERAL public, the words "hazard" and
S"chemical" are almost synonymous, particularly in
light of Bhopal and recent incidents in the United
States. Indeed, an ABC television movie, Acceptable
Risk, presented the U.S. chemical industry in an un-
favorable light with respect to public safety. Further-
more, liability costs have increased to such an extent
that the possibility of all insurance for environmental
liability being phased out exists [1]. Yet chemical en-
gineering curricula still generally do not cover the con-
cepts of hazard, emissions control, health and safety,
or risk assessment. For example, how many of our
students are familiar with flammable limits or
threshold limit values (TLVs) by the time they
graduate? [2]
Today engineers as designers and managers have
responsibilities that have not traditionally been consi-
dered as "true engineering." They include providing
worker and community safety and health protection.
Yet, there is still a perception by faculty and students
that other technical specialties are more rewarding
and intellectually appealing, and that safety is of sec-

Marvin Fleischman is professor of chemical engineering at the Uni-
versity of Louisville and a former department chair. He received his
BChE from City College of New York and his MS and PhD from the
University of Cincinnati. He has worked for Monsanto, Exxon, Amoco
and the U.S. Public Health Service. His research interests include waste
management and health effects.

ondary importance to engineers. While health and
safety may not have been mainline chemical engineer-
ing, they are typical issues which graduate chemical
engineers in industry and government regularly en-
counter. The relevance of health and safety to chemi-
cal engineering is demonstrated by the existence in
AIChE of the Safety and Health Division, the En-
vironmental Division, relevant continuing education
courses, and the recently established Center for
Chemical Process Safety (CCPS). Consideration is
also being given to the addition of specific require-
ments on safety in the ABET curricular requirements,
and to a recommendation that questions on these top-
ics be included in the professional engineering exam.
This paper is intended to be a philosophical com-
mentary on the need and rationale for incorporating
safety and health into the chemical engineering cur-
riculum. The how, where, and what are presented as
broad and general guidelines.

Engineering educators are in a position to enhance
the sensitivity of future engineers to societal respon-
sibilities such as safety and health. The undergraduate
program is a logical place to provide engineers with a
background for recognizing potential hazards and pro-
viding safe designs. It is neither necessary nor desir-
able to make all chemical engineering students ex-
perts in safety and health. However, increasing stu-
dent awareness, interest, and knowledge of health and
safety protection is like preventive medicine, and it is
much less grievous and expensive than remedial ac-
Given the changing and apparently diminishing
employment opportunities in chemical engineering, an
increased awareness and knowledge of health, safety,
and environmental concepts could enhance oppor-
tunities for jobs and create new areas for graduate
students and research. If chemical engineering doesn't
fill these gaps, other disciplines will. The argument
can and will be made that the chemical engineering
curriculum is already overcrowded. However, we
0 Copyright ChE Division ASEE 1988


Engineering educators are in a position to enhance the sensitivity of future engineers to
societal responsibilities such as safety and health. The undergraduate program is a logical place to
provide engineers with a background for recognizing potential hazards and providing safe designs.

need to re-examine the importance and priorities of
what we teach, and perhaps reduce the extent to
which certain topics are still covered.
"Some professors might argue that their job is to
provide a fundamental grounding in the principles of
chemical engineering and that many applications can
come when the graduate enters industry. It seems
that many U.S. chemical engineering educators may
believe this, as they do not include safety and loss
prevention in their courses." [3] Even though industry
is sensitized to these issues, not all students or faculty
may be aware of industry's interest.
University representatives from major employers
of chemical engineering graduates have indicated that
their companies would welcome more emphasis on
safety in the undergraduate program. New graduates
are quite surprised at how much time is spent on
safety in industry. Since companies rely on engineers
to build safety into equipment, process, and plant de-
signs, industry may well expect future BS graduates
to have a greater awareness of safety and health.

Due to their education, experience, and skills, en-
gineers are better equipped than most people to
foresee technologically induced problems and to de-
sign and implement appropriate preventive measures.
In particular, many health and safety issues lend
themselves to application of chemical engineering
principles and techniques. Optimal design and selec-
tion of safety devices are based on the same chemical
engineering principles as the process development it-
self. Plant and process design are important compo-
nents of chemical engineering education, and preven-
tion and control of emissions are important factors in
Current research by EPA and NIOSH such as the
assessment of new technology (e.g., semiconductors,
biotechnology) and prediction of hazardous emissions
using a unit processes/unit operations approach (e.g.,
alkylation, drying, etc.), rely heavily on traditional
chemical engineering. The latter approach is an excel-
lent teaching method since students must understand
the system and its components to identify, control,
and model emissions. Participation in an emissions
monitoring survey is also an excellent way to learn
about a particular process.

"The interdisciplinary nature of safety also ex-
poses students to non-engineering concepts and
methods. Students are also made aware of the non-
technical consequences of industrial activities, where
not the machine, but man and the environment are
central." [4]

A number of different approaches to bringing
safety and health into the undergraduate engineering
curriculum are possible. They include

Integration into existing engineering courses
Elective courses and/or options, either interdisciplinary or
in the major discipline
Strong emphasis in the capstone design project
Plant visits that emphasize engineering controls and other
aspects of safety and health
Student co-op and summer positions and research projects
directly related to safety and health
Use of qualified guest speakers from industry, NIOSH,
etc., for seminars, short courses, or to assist in classroom

A logical place to emphasize occupational health,
safety, and environment is in the process design
courses and/or elective courses. However, it is ques-
tionable whether or not sufficient concepts and
methods can be taught and learned in one or two
courses. Integration of safety and health into existing
courses is probably the most effective approach since
there could be continual reinforcement over a four to
five year period. However, this approach is probably
the most difficult to implement because of the lack of
engineering and science textbooks which incorporate
safety and health, coupled with the lack of faculty with
sufficient background and interest to teach such mate-
Textbooks seem to have the greatest impact in un-
dergraduate engineering courses, and most chemical
engineering programs generally use many of the same
textbooks. There is both room and a need for example
and assignment problems and discussion which relate
the subject matter to safety and health. NIOSH is
developing a coordinated effort with the authors of
textbooks and publishers to include safety and health
in new and revised texts. Also, the Safety and Health


Division of AIChE and CCPS is attempting to develop
supplemental instructional material.
In addition to revised textbooks, supplemental re-
source materials for incorporating safety into en-
gineering and science courses are either available,
being developed, under consideration, or are hereby
proposed. These include

Hazard workshop modules available from the Institute of
Chemical Engineers, UK
Case studies and engineering problems through NIOSH,
ASEE, AIChE Safety and Health Division, Stanford Re-
search Institute
API Safety Digests
Government publications available from NIOSH, EPA,
DOE, etc. [5]
Example problems derived from journal articles [6]
Information available from the Chemical Manufacturers

This paper is intended to be
a philosophical commentary on the need and
rationale for incorporating safety and health into
the chemical engineering curriculum.

Association, manufacturers of safety equipment, insur-
ance companies, etc. [7]
Industrial and laboratory safety manuals and other supple-
mental books [8, 9]

The National Institute for Occupational Safety and
Health has established an engineering faculty net-
work, offers short courses, and can provide workplace
safety and health videos.

There are many important health and safety con-
cepts that could or should be covered in the cur-
riculum. For chemical engineering students, many of
the concepts can be related to or derived from their
existing knowledge base.
Suggested topics include

Risk assessment and management, fault and event tree
analysis, hazard recognition and evaluation, and typical
hazards, e.g., overheating, reaction with water
Loss prevention, analysis of accidents
Monitoring and surveillance instrumentation and methods
Intrinsic materials hazards, e.g., toxicity, flammability,
reactivity, and indices, e.g., threshold limit values,
flammable limits, hazard ratings
Environmental and safety and health regulations, e.g.,
TSCA and OSHA acts
Occupational diseases and toxicology

Engineering controls in general, e.g., ventilation, isola-
tion, enclosures
Sources, generation and specific control of emissions,
e.g., dusts, mists, vapors

The above list is not complete and there is over-
lap. The exact choice of topics would depend upon the
specific course as well as the interests and knowledge
of the instructor.

Most required chemistry and chemical engineering
courses lend themselves to addressing safety and
health using materials supplemental to the text as de-
scribed earlier. Some preliminary thoughts follow.

Chemistry. Chemical hazards and toxicity could
be illustrated in both general and organic chemistry.
Permissible exposure and flammability limits could be
covered in the conversion of units. The labs, including
chemical engineering lab, could also enhance student
awareness of safety by requiring use of appropriate
protective equipment, e.g., goggles, by emphasizing
safe work practices and proper handling and disposal
of chemicals, and by acquainting students with the
need for appropriate protective clothing, monitoring
instruments, and emergency response procedures.
Safety quizzes could be given prior to doing an exper-
iment, and student teams could be used to do safety
checks. A videotape on chemical lab safety is available
from Rohm and Haas.

Material Balances and Thermodynamics. Expo-
sure and flammable limits could also be illustrated in
material balances via units conversions. Ventilation
could also be described here. Solvent vapor hazards
and work place concentrations could be related to
vapor pressure through a material balance involving
a spill or process release of a volatile substance. Pro-
cess leaks and fugitive emissions could also be illus-
trated through material balances and use of ther-
modynamic properties.

Transport Operations. Heat stress can be
explained in Heat Transfer and ventilation covered in
Fluid Flow. Sources of emissions from various types
of equipment, and equipment failure modes (e.g.,
leaks from pumps, valves and vents, seal failures, etc.)
could also be discussed in these courses, and proper
selection of equipment for a specific system could be
demonstrated. As a simplistic example, toxic volatile
slurries would not be filtered through a plate and
frame press-instead, a separation method such as


centrifugation would be used because of the capability
to better control emissions.

Transport Phenomena and Chemical Engineering
Analysis. Emissions, transport, and workplace con-

Examples of Safety and Health Cconcepts
for Various Courses

Material and

Fluid Flow

Heat Transfer





* Units/Conversions-threshold limit
values, flammable limits, relation-
ship of concentrations to regulations
* Mass Balances-number of room
changes for ventilation
* Adiabatic Flame Temperatures-
auto ignition point for flammable
and explosive compound
* Ventilation design principles, e.g.,
particle capture velocity
* Convection-heat stress
* Equilibrium properties and gas laws--
predicting maximum vapor concen-
trations of hazardous solvents in
the workplace
* Physical properties-relationship of
properties such as vapor pressure and
solubility to biological effects
* Selection of equipment compatible with
potential emissions of hazardous
* Mass and momentum transport-
modeling to predict workplace and
atmospheric concentrations of haz-
ardous vapors resulting from evap-
oration e.g., spills, fugitive emis-
* Simultaneous diffusion and chemical
reaction-human up-take and biolog-
ical transport of pollutants to evalu-
ate body burden, biological effects
and relationship to ambient concen-
* Process conditions for run-away
exothermic reactions
* Reaction kinetics of flame propagation

* Selection and use of protective equip-
ment and monitoring devices
* Appropriate safety practices
* Selection of chemicals for experiments
to avoid use of hazardous materials
* Environmental and health and safety
regulations, e.g., RCRA, OSHA
* Waste and hazards minimization
* Engineering controls

centration of hazardous materials can be predicted by
using transport properties and modeling. This could
also be related to various aspects of risk assessment,
and computer simulation could be used in this assess-

Kinetics and Reactor Design. Runaway reactions
and accident analysis are possible topics here, e.g.,
the relationship between available cooling water tem-
perature, catalyst concentration, and control of a
batch reaction could be examined. [10]

Chemical Engineering Laboratory. Since many
undergraduate laboratories resemble an industrial
pilot plant, it is one of the courses which could have
the greatest impact on student knowledge of safety
and health. Ideas for teaching laboratory safety were
presented earlier, and a recent paper describes the
teaching of safety in the unit operations laboratory at
Michigan Tech. [11] Equipment aspects described
under transport operations could also be presented in
the undergraduate laboratory.
The concepts of materials substitution and en-
gineering controls to avoid potential hazards could be
reinforced in the laboratory, e.g., using carbon dioxide
rather than ammonia in a gas absorbing or stripping
experiment, or if ammonia is used, providing proper
ventilation. Also, if students are required to wear pro-
tective equipment (e.g., face mask) they might become
more conscious of the need for safety and health.

Plant/Process Design. This is probably the most
obvious place to teach safety and health and is the
other course area which could have the greatest im-
pact on student knowledge. Most of the previously
mentioned safety and health concepts can be included
in the design courses (e.g., failure and 'hazards
analysis, selection of equipment, engineering controls,
etc.). Safety and environmental audits could be done
using the flowsheet and plant layout to identify possi-
ble points and types and composition of emissions; lo-
cations for and types of monitoring equipment; place-
ment of engineering controls, etc.
The design courses may also be the best place to
cover other engineering controls such as minimization
of hazards by materials substitution, recycle, process
change, or change of operating conditions.


A rationale for incorporating health and safety into
the chemical engineering curriculum has been pre-
sented, and various approaches for doing so have been


discussed. There will be difficulties in accomplishing
this goal, however, including the lack of faculty back-
ground. But the situation seems to be changing.
Perhaps faculty would be more inclined to include
safety and health in their courses if it was relatively
easy for them to do so. This could be accomplished by
the inclusion of relevant material in textbooks and a
ready availability of supplemental materials with
guidelines on where and how they could be used. Em-
phasis should be placed not only on problem solutions,
but also on the reading and understanding of the con-
cepts presented in the problem statement (e.g., the
process, specific reactions). Professors could demon-
strate to students, preferably quantitatively, the ethi-
cal and economic ramifications of what they are doing.
Reinforcement of why and how safety and health are
important and integral to chemical engineering is im-
Faculty background and interest in safety and
health could be heightened by providing opportunities
for related research, through consulting, and through
summer positions. The motivation to present safety
and health materials in the curriculum would also be
increased if industry expressed a desire for graduate
engineers with more awareness of health, safety and
loss prevention. Lastly, more specific accreditation re-

quirements for safety and health could be initiated by

1. Paustenbach, D. J., "Should Engineering Schools be Expected
to Educate Students in Environmental and Occupational
Health?" ABET Annual Industry/Government Symp., Knox-
ville, TN, October 23, 1984.
2. "Threshold Limit Values for Chemical Substances and Physi-
cal Changes in the Workroom Environment," Amer. Conf. of
Gov. Ind. Hygienists, Cincinnati.
3. Kletz, T. A., "Safety in Design," CEP March 1984, p. 11.
4. Mewis, J., "How Much Safety Do We Need in ChE Educa-
tion," Chem. Eng. Ed., Spring 1984, pp. 82-86.
5. "Control Technology in the Plastics and Resins Industry," Di-
vision of Physi. Sci. and Eng., Cincinnati, OH, January 1981.
6. Popendorf, W., "Vapor Pressure and Solvent Vapor Hazards,"
Am. Ind. Hyg. Assn. Jour. 45(10): 719-726 (1984).
7. Industrial Hygiene News, Reinbach Pub., Pittsburgh.
8. Cralley and Cralley, Aspects of Industrial Hygiene, Vols. I &
II, MacMillan, 1983.
9. "Guidelines for Hazard Evaluation Procedures," AIChE, 1985.
10. Russel, W. M., "Hazard Control of Plant Process Changes,"
AIChE, 81st Nat. Mtg., Kansas City, 1975.
11. Pintar, A. J., "Teaching Safety in the Unit Operations Labora-
tory," AIChE An. Mtg., Chicago, 1985.
12. Paustenbach, D. J., "Should Engineering Schools Address En-
vironmental and Occupational Health Issues?" J. Prof. Issues
in Eng. (ASCE) 113(2): 93-11 (1987). O

Continued from page 21.
from the XA versus t curve given by using the Curve
Fitting Program. In this program, the original data
can be entered, and the transformation function de-
fined by Eq. (9) to determine dXA/dt at various XA.
Either the Curve Fitting Program or the Multiple
Regression Program can then be used again with
transformation functions to estimate both k and a from
Eq. (10). For this example the results are a= 1.43 and
k= 0.0185 (1/gmole)043/s.
Verification of the model is efficiently obtained by
solving Eq. (8) with the Differential Equations
Simulator of the computational package. The equa-
tions needed for this simulation are given in Figure 6.
A tabular output of the pressure P versus time from
this program is also inserted into Figure 6 where
there is good agreement between experiment and the

Use of the Computational Package in ChE Curricula
At BGU, the computational package has been used
in a third year undergraduate course, "Introduction

to Computation for Chemical Engineers," and in a
graduate course, "Chemical Process Simulation." In
the undergraduate course, the students learn to for-
mulate the equations which describe various phenom-
ena or process equipment (e.g., dew point/bubble
point calculations; batch, CSTR and plug flow reac-
tors) and become familiar with the numerical methods
to solve the appropriate equations. Problem assign-
ments must be solved in three different ways: a com-
plete student program in Fortran, a student program
which calls for subroutines from the IMSL subroutine
library, and a POLYMATH program. Students con-
sistently relate that the most convenient and least
time consuming of the three solutions involves the use
of POLYMATH, and many continue to use the pack-
age in other courses.
In the graduate "Chemical Process Simulation"
course at BGU, the computational package is used to
experience different iterative methods for the solution
of recycle systems (as in Example 2) and to simulate
fairly complicated process units such as are found in
distillation columns. In the course they also use com-
mercially available simulators where results can be


The equations:
d (xa/d (t)=kxcaa0(alp-l) x (-xa) alp
p:1B1. 32
=.01851 Integration Results
alp=1.43 -."- 101
Initial values. t xa- B.8 19 i3
Final value: tf 5990, A 25.00 97.479
59.r M 94,041
75, M9 9C 953
1919.6 88.179
125.9 85.654
158. 9 83,373
175.90 81.298
29N. 79.364
225.9, 77. 61
259.9 75.999
e. Enter/Chane/Delete equations, t. E/C/D the title,
i. Change initial values. f. Change final value.
r. Restart from the current conditions.
F8 uhen done. F6 for helpful information.

FIGURE 6. Equations and constants for Example 3 (insert
is results table).

compared for simple systems. This use of the
POLYMATH package contributes to the student's un-
derstanding of the various solution methods.
At UConn, the POLYMATH package is given to
all juniors in chemical engineering. Two three-hour
workshops are given to the students at the beginning
of the academic year which cover the five programs
and their capabilities. A problem assignment with typ-
ical chemical engineering applications completes the
introduction to the package. All students can receive
their own set of POLYMATH programs or utilize
them in several PC computer labs on campus.
Undergraduate students are then able to utilize
the package in any of their courses as appropriate.
This allows use of the computer to solve numerical
problems which previously would have been very
time-comsuming or not even attempted. Particular
use has been made in chemical reaction engineering
and process dynamics and control courses. Details re-
garding the numerical analysis techniques are pro-
vided in an elective "Chemical Engineering Analysis"
course. In general, students may use POLYMATH
in conjunction with many of their regular courses and
in some independent study courses as well.

Our experience has clearly demonstrated that a
personal computer can be used to considerable advan-
tage in numerical problem solving in many chemical
engineering courses. The capability to easily apply
numerical methods with packages such as
POLYMATH will allow the solution of more realistic
problems without requiring excessive student time.
This new calculation tool will allow the appropriate
use of computing within our curricula. It supports the

concept of providing open-ended realistic problem sol-
ving which is being emphasized by ABET require-
Students and faculty seem to enjoy problem solv-
ing that is interactive, user-friendly, and readily ac-
complished through the use of personal computers.
We expect that the capability of solving realistic prob-
lems efficiently will have a major impact on chemical
engineering education by shifting the emphasis from
the technical details of the computer solution to the
formulation/analysis of the particular problem or pro-
cess. It is a continuing challenge to the educator, how-
ever, to make sure that the student understands the
numerical procedures that are being applied to obtain
the solutions.

a = order of reaction
C = concentration
E = equipment transfer matrix
F = vector of feed flow rates
f = vector of functions
H = estimate of the inverse of the matrix of partial
k = reaction rate coefficient
L = vector of liquid flow rates
P = variable in Eq. (6)
P = total pressure
Po = initial total pressure
r = rate of reaction
r = calculated value of a "tear" stream
t = time
v = volumetric flow rate
V = volume of a reactor
V = vector of vapor flow rates
XA = conversion of component A
y = variable in Eq. (6)
z = estimated value of a "tear" stream

1. Fogler, H. S., Elements of Chemical Reaction Engineering,
Prentice Hall, Inc., Englewood Cliffs, New Jersey (1986).
2. Henley, E. J. and E. M. Rosen, Material and Energy Balance
Computations, John Wiley & Sons, Inc., New York, NY (1969).
3. Hill, C. G., Chemical Engineering Kinetics & Reactor Design,
John Wiley & Sons, Inc., New York, NY (1977).
4. Shacham, M., M. B. Cutlip and P. D. Babcock, "Simulation
Package for Small-Scale Systems," Microprocessors and Micro-
systems, 9, 76 (1985).
5. Shacham, M., S. P. Macchietto, L. F. Stutzman and P. D. Bab-
cock, "Equation Oriented Approach to Process Flowsheeting,"
Comput. Chem. Engng., 6, 79 (1982).
6. Univ. of Texas, Septenary Report, "Chemical Eng. for the Fu-
ture," Chem. Eng. Prog., 81, 9 (1985). [





Texas A&M University
College Station, TX 77843


Most undergraduate [1, 2] and graduate [3, 4] texts
in chemical engineering thermodynamics illustrate the
separation of a single liquid phase into two liquid
phases for a binary system by graphs of the change of
the Gibbs energy upon mixing, AmG, versus composi-
tion, xl, at constant temperature, T, and pressure, P.
The first graph, Figure 1, is often for a symmetric,
non-ideal liquid solution obeying the one-constant
Margules model [1]

(AmG/RT) a a Gm xG1 x2G2
= x ian xi + x2 An X2 + Axx2

where G1 and G2 are the pure component molar Gil
energies, Gm the molar Gibbs energy of the liquid m
ture taken as a single phase, and A, which depel
only on temperature, indicates the deviation from
ideal solution (where A= 0). In terms of the exc
Gibbs energy, GE, and activity coefficients, yi, Eq.
(GE/RT) a Ax x2 a x1 in yl + x2 an Y2

An y, = Ax2

and An y2 = Ax1

Single-phase stability [5] is observed as long as

(a2Gm/aX12)p,T 0 or [a2(AmG/RT)/ X2]p ,T 0
For our symmetric binary solution

0' la( [ G/RT)/ax]pT = An(X1/X2) + A(x2 x)
a [a2(AmG/RT)/ax12]pT = (X2X)-1 2A (4)
Thus, Figure 1 provides single-phase stability over
the entire composition range for A < 2 because
(x2X1)-1 reaches its minimum value of four at x, = 1/2,
0 Copyright ChE Division ASEE 1988

ix- Symmetric Solution
nds A .50
A 1.75
(1) Mole Fraction X,

(2) FIGURE 1. The dimensionless change in the Gibbs
energy upon mixing, AmG/RT, versus composition x, at
a constant temperature and pressure for a symmetric
solution. IP = inflection point.

the vertical line of symmetry. However, when A > 2,
we observe a pair of symmetric inflection points (IP)
at(x2xl)-1 = 2A plus a pair of symmetric minima at

An (X /X2) = A(x1 x2)
Between the minima and inflection points the solution
(as a single phase) is termed metastable whereas be-
tween the inflection points it violates material stabil-
ity [5] and so is termed unstable. Physically, a single-
phase liquid is stable between the pure-component
end points and the minima (labeled a and 13) whereas
two liquid phases of composition


xa and xI

are stable for any overall system composition zI falling
between the minima.
Numerical determination of

xa (= 0.5 e) and xi (= 0.5 + E)
is a simple task for symmetric solutions because by
Eq. (3)
An [(0.5 + e)/(0.5 e)] = 2EA (5) F
Given a value of A, Eq. (5) should be solved for E by
trial; a useful first approximation is

2 (J 1 + 8 (A 2) i (6)

based upon the series expansion of

an (1 y) Z I (ynln)
which is exact as e approaches zero. At A = 2.3, Eq.
(6) provides E within 2.5%. However, when A = 3.0,
E by Eq. (6) is 13.8% too high and the limiting value
of E = 0.5 is reached by Eq. (6) for A = 3.067.
For the symmetric solution, upper and lower crit-
ical solution temperatures (CST) occur at A = 2. If

(dA/dT) > 0 at A = 2

then we have a lower CST; conversely when

r /

Philip T. (Toby) Eubank is professor of chemical engine
Texas A&M University. He received his BS degree from Rose-H
Institute and his PhD from Northwestern University. His resea
terests are in the thermo-physical properties of fluids and flui
tures plus electrical discharge machining. (L)
Maria A. Barrufet is a research assistant and a PhD candi
Texas A&M University. She received her BS and MS degrees fr
National University of Salta (Argentina) and the Southern N
University of Bahia Blanca (Argentina). Her research interests ar
modynamics of mixtures and optimization of vapor-liquid equil
predictions through mixing combining rules. (R)

0 1
Mole Fraction X,

FIGURE 2. Similar to Figure 1 but for an unsymmetric
solution showing two minima. The intercepts ii equal
(Gi G,)/RT, the difference between the chemical poten-
tial in the solution of component i and that of pure i.

(dA/dT) < 0 at A = 2
we have an upper CST.


The equilibrium tie-line between the coexisting liq-
uid phases a and p of Figure 1 is horizontal because
the solution is symmetric. Figure 2 illustrates the tie-
line for the general case of an unsymmetric solution.
This tie-line must be tangent to the curve at both a
and p in order to have equality of the chemical poten-
S tial of species i,

The method of tangent intercepts [3] provides the in-
ring at tercepts I, and 12 (Figure 2) in terms of partial molar
lulman properties

rcn in-
d mix-

date at
om the
re ther-

I, = (01 G)/RT

and 12 = (G G)/RT (7)

For a tangent common to points a and 3,

a1 1 1
and likewise for the second component. A minimiza-


... it makes little difference whether we are
examining liquid-liquid ... or gas-liquid phase
separation such as dew and bubble points . the
"inside-out" algorithm remains efficient for
determination of the equilibrium states for any
type of phase transition for a binary system.

tion of Gm (or maximization of system entropy) pro-
vides the number of phases that are stable for a given
overall composition z1. A single phase rich in compo-
nent 2 is thus stable from

x =0

to xi

two phases are stable for z, between

xa and x0

and a single phase rich in component 1 is stable from

x1 to x = 1
to x1 1
The curve of Figure 2 represents the single-phase Gm
whereas the tangent line represents the two-phase
Use of the tangent plane criterion for stability in
phase equilibria has enjoyed considerable interest in
the past decade [6, 7, 8, 9]. Most of the algorithms are
based upon Newton's method and are different from
that of the next section.
Because the "curvature,"

[a2 (AG/RT)/ax,2 1,T

is positive infinite at both pure-component end points,
it is obvious that the inflection points of the curve
come in pairs. For general, unsymmetric solutions, it
is further obvious that the equilibrium points (a, 3)
do not occur at minima and, indeed as in Figure 3,
there may be phase separation with only one
minimum. In conclusion, the presence of one or more
pairs of inflection points are a necessary and sufficient
condition for phase separation for an overall composi-
tion z, bounded by

x and xl

Each pair of IP provides a tangent line, which is an
equilibrium tie line so long as it does not intersect the
curve at a composition outside the two tangent compo-
sitions (as in Figure 3). The tangent line for each IP
pair has tangent compositions outside the composi-
tions at the IP. However, these tangent compositions
do not represent stable equilibrium points when their
tangent line cuts the curve at another point as in Fi-

gure 4. Here the interior IP pair is "wasted" and
meaningless as far as real fluid behavior is concerned.

Assume that we have an equation,


for the curve representing the single-phase system.
We first set

0"(x ) -= 0

and solve for the location of the IP pairs. Imagine for
simplicity that only one pair occurs as in Figure 2. To
find the tangent line,

xI and x

we recognize that the tangent line:

1) Must pass through the points a and P.

Mole Fraction X1
FIGURE 3. An unsymmetric solution exhibiting four in-
flection points (IP) which produce double phase separa-
tion, da/f and 8/y, but only a single minimum. Note that
the tangent lines a and b do not cross the curve.


2) Must be tangent to the 4 curve at a and at p3
so that

=' (x) = '(x = m

the slope of the tangent line.
3) Cannot cut the 4 curve at any x--that is, must
lie entirely below the 4 curve.

The equation of the tangent line is thus

TL = 1 + mx = IX m(1 xi) (8)
m = p'(x ) = *'(xl) = w[(xa) -(xl)]/(x' x) (9)

Our new algorithm is based upon the observations
that (1) the equilibrium points a and p must lie outside
the IP pair and (2) a straight line drawn through the
IP pair will have a slope very near to that of the cor-
responding tangent line. Rather than seeking

by approaching from

and x

the pure end-points, this al-

Mole Fraction X,
FIGURE 4. An unsymmetric solution exhibiting four in-
flection points (IP) which produce only a single phase
separation, a/f, but three local minima. Note that the
tangent lines b and c cross the curve and so are disal-
lowed in determination of phase equilibria.

gorithm is an "inside-out approach." The general nu-
merical procedure for this algorithm follows (see Fi-
gure 2):

1) Set *"(x ) = 0 and solve for the IP pairs,

x1o and xbo

2) Approximate m by
mo = [4(x)o- (X O)]/(Xa xb0)

3) Solve the equation
1'(x ) = mo

for the roots

and xbl

4) Approximate m by

m1 = [(xa) (xl)]/(x' xb')

5) Repeat step (3) to provide
Xa2 and Xb,

and continue to repeat steps (2) and (3) until

(xn xan-1) s 104 (x xbn-1
for example, as a convergence criterion. Then

an a
x = X

and x1 x1

a bn an bn
mn = [(x 1) ((x )]/(x x1")

We now show two numerical examples of the above
algorithm as applied to liquid solutions qualitatively
similar to Figures 2 and 3, respectively. All the calcu-
lations in this article were performed with a simple
hand calculator.
The qualitative nature of both Figures 2 and 3 is
captured by the simple mathematical model

0 = x An x + X2 An x2 + BX1X2(1 + xl)


where B is only a function of T. This empirical equa-
tion, which is of the two-constant Margules form
(A12 = B, A2 = 2B), may be differentiated to yield
i' = An (x /x2) + B(1 3x2) (11)

0" = (X X2)-1 6Bx1


When set to zero, Eq. (11) provides the location of
extrema of whereas Eq. (12) provides the location
of IP pairs. It can be shown that no IP pairs exist for
B (9/8) whereas only a single pair exists for B >
(9/8). For (9/8) < B < 1.484774 only a single minimum
of ( exists as in Figure 3. However, when B >
1.484774, two minima and one maximum occurs as in
Figure 2.

Case I: Let B = (7/4) to correspond qualitatively
to Figure 2. Eq. (11) provides a minima at x, = 0.1676
and at 0.9537 whereas Eq. (12) yields IP at x, =
0.3976 and 0.8758. Following the steps of the new al-
gorithm at the end of the previous section

x = 0.3976

so that

and x = 0.8758


[-8.6223 (-1.8142)]10-2
0 (0.3976 0.8758) 0 37

the first estimate of the slope of the tangent line. Eq.
(11) is now solved for <' = mo to obtain the roots

x = 0.1973

and xI = 0.9631

From Eq. (11),

x = 0.3696

( = 0.233290
respectively. Then,

m = 0.25645

x = 0.3722

= 0.232600
respectively. Then,

and x = 0.8891


and x = 0.8902

and 0.099756

m2 = 0.25646

xa = 0.3722 and xb3 = 0.8902

are our equilibrium values of

x and xl

-16.4800 (3.5874)]10-2 0.1
m1 (0.1973 0.9631) = 016835

the second estimate of the slope of the tangent line.
Eq. (11) now provides the roots

x1 = 0.2036

and 1 = 0.9645

m 1-16.3820 (-3.5657)]10-2 = 0.16844
2 (0.2036 0.9645)

Eq. (11) yields

xa3 = 0.2036

and xb3 = 0.9645

so that our convergence criterion has been met. Thus

x = 0.2036

and x = 0.9645

are the equilibrium points. Note their position relative
to the minima.
Case II: Let B = (4/3) to correspond qualitatively
to Figure 3 on the right side of the minimum. Now
there is but one minimum at x, = 0.2546 plus the IP
pair at x, = 0.5000 and 0.8090 where d = -0.193147
and -0.114967, respectively. The first approximation
of the tangent slope is thus
S(-0.114967 + 0.193147)
o ( 0.8090 0.5000) = 0.25299

The new algorithm also provides more rapid con-
vergence for more complicated forms of ( requiring
the use of a digital computer. Examples are where d
is given by the van Laar equation, general two-con-
stant forms of the Margules equation, or from equa-
tions of state (EOS) with mixture combining rules
(MCR). Further, it makes little difference as to
whether we are examining liquid-liquid phase separa-
tion or gas-liquid phase separation such as dew and
bubble points (with < from EOS/MCR). This "inside-
out" algorithm remains efficient for determination of
the equilibrium states for any type of phase transition
for a binary system.
Cases of more than one pair of inflection points,
Figures 3 and 4, are solved easily by the present al-
gorithm provided an overide computer program is
used to first calculate the slope of line through the
paired inflection points. In the case of Figure 3, these
slopes are of opposite sign so two separate regions of
liquid immiscibility exist. The equilibrium points (a,
3)and (8, y) may then be found by double application
of the present algorithm as illustrated by Case II for
(8, y) on the right side of Figure 3. With Figure 4, the
overide program would see that the slopes of lines
through the paired IP are the same. It would then
check the slope from the two outermost IP. Knowing
this slope to be between the original two in magnitude


is a strong indication that the final equilibrium tangent
is a rather than b or c of Figure 4. These comments
are only guidelines for Figures 3 and 4-see [6], Fi-
gure 5, for a different case involving double pairs of
IP. The only rule is that no true equilibrium tangent
may cut the Gibbs' curve at any composition.
We are presently working on procedures to extend
this algorithum to ternary and higher component sys-


We wish to thank the Exxcn Research and En-
gineering Co. and the National Science Foundation
(Grant #CBT-8420547) for their financial support. We
also acknowledge technical discussions with C. D. Hol-
land, particularly in regard to extension to multicom-
ponent systems.


1. J. M. Smith and H. C. Van Ness, Introduction to Chemical
Engineering Thermodynamics, Fourth Ed., p. 450, McGraw-
Hill (1987).
2. S. I. Sandler, Chemical and Engineering Thermodynamics,
pp. 451-453, Wiley (1977).
3. H. C. Van Ness and M. M. Abbott, Classical Thermodynamics
ofNon-Electrolyte Solutions, pp. 382-390, McGraw-Hill (1982).
4. M. Modell and R. C. Reid, Thermodynamics and Its Applica-
tions, Second Ed., p. 247, Prentice-Hall (1983).
5. J. S. Rowlinson and F. L. Swinton. Liquids and Liquid Mix-
tures, Third Ed., Butterworths (1982).
6. L. E. Baker, A. C. Pierce and K. D. Luks, Soc. Pet. Eng. J.,
22, 731, (1982).
7. M. L. Michelsen, Fluid Phase Equil., 4, 1, (1980); 8, 1, (1982);
8, 21, (1982); 16, 57, (1984); 23, 181, (1985); 30, 15, (1986); 33,
13, (1987).
8. L. X. Nghiem and Y.-K. Li, Fluid Phase Equil. 17, 77, (1984).
9. L. X. Nghiem, Y.-K Li and R. A. Heidemann, Fluid Phase
Equil., 21, 39, (1985). O

REVIEW: Reactor Engineering
Continued from page 7.

California Professional Engineers exam. Some involve
new technologies (semiconductor processing, biotech-
nology) and some require numerical solutions. In sev-
eral chapters excellent problems on critiques of jour-
nal articles are given.
A series of accompanying interactive programs for
personal computers is available on floppy disks,
though they must be purchased separately from the
University of Michigan. These are interesting prob-
lems that can be used as homework assignments since
they provide the student a coded grade. Students find
the programs to be both fun and helpful for learning
reactor design.

A few aspects of the book could be improved. As
done in essentially all reactor design texts, fractional
conversion is used as a dependent variable and solu-
tions start with an integrated form of the design equa-
tion. A more general approach, which is more easily
extended to multiple reactions and complicated reac-
tors, would be to use flow rates and number of moles
as dependent variables and start with the differential
form of the design equation. The energy balances in
Chapter 8 are complicated by using variable heat
capacities and symbols for several types of heat
capacities. In Chapter 6, the rate of reaction is incor-
rectly shown as being proportional to the square of
the total site concentration on the catalyst surface.
Also, as done in many texts, more significant figures
are given in the solutions to the example problems
than are justified by the data presented.
In summary, this is an excellent undergraduate
text for reactor design and it will likely be adopted by
a large number of departments. It could also be used
as a graduate text if supplemented. []

Rduued by 39 USC- 685J
quarterly 4 See Attached Rates
LLLALD. 0INIO.ENRiG E DiUTIUN, Do1m J17. Chemical Englinering Depart-nt,
bulversity of FloriDa, Gainerville, Alacnua, Florida 32611
CGemical hnaii ering Division, American Society for Engineerlng Education,
11 UuPoia Circle., WaaDninto, DC 20030
PUBLISHER INme an Compli ea lH Addm=
ASLE Lnemical mn~inering Division, 11 UPont Cirtle, Waaaington, DC 200UJ
EDITOR SN.. ,,.dC r oplI. .I aDd.a );
Kay FanieS. Cnemical GtineerinBB Departmtnt, KOom 319.
SUniversity of Florida. Gainesville, FL 32611
MAUGING EDITOR 15m.. 0 C-P lIF MD,o Add10
CarolD ocumD, Cnaulical nih iLeerinB Department, Roo 317,
University of Florida, .aioeavllle, FL 32611

'C I ..O.O a t P .r AD, a.. ,. ,...
listed aovew Ediror aa lfiln i ahonv.


Thr prose, f uncton, and nnprfit tatus of thn 'rization ind th exempt status for FnTfral Lncome t puMgoo [Ch= i k o1e-
A TOTALNO COPIES fe1 1 R..) R405 196U

I certEE fythtthestatements made byO O THEIR L AN ITLEOFEDITOR HE INESANGER NER

G T STAL 11-,,, T. : ll ...O ,,, e ..........l N I I I 2400 1.5.
I Certify that the statemeTntsm ad mTOR
me above are cOeCt and complete d tor
EN 3526, R .,, 1, -! T" I. ...... .t. ri. 1 r,


J laboratory


Organize It to Parallel Industrial Process Development

Purdue University
West Lafayette, IN 47907

OUR FINAL CHEMICAL engineering laboratory
course essentially doubled because of both a rec-
ord enrollment of seniors and a simultaneous cur-
riculum change that offered this course only once a
year. Faced with 180 seniors in a single chemical en-
gineering laboratory course, we realized that some-
thing had to be done to relieve the overwhelming
strain that this large enrollment exerted on our man-
power and facilities. Smaller classes of a previous dec-



Roger E. Eckert is a professor of chemical engineering at Purdue
University. After he received his BS in chemical engineering at Prince-
ton University and his MS and PhD at the University of Illinois, he
joined DuPont, where his major assignments were process research
and development and application of mathematics and statistics. Since
1964 he has taught both undergraduate and graduate courses in
polymers and in statistics, and also many required courses at Purdue.
His current research is on rheology and mechanistic model discrimina-
Robert M. Ybarra is a lecturer in chemical engineering at the Uni-
versity of Missouri-Rolla. He received his BS in chemical engineering
from the University of California-Santa Barbara and his MS and PhD
in chemical engineering from Purdue University. He worked two years
for DuPont's Textile Fibers Department before returning to Purdue as a
Postdoctoral Research Associate and Visiting Instructor. He joined UMR
in 1984. His research interests are in rheology and two-phase liquid
*Present address: Department of Chemical Engineering, 143
Schrenk Hall, University of Missouri-Rolla, Rolla, MO 65401-0249

ade permitted informal instruction, and the learning
experience was tutorial. A large enrollment requires
an alternative approach which is necessarily more
structured. Therefore, we needed a structure that
would allow efficient and timely communication among
students and faculty. Our course size and organization
contrast with that of Rochefort et al. [1], who de-
veloped a process laboratory for a class size under
Since the laboratory would run all day, all week,
we thought, "Why not organize the course to operate
as an industrial process development department?"
Just as industry subdivides process development
along product lines, we grouped the laboratory pro-
jects into divisions of commonality managed by a fac-
ulty member and supervised by a teaching assistant.
Although this reorganization did not reduce the facul-
ty's workload, it did make the teaching more efficient.
Pedagogically a laboratory course should culmi-
nate with students preparing excellent technical re-
ports. However, such a goal is often thwarted because
most student effort occurs at the end-the all-nighter!
To prevent this last-minute rush, we patterned our
guidance after that which we experienced in industry
where management is continually updated through
scheduled meetings and written memos.
The key meeting that we stress is the preparation
conference. At this meeting, several groups in a com-
mon project area orally present their proposed work.
The faculty member reviews the proposals and offers
suggestions to improve or redirect the work. Since
many of the ideas discussed at the conference are of
mutual benefit to all the participants, we find these
multiple-group sessions are an effective way to reach
a large but manageable number of students. Other
written and oral follow-ups, such as sample calcula-
tions and an oral progress report, are scheduled when
criticism is most useful-not saved for grading at the
In the course we reorganized, students must com-
plete conceptual process designs based on laboratory
data. These designs range from a single piece of equip-
@ Copyright ChE Division ASEE 1988


Faced with 180 seniors in a single chemical engineering laboratory course, we realized that
something had to be done ... Since the laboratory would run all day, all week, we thought, "Why not
organize the course to operate as an industrial process development department?"

ment to an entire plant. The course is divided into
three project cycles, each lasting four and a half weeks
or nine three-hour laboratory periods. During a cycle,
a design project from one of the three areas (Table 1)
is completed.
In the past a faculty member typically handled one
or two laboratory divisions with all twelve of the pro-
jects in Table 1. This situation led to little improve-
ment in the projects. Twelve divisions were needed to
accommodate the 180 students. We decided to follow
an industrial-type line organization (Figure 1) and to
restructure the course along project areas. Thus, each
area was headed by a faculty member (manager) and
a graduate teaching assistant (supervisor). Assistants
supervised laboratory divisions where all projects
were conducted, but their grading assignment was in
a limited project area. By concentrating each indi-
vidual's effort to fewer projects, we could

improve our guidance of the students
increase the possibility of improving our laboratory pro-

To further simulate an industrial assignment in
process engineering, we give the group members spe-
cific job titles; these job responsibilities rotate
through the three cycles

Laboratory Design Project List



R1 Continuous Stirred-Tank Reactor
R2 Continuous Esterification in a Tubular Re-
R3 Gas-Phase Dehydrogenation
R4 Catalytic Cracking
S1 Extraction in an Agitated Staged Column
S2 Continuous Fractionation in a Bubble-Cap
S3 Continuous Fractionation in a Sieve-Tray
S4 Batch Fractionation in a Packed Column

Transfer Operations

Gas Absorption in a Packed Bed
Water Cooling in a Servel Tower
Air-Water Contact in a Fluidized Bed
Continuous Drying

FIGURE 1. Organization chart for Reaction/Reactor pro-

The lead engineer is responsible for the execution of the
The experimental engineer is responsible for experimental
aspect of the project
The design engineer is responsible for design aspect of the


In this section, we will outline the activities of the
students as they develop their process designs. Table
2 is typical of a project outline the lead engineer
should develop when planning the project's key
events. With such an outline, the group has an idea of
what must be accomplished and when. We will refer
to the six written and oral assignments (A1-A6) that
we require as we develop what activities occur during
a project cycle.

Design Problem

An open-ended design is presented to the students
during the first period of the project cycle. We have
selected the reaction project R3, Gas-Phase Dehydro-
genation, to serve as an example throughout this arti-
cle. Part of its statement is

Acetone can be produced by gas-phase dehydrogena-



tion of isopropanol catalyzed by 0.5% platinum on silica-
gel support at temperatures near 2000C. The reaction is
(CH )2CHOH + (CH)2O + H2 (1)

When the temperature is high or the residence time is
long, a dehydration reaction also takes place and produces
propylene as follows:
(CH3)2CHOH CH2CHCH3 + H20 (2)

The product desired is 99% acetone; the Sales Division
will estimate the market later. If there were no complica-
tions from the competing reaction, this could conceivably
be obtained in a single pass through the tubular reactor.
However, by-product formation may not permit this, and
it may be necessary to use a lower conversion followed
by a purification (probably by distillation) with the iso-
propanol recycled. Obtain the dependence of conversion
and fractional yields on processing factors; the design
could be optimized using this information.

Preparation Conference
The idea of a planning discussion is not a new one;
however, we feel the emphasis that we place on this
conference is. In the past an informal conference
would take place before the students knew what the
equipment could do. The professor would question the
students to determine if they understood how to oper-
ate the equipment and what engineering principles the
equipment demonstrated. Often times the students'
answers were vague; the meeting would degenerate
and end with the professor telling the students, "Go
back and research the problem further." The end re-
sult was an unproductive exchange of ideas.
In a well-run industrial organization, sound plan-
ning is a must. Planning requires sufficient prepara-
tion, and ideas that are forwarded must be well-con-
ceived. At the planning meetings, the engineers are
generally those who present the ideas while manage-
ment sits in review. These formal meetings are much
more productive than an informal gathering where the
manager, rather than the engineer, forces the issue.
For these reasons, we moved the conference from
the first to the third period and instituted a formal
planning meeting, which we felt would be a
springboard for the project. Much spadework must be
done before the preparation conference, for example,
to identify the operating range of the equipment,
calibrate analytical instruments and make a few scout-
ing runs to iron out experimental difficulties. The for-
mal presentation forces the students to gather their
thoughts into a more coherent package. Students who
are well-prepared will spend less time in subsequent
laboratory periods deciding what they need to do
next. The required text of Holman [2] and a guidance

document, which details each engineer's role, help
them to prepare for their conference.
These conferences last about thirty minutes and
include a five-minute presentation from each group
member. The lead engineer chairs the preparation
conference. Each engineer also develops a one-page
outline with attachments that can be reviewed during
the talk. Here the objectives, approach, and division
of effort are established. We challenge the students to
create an industrial situation, or "scenario," as a basis
for their proposed work. This scenario establishes an
engineering need and thus better defines the project.
For the dehydrogenation example, project objec-
tives are first defined from the chemical reactions.
Acetone production involves parallel reactions and re-
quires that the kinetics of each be determined as a
function of temperature. From published kinetics on
other catalysts, the leader selects a temperature that
favors acetone production. A conversion is chosen and
the reactor and distillation columns are sized. Vapor-
liquid equilibrium data can be obtained from literature
to design the separation units. By obtaining rough es-
timates for the kinetics, the group then has a feel for
the order of magnitude of the rate constant. This is
an extremely important exercise to go through be-
cause few experimenters enter the laboratory without
some idea of what they expect to find.
The experimental engineer must calibrate a gas
chromatograph with known compositions of simulated
product mixtures. Reaction scouting experiments are
essential because they enable the engineers to sharpen
their experimental technique and provide information
needed for the experimental design. The time re-
quired to conduct a single run is needed to define the
maximum number of runs that can be performed dur-
ing the project. The range of important factors can
also be obtained, preferably before the preparation
conference. With the above information, the experi-
mental engineer can design an experiment from which
the kinetics can be modeled. Typically a first-order
kinetic model with an Arrhenius temperature depen-
dence is tested for adequacy. Space-time and temper-
ature are the factors, and conversion is the response.
If the first-order model is not adequate, then other
kinetic models, such as higher orders, must be
evaluated. The experimental engineer's design must
have enough levels of each factor and enough repli-
cates so a suitable kinetic model can be built.
The design engineer presents a flow sheet of the
conceptual process based on limited data or estimates.
Flows and compositions are projected, and the need
for the experimental data is established. Thus the de-
sign engineer learns how to design the various equip-


ment items in parallel with experimentation. For
example, the distillation columns to separate the reac-
tion products can be designed for the number of stages
required to attain a desired product and recycle
purities. Energy balances for reactors, reboilers, and
condensers are made. When production rates are ob-
tained, final flowsheet values are ratios of the prelim-
inary ones.
We have found that jointly conducting all four con-
ferences in a common area is beneficial. From the pre-
sentations and discussions the students learn much
about the related projects of other groups. Since they
will never conduct these other projects, they broaden
their technical knowledge as well as learn how to pre-
sent and discuss a project. Furthermore, their in-
terest is focused on similar concepts that can be valu-
able for their own project.
The lead engineer prepares minutes of the confer-
ence and discusses the disposition of unresolved items.
Such items often include redirection of the project,
correction of erroneous concepts and equations, and
collection of needed literature data. As in an efficient
industrial planning meeting, much is accomplished in
a limited time through a well-prepared conference and
the written minutes.

Sample Calculations
With transfers, promotions and project changes so
frequent in industry, supervisors are often unaware
of the project details. This lack of detailed knowledge
is simulated in education by changes in teaching as-
signments. Therefore the engineer must regularly re-
port details of the project to inform the supervisor as
well as clarify his own understanding.
Two additional checkpoints of this type that appear
in Table 2 are the sample experimental and design
calculations; they are given to the supervisor at
periods 5 and 6. These calculations explain the details
to the group supervisor and require him to update his
understanding of the work. Furthermore, the super-
visor has an opportunity to question these details and
explanations before extensive calculations are com-
pleted. This review of the sample calculations helps
the teaching assistants play a more active role in the

Oral Progress and Final Design Reports
The students receive a document that describes
the content of each section of the final design report.
In many other project courses final results are pre-
sented in an oral report. At this time the instructor
should call attention to errors, lest the student audi-
ence believe the results presented are correct. The

instructor must be negative in pointing out errors,
discrepancies and inadequacies; unfortunately these
errors are never corrected because the final report is
already completed.
With large enrollments, faculty-student contact
must be efficient and effective. By presenting oral
progress reports one week before the final written
report is due, improvements can be suggested and
carried out by the group. Typical action includes cor-
recting misconceptions, gathering needed additional
data, better modeling of the process steps, and im-
proving the design. As an example, one group pre-
sented a dehydrogenation reactor designed for 11%
conversion because this was the highest they obtained
in the laboratory. Their data were from a laboratory
reactor and should have been used only to model the
kinetics. The selected conversion was ridiculously low
for a plant design. Through our questions and com-
ments we were able to lead them to a reasonable de-
sign report.
Our approach parallels industrial practice where
revisions resulting from meetings are commonplace,
and final reports are almost always issued after the
information is disseminated at a meeting of the con-
cerned parties. Industry cannot afford the time delay
of waiting for final report preparation before making
business decisions.

Outline of a Project Flow Sheet for Dehydrogenation
1,2 establish analytical methods
conduct scouting experiments
design experiments
conceive process
3 Al Preparation establish kinetics
Conference 0 collect non-experimental data
4 A2 Minutes of needed for design
5 A3 Sample replicate data
test models
6 A4 Sample Design specify process design
7,8 A5 Oral Pro- analyze experimental error
gress Report assemble final tables and figures
complete design
9 A6 Final Report integrate and proofread final re-


Scheduled meetings already discussed are the
preparation conference and oral progress report. Two
additional places for contact are the laboratory and
consultation sessions. The manager should personally
interact with the engineers in the laboratory, but can-
not continuously supervise them since the laboratory
operates all day, every day. Still, morale of the en-
gineers and supervisors is boosted by the manager's
interest in the project. In the laboratory, the manager
assesses the project status and progress, and some-
times rolls up his sleeves to fix or operate the equip-
ment. He questions the engineers and offers advice to
help the group stay "on-track." The group that is on
top of their project has an opportunity to review prog-
ress by telling the manager about their accomplish-
ments and future work. A group that is not well-pre-
pared can lapse into details of technical problems that
they should solve themselves. Of course, they are so
For example, students raise questions about how
to determine feed composition in the dehydrogena-
tion. They know how to obtain the relative amounts
of isopropanol, acetone, and water in the product
stream from the GC calibration. However, the amount
of isopropanol in the feed is needed to determine con-
version and yields. In the laboratory reactor iso-
propanol is fed to the dehydrogenation reactor by
saturating a noncondensable carrier gas in a bubbler.
A typical student question is, "How do we measure
the feed composition?" Our response is, "You can't!
What can you do?" They should assume that the car-
rier gas is saturated with isopropanol because nitro-
gen/isopropanol mixtures are not available for calibra-
The consultation sessions regularly bring the
groups together on a project-basis. At the beginning
of a project cycle, information already discussed under
"Preparation Conference" is amplified and applied to
the specific projects. After the conference, these ses-
sions are used to

exchange technical information that helps solve experi-
mental difficulties
consider alternative designs that promote the divergent
thinking portion of the problem solving process
elicit further information about the project that helps the
convergence to a solution.

Examples of improvements from both the experi-
mental and the design parts of the project are pre-

sented here. These improvements are products of how
we reorganized this course. On the dehydrogenation
project we observed erratic behavior of the catalyst
that was difficult to trace to the temperature or chem-
ical exposure history. Switching from nitrogen carrier
gas to nitrogen containing 10% hydrogen helped to
maintain the catalyst's activity, and thus more consis-
tent conversions were obtained. Gas sampling of feed
and product streams at high flow rates was also a
problem. A low-volume, three-way valve was installed
which cured the sampling problem. These improve-
ments were conceived and executed because the con-
cerned faculty member specialized in the project, and
he carefully observed and listened to many student
Student engineers often decide the design needs a
"safety factor" and therefore increase the equipment
size by, say, 50%. We point out that such overdesign
can kill a project through excessive capital costs and
that a more objective method is needed. Our method
uses the experimental uncertainty in the key design
parameter. The design engineer must consider this
uncertainty to assure that the plant will operate at
the designed capacity. For design purposes a conser-
vative limit on the parameter is selected instead of its
best estimate. For example, if the kinetic constant of
a first-order reaction, k, is the key parameter in de-
signing the reactor, the lower limit of the confidence
interval on k can be used to size this reactor. A some-
what larger reactor will be designed to provide a
safety factor for the uncertainty in k. By intercepting
a loose and inappropriate use of a "safety factor" at
the oral progress report we offer the engineers an
opportunity to strengthen their final report.

Grades are based on the elements described in this
section; we attempt to parallel the responsibility and
rewards that a manager assigns to his engineers on a
process development project. Each engineer is graded
on the preparation conference, laboratory perfor-
mance, the written report and the oral report. The
responsibilities for the preparation conference have
been described; performance involves efficiently
executing these tasks during the laboratory.
The written report is evaluated for both technical
content and communication skill. The technical grade
is assessed from sections of the engineer's primary
responsibility. Also, each member shares responsibil-
ity with others in the group. In industry, responsibil-
ity and rewards are closely related, and this concept
underlies our decision to have the lead engineer
graded on all sections of the report. The experimental


and design engineers aid in coordinating the report so
each receive an added one-fourth of the technical
grade of the leader's and other engineer's sections. A
separate communications instructor assigns a writing
grade for content, organization, development, style,
and mechanics of the sections written by each en-
The oral report is graded for skill shown in com-
municating the scientific and engineering concepts
with an intelligent but uninformed technical audience.
The grade is based on the choice of subject matter,
organization, balance, and the content of displays.
Two separate evaluations are made; one is by the tech-
nical supervisor and the other by the communications
instructor for the presentation in each respect.


Students evaluated aspects of the course on a five-
choice scale from "strongly agree" to "strongly disag-
ree"; written comments were also solicited. Since the
students have limited industrial experience we did not
attempt to have them evaluate how well our approach
parallels industry. That could be the subject of a ques-
tionnaire after they have had working experience.
They overwhelmingly concluded that the projects
reinforced and extended the concepts of prerequisite
courses in general. Specifically, the junior chemical
engineering laboratory course was cited in this cate-
gory. But also, they wanted better coordination be-
tween the prerequisite courses and the laboratory
projects. The reaction projects, which required full
plant designs, were rated as challenging and practical;
however, the students thought more time should be
allotted to complete the difficult projects, or else the
scope of such projects should be reduced. The open-
ended problem statement was credited with improving
their engineering problem-solving skills. Further, the
group seminar approach to the reaction project prep-
aration conference was judged to increase their know-
ledge of the reaction area.
Most students expressed a need for meeting with
the manager between the preparation conference and
the oral progress report. They also strongly favored
a class period to transmit initial information about
each project. Further, they thought manager visits to
the laboratory were essential, but supervisors were
not considered helpful in the execution of the projects.
They thought a Laboratory Center with computer ter-
minals, where calculations and report drafts could be
done, would have increased their efficiency. Finally,
equipment was judged to operate poorly. (So what's

In an unusual comment, one student suggested
that the communications portion be considered a sepa-
rate course so that it could be listed on their resumes;
it might help in job search!


Laboratory projects have been developed over
many years, and we recognize the contributions of the
faculty members involved. Many of these projects
were devised by John M. Woods. In addition to the
authors, in the two years preceding this organization,
A. H. Emery, N. H. L. Wang, and D. P. Kessler
contributed while they served as managers and Frank
S. Oreovicz sharpened the students' communication
skills. E. I. Franses taught the prerequisite labora-
tory course and contributed ideas through that course.
We also acknowledge R. P. Andres, school head, for
supporting this concept of teaching the senior chemi-
cal laboratory course for 180 students.


1. Rochefort, S., S. Middleman and P. C. Chau, "An Innovative
Process Laboratory," Chem. Eng. Educ. 19, 150 (1985).
2. Holman, J. P., Experimental Methods for Engineers, 4th ed.,
McGraw-Hill, New York (1984). O

Continued from page 25.
cations to Data Acquisition and Control, Van Nostrand-
Reinhold, 1983.
2. Mano, M. M., Digital Design, Prentice-Hall, Englewood Cliffs,
New Jersey, 1984.

0 0
0 1
1 0
1 1


0 1
1 0




Al Ba NC C D1 Y1
7420 W1 4-Inpout NMO gate Y AB.C.O



5V A6 6 Y A5 YS A4 Y4

Al Yl Ai Yz A3 Y3 7
7404 Hex Inverter Y = 1







Brigham Young University
Provo, UT 84602

IT IS A WELL-KNOWN fact that an engineer's ability
to communicate, both orally and in writing, largely
determines his success in advancing up the career lad-
der. In view of this fact, most formal engineering edu-
cation programs require some degree of proficiency in
technical writing. However, few offer training or ex-
perience of any kind in making oral presentations.
The importance of teaching engineers to make ef-
fective oral presentations is becoming recognized. In
a recent report on the future of chemical engineering
education, a group of prominent industrial leaders
suggested the use of oral presentations in at least one
course each year as a means of improving communica-
tion skills of graduating engineers [1]. In a recent
news article on chemical engineering schools and what
they are doing to adapt to the changing needs of the
chemical process industries, Donald Dinsel, DuPont's
professional-staffing manager, was quoted as saying,
"We like engineers who have taken writing and speech
classes" [2] (italics added). In the same article, both
Jordan Spencer, acting chairman of Columbia Univer-
sity's chemical engineering department, and Allan
Myerson, head of the chemical engineering depart-
ment at Polytechnic Institute of New York, were
quoted as saying that they are pushing for required
exercises in oral presentation for their students.
Since approximately 1977, undergraduate students
in chemical engineering at Brigham Young University
(BYU) have been required to take a course on making
oral technical presentations during their senior year.
The purpose of this course is threefold:

To provide training and experience in oral presentation of
technical material
To provide familiarity with the chemical engineering liter-
To provide exposure to a wide variety of engineering topics

The first two objectives are accomplished by re-
quiring each student to prepare and deliver two oral

B. Scott Brewster obtained his PhD degree from the University of
Utah in 1979. He subsequently joined the faculty at Brigham Young
University, where he taught courses and conducted research in model-
ing of solid propellant combustion, catalytic reactors, and simulation
of fossil fuels processes. He is currently a research associate in the
Combustion Lab at BYU, where he is conducting research on modeling
of entrained- and fixed-bed coal conversion processes. (L)
William C. Hecker received his BS and MS degrees from Brigham
Young University and his PhD degree from the University of California,
Berkeley. He joined the chemical engineering faculty at BYU in 1982
after several years in industry, and his research and teaching interests
include heterogeneous catalysis, chemical kinetics, infrared spectros-
copy, automotive emissions control, nitric oxide reduction, ammonia
oxidation, and coal char combustion. (R).

presentations on technical subjects. The third objec-
tive is accomplished by requiring each student to crit-
ically evaluate all of the presentations made by his
classmates. This paper discusses the mechanics of the
course and our experiences in teaching it.

The course is a one-semester (15-week) course for
which one semester hour of undergraduate credit is
awarded. Class size typically ranges from fifteen to
twenty students. The course meets twice each week
for fifty minutes, and 100% attendance is required of
each student. Students who miss a class period must
either attend one of the other sections of the course
or view the proceedings of the class they missed on
video tape. Although this requirement might seem
0 Copyright ChE Division ASEE 1988


stiff, we feel it is important that there be a good audi-
ence for each presentation.
The first two weeks of the course (four meetings)
are spent organizing the course and discussing tech-
niques for making good oral technical presentations.
Beginning with the third week, two presentations are
given by two different students each day. The presen-
tations are limited to fifteen minutes in order to allow
time for questions and evaluation. All students are
required to give their first presentation before any of
them give their second.
The presentation is based on a recent article
selected by the student from the chemical engineering
literature. A list of approved articles is handed out at
the beginning of the semester, but students do not
have to restrict themselves to this list. Other articles
may be selected, but are subject to approval by the
instructor. In preparing their presentations, students
are encouraged to critically evaluate the articles based
on their training and not merely accept everything
they read. They are encouraged to form their own
opinions and conclusions about the subject matter and
to be able to defend their position.
In order to help the students prepare for their pre-
sentations, they must write an outline of their pro-
posed talk and prepare the visual aids they expect to
use. Students are encouraged to make transparencies
and to use the overhead projector rather than writing
on the blackboard. The students are told that the
transparencies should be neat (drawing aids should be
used) and simple. Materials for making the trans-
parencies are made available.
After preparing the outline and transparencies for
the proposed talk, each student meets with the profes-
sor to evaluate his or her advance preparation. This
meeting must take place at least one full week prior
to the date of delivery in order to receive full credit,
and an advanced preparation approval form must be
filled out by the student prior to the meeting. On this
form, the student records the title of the presentation,
the title and source of the literature article he/she is
using for reference, two questions that the audience
should be able to answer at the conclusion of the talk,
and the answers to the two questions. At this meet-
ing, which normally lasts about ten minutes, the pro-
fessor reviews the student's outline, the visual aids,
and the two questions for the audience. It is stressed
in class that this meeting cannot take place until the
student has constructed a formal outline and com-
pletely prepared his visual aids. This requirement
forces the student to prepare well in advance, it gives
the professor an opportunity to review the student's
preparation and to make suggestions well in advance

In a recent report on the future of
chemical engineering education, a group of
prominent industrial leaders suggested the use of
oral presentations in at least one course each year as
a means of improving communication skills ...

of the actual presentation, and it gives the student
ample time (one week) to act on the professor's sug-
gestions and to polish his or her delivery.
In addition to giving two technical presentations,
each student must act as session chairman for one day.
The responsibilities of the session chairman include
preparing and distributing a program announcement
at the beginning of the session, introducing the speak-
ers, arranging for any special audio-visual needs,
keeping the speakers within the allotted time limits,
and conducting the discussion that follows each talk.
The program announcement typically contains the
names of the speakers, the titles of their talks, and
the two questions they expect the audience to be able
to answer. The session chairmen also have the option
of providing refreshments if they wish to do so. Re-
freshments seem to help ease tension and enhance in-
teraction between the presenters and the audience.
Following the talks and questions, each person in
the audience (including the instructor) completes an
evaluation form for each speaker. A sample evaluation
form is shown in Figure 1. The presentations are rated
as excellent, good, fair, or poor for several specific
items in each of the following areas: organization,
preparation, and delivery. For example, under organi-
zation the presentation is rated on its effectiveness in
preparing the audience at the beginning, maintaining
a smooth and continuous flow of thought, summarizing
the main points at the conclusion, and keeping within
the time limit of fifteen minutes. An overall rating on
a 1-to-10 scale is then given, where "0" means that the
presentation was "absolutely worthless," and "10"
means that the presentation was "outstanding beyond
description." The evaluation form also calls for specific
comments from the evaluators. The overall ratings by
the students are averaged in order to obtain a score
for "peer evaluation" of each presentation. After the
peer evaluation is computed by the instructor, the
names of the evaluators are removed, and the forms
are given to the speaker.
Each presentation is video-taped. This ac-
complishes two purposes. First, it allows students
who must be absent on any particular day to view the
presentations and complete the required evaluation
forms at a later date, and second, it allows the speak-
ers to view their own presentations. Each student is


required to view his own presentation within one
week of giving it and to complete a self-evaluation
form. This form asks the student to respond to the
criticisms of his classmates and to evaluate his own
strong and weak points. After completing the self-
evaluation form, the student meets privately with the
instructor to discuss his performance.
The course is graded on a pass/fail basis. In addi-
tion to receiving scores for advanced preparation,
peer evaluation, self-evaluation, and instructor evalu-
ation for each presentation, each student also receives
a score for class participation. This motivates the stu-
dent to attend each session of the class (or make it up
as discussed previously), to participate in the discus-
sions, listen courteously, arrive on time and perform
as session chairman satisfactorily. All of the above
scores are counted equally in determining the final


The value of this course in the overall educational
experience of our students has become increasingly

Evaluator Date
ORGANIZATION: Consider the following: lent Good Fair Poor
Introduction (Did it prepare the audience?)
Flow of thought
Summary (Did it summarize the main points?)
PREPARATION: Consider the following:
Smoothness (Was the presentation well-
Clarity (Could the audience understand the
Technical Content (Were the presubmitted
questions answered adequately?)
Visual Aids (Quantity? Quality?)
Explanation of Figures and Graphs (Was it
DELIVERY Consider the following:
Visual contact with the audience
Clarity of speech
Tone of voice
Scale: O=Absolutely worthless
2=Many serious deficiencies
4=-A few serious deficiencies, otherwise okay
6=A few strong points; pretty good
8=Many strong points; very good
10=Outstanding beyond description

FIGURE 1. Sample evaluation form

apparent over the past several years. When the
course was first initiated, the intent was to offer an
opportunity to the faculty and the brighter students
to make technical presentations to the senior class.
Over the years, it has evolved to the point where the
students do all of the presenting, most of the conduct-
ing, a good share of the evaluating, and where every
student is required to actively participate. Nearly all
of the students seem to appreciate the experience ...
especially once it is over.
A survey of class members recently taken at the
end of the semester asked two questions: (1) "Was the
class worthwhile?" and (2) "What did you gain from
it?" The answers to the first question were one-
hundred percent "yes." The answers to the second
question were more varied, but quite informative.
Over half the class said they gained valuable experi-
ence in speaking. One student wrote, "I gained much
valuable experience in front of technical peers talking
technical." Nearly half the class said they gained good
experience in preparing oral presentations. For exam-
ple, one student said he learned "the importance of
practicing a talk before giving it," and another said he
gained "insight into how to prepare a speech to make
it interesting." Twenty-five percent of the students
indicated that they gained confidence in giving oral
presentations. One stated, "I know that I know the
material better than the others and can tell them what
I know." A couple of students mentioned that they
had gained valuable experience in preparing and using
visual aids, particularly transparencies with the over-
head projector. Finally, twenty-five percent men-
tioned they gained important new knowledge by lis-
tening to the talks of their peers. One mentioned that
he was "exposed to a variety of very interesting chem-
ical engineering topics," while another said that he
"gained valuable knowledge of new processes and
products . not covered in regular classes."
The course does not offer enough experience for
every student to become a polished speaker, but it
does offer an excellent opportunity for improvement.
During the second round of talks, students have the
opportunity to immediately apply the things they
learned during the first round, and it is gratifying to
see the improvements in their ability to orally present
technical material.


1. "Chemical Engineering Education for the Future," CEP, Oc-
tober 1985, 9-14.
2. Farrell, Pia, "Chemical Engineering Schools: What they're
doing to adapt," Chemical Engineering, April 15, 1985, 23-
25. E


1 book reviews

by C.A. Miller and P. Neogi
Marcel Dekker, Inc., 270 Madison Ave., New York,
NY 10016; 376 pages $69.50 (1985)
Reviewed by
John C. Berg
University of Washington
The subject of interfacial phenomena is increas-
ingly capturing the attention of chemical engineers as
they are called upon to solve problems dealing with
surfactants, thin films, coatings, natural and synthetic
fibers, adhesives, lubricants, foams, pigments, pow-
ders, aerosols, emulsions and other materials for
which surface properties play a dominant role. These
properties are recognized as important both in the
traditional fluid-phase separation processes such as
distillation, absorption and extraction, and in the
emerging technologies of composite materials design
(polymer-fiber, polymer-ceramic, metal-ceramic, etc.)
and bioprocessing. The underlying science of interfa-
cial phenomena, however, is rarely treated in courses
in chemistry, physics, or engineering. This excellent
text should do much to encourage its inclusion in
chemical engineering curricula.
While a number of books on interfacial phenomena
or surface and colloid science are presently in print,
this book is unique among them in giving a com-
prehensive treatment to nonequilibrium flow and
transport behavior. It treats flowing and deforming
waves, films, drops, bubbles and jets, and considers
situations in which heat or species transfer across fluid
interfaces is taking place. It thus describes the range
of important phenomena (the Marangoni effects) in
which the potential energy of unequilibrated phases is
converted, by means of the interface, into the kinetic
energy of spontaneous convection. It also deals with
the damping and stability effects of surfactant films.
The organization of the book and its topical cover-
age are consistent with the general objective of the
authors "to combine in one text an account of non-
equilibrium interfacial phenomena such as wave mo-
tion and Marangoni flow with enough background on
the fundamentals of interfaces to enable the dynamic
analyses to be understood and to provide an initial
overview of the field." The first four chapters give the
background material, and deal with interfacial ten-

sion, wetting and contact angles, colloidal dispersions,
and surfactants, respectively. The material is de-
veloped in a manner which anticipates the subsequent
chapters dealing with the dynamics of interfacial sys-
tems. Many of the approaches used are novel and pro-
vide new insights into old concepts. For example, the
"mechanical" derivation of the Young-Laplace equa-
tion introduces the use of the control volume embrac-
ing the interface (the "pill box") upon which force bal-
ances, material balances, enthalpy balances, etc. can
be drawn. This approach reveals more clearly than
others the nature of the various "surface excess"
quantities, the equivalence between the "force" and
"energy" definitions of interfacial tension and the ap-
proximations entailed in the usual formulation of the
Young-Laplace equation. Finally, it sets the stage for
the development of the boundary conditions to the
thermal energy and convective diffusion equations.
The final three chapters, constituting nearly half
the text, deal with interfaces in motion-stability and
wave motion, transport effects on interfacial phenom-
ena, and dynamic interfaces, respectively. These
chapters are the centerpiece of the book. They pre-
sume prior knowledge of basic fluid mechanics and
transport phenomena, but include sufficient develop-
ment of such techniques as linear stability analysis
and matched asympotic expansions that these topics
may be used in the analysis of a variety of problems.
The text is written as a teaching tool. The style is
lucid and expository. Each chapter contains several
solved example problems and is followed by an ample
set of problems to be solved by the reader. Some are
quite challenging, and many represent topics of cur-
rent research interest. Each chapter also contains a
full listing of references, both to other textbooks and
to literature articles.
If there is any criticism at all to make of this text,
it is its brevity. Many topics have been left out al-
together: adsorption equilibria and dynamics, kinetic
phenomena of colloids (sedimentation and diffusion),
the rheology of colloids, electrocapillary phenomena,
and others. Also, very little attention is given to ex-
perimental techniques. Given the length of the text,
however, the choice of material provides a coherent
whole. It is a book which anyone interested in interfa-
cial phenomena, whether teacher, student, or prac-
titioner, should own. E





Villanova University
Villanova, PA 19085
Bucknell University
Lewisburg, PA 17837

METHODS FOR SOLVING multiple-effect evaporators fall
into three broad classes: a) traditional trial-and-error
taught in most textbooks [1, 2, 3], b) simplified methods
used as first-order design approximations [4], and c) com-
plex methods such as dynamic programming and evolutionary
operations [5, 6, 7]. In the past, the only one of the above
ever taught in the classroom (if evaporators were taught at
all) was the traditional trial-and-error method. This method
is extremely limited and almost useless for evaporator series

Donald D. Joye is associate professor of chemical engineering at Villanova
University. He joined the staff in 1981 after previous teaching and industrial
experience. He received his BSE degree in chemical engineering from Princeton
University in 1967 and his MS and PhD from Lehigh University in 1969 and
1972, respectively. Dr. Joye is presently active in research in heat transfer, fluid
mechanics, the unit operations and wastewater treatment. Drs. Koko and Joye
shared an office and many a stimulating conversation on the nature of things
during their graduate school days at Lehigh. (L)
F. William Koko, Jr. is an associate professor of chemical engineering at
Bucknell University, where he has been teaching since 1973. Dr. Koko obtained
all his degrees in chemical engineering at Lehigh University (PhD, 1976). He also
has six years experience as a process engineer with du Pont. His major interests
are data analysis, numerical analysis, process control and computerized
laboratories. (R)
Copyright ChE Division ASEE 1988

larger than three, yet virtually all unit operations textbooks
discuss it in some detail. The simplified design methods are
easy to use (assuming no boiling point rise), but are not
flexible or particularly accurate; and the computer methods,
although accurate and flexible, are generally much too
sophisticated for effective teaching, especially if the en-
gineering aspects of an evaporator train are to be taught as
In the classroom, the subject of evaporators is getting
squeezed out of the undergraduate curriculum. This is
partly the result of the difficulty in teaching it without get-
ting hopelessly bogged down in the analysis. A new method
is presented here, the details of which have been recently
published elsewhere [8, 9], which is both simple enough to
use in the classroom and accurate and flexible enough to be
used as a design tool in practice.


In searching for more efficient ways to solve the
evaporator system, it becomes clear that a different ap-
proach than the three previously mentioned is required. The
evaporator series can be viewed as another stagewise unit
operation, where fundamental equations for heat and mass
balances can be written for each stage, with the heat trans-
fer rate equation substituting for an equilibrium relation-
ship. These equations, when written in such a manner, are
not easy to solve. The evaporator train equations, shown
below for the backward-feed system illustrated in Figure 1,
do not lend themselves to solving directly as a set of non-
linear algebraic equations. In fact, almost none of these
kinds of nonlinear equations are easy to solve.
The equations for the ith stage in backward feed are:

mass balance
L+ L Vi = 0 (1)
heat balance
AH(x,T)Vi + h(x,T)L+ H(x,T)Vi h(x,T)Li = 0 (2)
heat transfer rate
[A(T) + SH(x,T,T )]Vi = Ui(x,T,...)Ai(Tsi_l -Ti) (3)

The variable Ai in Eq. (3) can also be written as ARi to allow
for variable area situations, where Ri is the ratio of area of
the ith stage to A. Setting Ri = 1.0 gives the most common
solution corresponding to equal-area evaporators, which
simplifies to Ai = A.


In the classroom, the subject of evaporators is getting squeezed out of
the undergraduate curriculum. This is partly the result of the difficulty in teaching
it without getting hopelessly bogged down in the analysis. A new method is presented here...

FIGURE 1. Backward feed, N-effect evaporator series

variable, AT. Then the set of equations is sequentially
linear, i.e., at any Ti the coefficients are defined. Thus the
resulting equations can be put into matrix form and solved
by any linear technique, such as Gaussian elimination. Of
course, iterations will be required if the new values of Ti are
significantly different from the old. Because the coefficients
are not strong functions of T, the equations don't change
very much from iteration to iteration, and convergence is
very rapid.
For example, Eqs. (1), (2), and (3) are rewritten below,
incorporating the changes discussed above, for a constant-
area solution.

aV.i +

eVi_ + fATi +
i-i i-i

(-1)Li + (-1)Vi +

bL. + cV. +
S i

gATi +

(1)Li+ = 0 (6)

dL+ = 0 (7)

kA = 0 (8)

The variable Tsi-1 is the saturation temperature of t
previous stage and can be computed as follows:

Tsi_ = Ti1 BPRi_

where BPR is the boiling point rise. The boiling point r
can be calculated from a knowledge of the solids concent
tion and a Duhring chart. The solids concentration can
computed from the solids balance

xiLi = solids = constant (5)

where x is the solids concentration.
In Eqs. (1) through (5) standard symbols have been used
for liquid (L) and vapor (V) flow rates, overall heat transfer
coefficient (U), temperature (T), heat transfer area (A), liq-
uid and vapor enthalpies (h and H, respectively) and heat
of vaporization at the saturation temperature (k). Subscript
"s" is used to indicate saturated conditions, and the variable
SH is the superheat in the vapor occurring as a result of a
boiling point rise. The superheat is the enthalpy rise above
saturated vapor enthalpy at constant pressure. Subscripting
is arranged so that externally supplied steam always enters
the first stage. At the ends of the series some of the vari-
ables are known. Corresponding equations for forward feed
can be generated by simply changing the i+1 subscript to
i-1 on all liquid flows (only). Mixed feed situations may be
handled by writing all the equations explicitly to show the
connections between stages.
The non-linearities in Eqs. (2) and (3) arise from two
sources; the coefficients are non-linear functions of x and T,
and in some cases other variables as indicated, and a cross-
product term, AT, exists. The hurdle of non-linearity can
be eliminated by redefining the cross-product term as a new

where the coefficients a-k are

a = AH(x,T)i- = [(Ts) + SH(x,T,T )]i_

(4) b = -h(x,T)i

ise c = -H(x,T)i
ra- d = h(x,T)i+
be e = [X(Ts) + SH(x,T,Ts) i-_

f = -U.

g = -f = Ui
k = -U.BPR

In general there are 3N + 4 unknowns and 3N equations.
Four boundary conditions can be specified, for example:

LN+1 = F = feed rate
L = P = product rate
AT0 = GA, thus G = T0
AT = G2A, and G2 = T

This results in a set of 3N + 4 linear, algebraic equations
with constant coefficients, which can be solved easily once
the values for the coefficients are known.
The method of solution, described in detail elsewhere [8,
9], can be summarized by the following:

Write the governing equations in linear format
Initialize temperature and composition in each effect by appor-
tioning the vapor flows and temperature drops


v,.I V,




Find the coefficients of the variables. These coefficients are the
enthalpy values, boiling point rises, and overall heat transfer coef-
ficients, all of which are functions of temperature and composi-
tion, or are specified as design (input) parameters
Solve the matrix of coefficients by any linear technique to obtain
values for the variables (L, V, AT and A) for each stage
Compute Ti (= AT,/A) and xi (= xNLN/Li)
When new values of T and x lead to significant changes in the coef-
ficients, do another iteration

Initialization to get first values for the coefficients may
be done by any method, including guessing, but the tradi-
tional method of apportioning vapor flows and T's leads to
more rapid convergence. This is especially important when
doing this by hand calculation but less important when a
fully computerized method is used.


The following example shows how students can solve a
triple-effect evaporator series easily, if they have enthalpy
tables, a Diihring chart, and a straightforward linear equa-
tion solver. Students were asked to find the steam rate and
the area for one, two, and three-stage systems given the
following information. The feed was aqueous sodium hy-

Feed = 20,000 kg/hr
x = 0.05
Xp = 0.60
TF = 3oc
To = 163%C, saturated steam

N = 17270 Pa
U = 787.8 W/m2*K = 2836 kJ/hr-m2-K (when N=2,3)
U1 = 157.8 W/m2*K = 567 kJ/hr-m2-K
Umid = 1701 kJ/hr-m2-K for middle stage when N=3

The units of U are changed to be consistent with enthalpy
units. Equations 2 and 3 can then be rewritten in a form to
make the coefficients and variables more obvious.

hLi + (Hs + SH)Vi hLi+1 (A + SH)Vi_- = 0 (6)
Ui(ATi1) UiBPRi_ (A) U(ATi) (X + SH)Vi- = 0 (7)

Here Hs is the vapor enthalpy at saturated conditions.
Preliminary calculations for the boundary conditions and
initial estimates of internal temperatures and flows are
shown below.
P = x F/x = 0.05(20000)/0.60 = 1667

Vtot = 20000 1667 = 18333
V, = V2 = 9166.7
x2 = XFF/(F V2) = 0.09
BPR2 = 2C (from Duhring chart)
BPRI = 60C (from Duhring chart)
T = 3C
T = 163%



= 57%C (from p2 = 17.27 kPa)
= T T2 BPR BPR2 = 44oC
= 1/(1/U1 + 1/U2) = 1/(1/567 + 1/2836)
= 472.5 kJ/hr-m2"K

AT1 = ATtt(Utot/Ul) = 36.7C
thus Tl = 126.3, Ts = 66.3
AT2 = ATot(Utot/U2) = 7.3C
thus T2 = 59, Ts2 = 57

One can see that the equal boilup assumption was used
for the calculation of the boiling point rise, and the driving
force temperature differences were apportioned by inverse
U relationship. These are the elements of the traditional
startup. The coefficients for the stage equations can now be
calculated and entered into matrix form according to Table
1. The order of the columns should always be consistent to
minimize confusion in a hand calculation. The last four rows
need not be solved as part of the matrix, but the calculations
are more convenient to do in the matrix than outside it.
When additional stages are to be investigated, the new rows
and columns can be entered as indicated in Table 1.
The above procedure is done to obtain a square matrix
of coefficients with dimension 3N + 4. Any linear technique
can be used to solve this easily with a given right-hand-side
vector. For example, an in-house package developed by
Koko was used at Bucknell, and PC-MATLAB was used at
Villanova. The IMSL package has been used successfully at
both places, and student-programed Gaussian elimination
can also be used, but the packages are very convenient.
These methods have worked in systems with up to ten
stages with no major difficulty.
The matrix of coefficients for the two-stage, backward
feed problem is shown below. The coefficients of L and V in
Eq. (1) are always 1 or 1, thus only the coefficients of Eq.
(2) and (3) need be calculated.

h2 = 221 kJ/kg (at 59C and 0.09)
h = 12 kJ/kg (at 3% and 0.05)
H1 = Hs + SH1 = 2619 + 1.88(60) = 2732 kJ/kg
H2 = Hs2 + SH2 = 2604 + 1.88(2) = 2608 kJ/kg
A0 = 2073 kJ/kg (at 163%C)


hi = 819 kJ/kg (at 126.3C and 0.60)
A = 2345 kJ/kg (at 66.3C)
(A + SH), = 2345 + 1.88(60) = 2458 kJ/kg

(U2BPR ) = (2836)(60) = 170160 kJ/hrnm2

The U's and BPR's are known from the preliminary cal-
culations. For calculation of superheat values, heat capacity
of water vapor was taken as 1.88 kJ/kg*K. Liquid enthalpies
were obtained from sodium hydroxide/water data [1, 2, 3],
and vapor enthalpies and heats of vaporization were ob-
tained from saturated steam tables [1, 3, 4].
The matrix is then solved by a linear equation solver.
The answers for the first iteration are shown in Table 1,
below the matrix. One can see that these values are very
close to the guesses, i.e. T1 = 125.9 compared to the guessed
value of 126.3. Thus, for a hand calculation, one iteration
has been sufficient.


This approach yields an algorithm that can be made ex-
tremely flexible and accurate. By far the most attractive
feature, besides its simplicity, is its inherent stability when
fully computerized.
Students have used the hand calculation method success-
fully, and they agreed that this method was much easier to
use and to understand than the traditional trial-and-error
method. We provided students with a simple computerized
linear algebraic equation solver that allowed changing indi-

vidual coefficients and re-solving the matrix. We found that
at least two hours of lecture were required to present the
concepts of an evaporator series, the balance equations, and
the tableau into which they had to place the coefficients.
The hardest concepts for junior chemical engineers to under-
stand were boiling point rise and the effect of superheat,
apportioning the available AT for the initial guess, and get-
ting the correct enthalpies when significant heat of solution
effects were present.
In the fully computerized method reported in detail else-
where [8, 9], we found that convergence could be monitored
by any single-valued function, such as the steam rate. The
fully computerized method converged from any starting
point and was at least an order of magnitude faster than
nonlinear equation solving techniques, if the latter con-
verged at all. The algorithm did not need a traditional equal
boilup start and easily handled 30-stage forward and back-
ward-feed systems with complex liquid behavior like that
exhibited by sodium hydroxide and water mixtures.
In addition, the computer routine can be incorporated
into a search program to find an economic optimum design.
The ratio of areas, Ri, may come into play here, if area were
to be used to compensate for varying U, for example, or
BPR changes.


The algorithm solves for the classical variables of in-
terest, viz., liquid and vapor flow rates, the temperature
and composition in each stage, the size of an individual

Matrix of Coefficients for the 2-Stage,
Backward Feed, Sodium Hydroxide Problem

order of vari

ATo Vo L1 AT1
0 0 -1 0
i = 1 0 -2073 819 0
567 -2073 0 -567
0 0 0 0
i=2 0 0 0 0
0 0 0 2836
(higher stage equations inserted here)

S0 0 1 0
boundary 0 0 0 0
conditions 1 0 0 0
0 0 0 0

Solution 211800

13360 1670 162920
T,= 125.9




0 -1 1
0 2608 -12
-2836 0 0

0 0
0 0
0 0
0 1
11910 76660
X2=.086 T2=58.9


(higher stage
inserted here)

0 0
0 0
0 0
0 0
0 0
-170160 0





evaporator (heat transfer area) and the required flow rate
of externally supplied steam. Practical consequences of the
design can also be investigated by looking at other variables
not normally of interest, such as heat transferred in each
effect, pressure in each effect, etc.
Using the computer program, two design problems were
made very apparent. One is what we term "boiling point
rise failure," and the other is what we term "sensible heat
demand failure." In systems with substantial boiling point
rise, as the number of stages increases the sum of the boiling
point rises also increases. It can increase to the point of
exceeding the overall temperature driving force (the tem-
perature difference between externally supplied steam and
final effect saturation temperature), in which case the sys-
tem breaks down [1]. The output of our algorithm showed
a negative area appearing when this boiling point rise failure
occurred. This is likely to happen in systems having a signif-
icant boiling point rise, such as sodium hydroxide/water,
when the number of stages is large. The maximum number
of stages where this occurs cannot be computed easily by
hand, but the computer method can do evaluations of all
cases from 1 to N stages, from which such information can
be readily seen.
A sensible heat demand failure occurs in a stage when
the incoming steam gives up all its heat to raise the temper-
ature of the incoming liquid to its boiling point, and none is
left over to evaporate the liquid. This can happen in the last
stage (feed stage) of a backward-feed system, but it cannot
happen in a forward-feed system [9], because temperature
is highest in the first (feed) stage and drops in each succes-
sive stage. This condition has not been generally recognized
as a cause for failure in evaporator systems, but it can occur
with fewer stages than would cause a boiling point rise fail-
ure in backward-feed systems with a large boiling point rise.
As the cost of energy increases, evaporator trains with more
stages will be economical, and both failure modes could be-
come even more important. One cannot have an arbitrarily
large number of stages in backward feed situations, or in
systems with a large boiling point rise. Some evaporator
systems, therefore, cannot have unlimited economy, the
ratio of total vapor evaporated to externally supplied steam.


The flexibility of the computer method is such that differ-
ent combinations of independent and dependent variables
can be used. For example, one could have A as an independ-
ent variable and solve for PN (the pressure in the last stage).
Also, functional dependencies of some of these variables
relative to A may be of some interest. One could use the
fully computerized model to find PN as a function of A by
selecting several values of PN and solving for corresponding
values of A, plotting these and backing out PN as a function
of A. In a similar manner one could back out N, the number
of effects, given A, and so on. Thus, this algorithm can be
used to calculate how many effects of a given size would be
required to do a given evaporation problem.

The model could also be used as the performance function
of a non-linear optimization program in which the independ-
ent variables, including Ri's, are treated as the design or
search variables and some estimate of overall cost (capital
plus operating costs) is used as the performance function. If
pricing functions for To and PN are available, they could be
included as search variables for the economic design.


In conclusion, the reasons for using this approach are

The functional form of the equations is readily apparent. The
method can be taught easily and does not get much more compli-
cated as the number of stages increases beyond two.
Non-linearities in enthalpy or boilingpoint rise are inconsequential
to the algorithm and easily incorporated.
Physical limitations on the system can be readily identified.
The solution scheme involves simple numerical methods, such as
successive substitution with Gaussian elimination for the solution
of the simultaneous linear algebraic equations at each iteration.
Small-N problems done by hand rarely need a second iteration.
The algorithm may be used in complex economic optimization
routines. The method encourages investigating other relationships
in the evaporator system.
The fully computerized algorithm is unusually efficient and stable,
with very favorable convergence characteristics shown even where
non-linear methods fail to converge.
This approach fits in well with computer methods for staged opera-
tions and may be very useful in that area, because it shows how
some non-linearities can be handled simply and efficiently.

The potential usefulness of this method is exciting be-
cause, (a) it is easy to teach as our experience has shown,
(b) it inspires investigation of other evaporator problems
previously ignored as too difficult or not worth the trouble,
and (c) it might have a general application in other unit
operations problems, such as absorption and non-ideal distil-
lation analysis.


1. McCabe, W. L., J. C. Smith and P. Harriott, Unit Operations of Chem-
ical Engineering, 4th ed., McGraw-Hill, NY, 1985.
2. Foust, A. S., L. A. Wenzel, C. W. Clump, L. Maus and L. B. Andersen,
Principles of Unit Operations, 2nd ed., J. Wiley & Sons, Inc., NY, 1980.
3. Geankoplis, C., Transport Processes and Unit Operations, 2nd ed.,
Allyn & Bacon, Boston, MA, 1983.
4. Perry, R. H., D. W. Green, and J. O. Maloney, Perry's Chemical En-
gineers' Handbook, 6th ed., McGraw-Hill, NY, 1984.
5. Itahara, S. and L. I. Stiel, I&EC/Proc. Des. Dev., 5(3), 309-315, (1966).
6. Itahara, S. and L. I. Stiel, I&EC/Proc. Des. Dev., 7(1), 6-11, (1968).
7. Bhatt. B. I., et al, Chem. Age India, 21(2), 1135-1144, (1970).
8. Lambert, R. N., D. D. Joye and F. W. Koko, Jr., Ind. Eng. Chem.
Research, 26(1), 100-104, (1987).
9. Koko, F. W., Jr. and D. D. Joye, Ind. Eng. Chem. Research, 26(1),
104-107, (1987).
10. Seader, J. D., Chem. Eng. Education, 19(2), 88-103, (1985).
11. Denn, M. M. and T. W. Fraser-Russell, Introduction to Chemical En-
gineering Analysis, J. Wiley & Sons, Inc., NY, 1972. D



Departmental Sponsors

The following 153 departments contributed to the support of CEE in 1988 with bulk subscriptions.

University of Akron
University of Alabama
University of Alberta
University of Arizona
University of Arkansas
Auburn University
Brigham Young University
University of British Columbia
Brown University
Bucknell University
University of Calgary
California State Polytechnic University
California Institute of Technology
University of California (Berkeley)
University of California (Davis)
University of California (Los Angeles)
University of California (Santa Barbara)
University of California at San Diego
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
Dartmouth College
University of Dayton
University of Delaware
Drexel University
University of Florida
Florida State University
Florida Institute of Technology
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
Johns Hopkins University
University of Kansas
Kansas State University
University of Kentucky
Lafayette College

Lakehead University
Lamar University
Laval University
Lehigh University
Loughborough University of Technology
Louisiana State University
Louisiana Technical University
University of Louisville
University of Lowell
University of Maine
Manhattan College
University of Maryland
Massachusetts Institute of Technology
McGill University
McMaster University
McNeese State University
University of Michigan
Michigan State University
Michigan Technical University
University of Minnesota
University of Missouri (Columbia)
University of Missouri (Rolla)
Monash University
Montana State University
University of Nebraska
University of New Hampshire
New Jersey Institute of Tech.
University of New Mexico
New Mexico State University
University of New South Wales
City University of New York
Polytechnic Institute of New York
State University of N.Y. at Buffalo
North Carolina A&T State University
North Carolina State University
University of North Dakota
Northeastern University
Northwestern University
University of Notre Dame
Nova Scotia Technical College
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
Princeton University
University of Puerto Rico

Purdue University
Queen's University
Rensselaer Polytechnic Institute
University of Rhode Island
Rice University
University of Rochester
Rose-Hulman Institute of Technology
Rutgers University
University of Saskatchewan
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
Stanford University
Stevens Institute of Technology
University of Sydney
Syracuse University
Tennessee Technological University
University of Tennessee
Texas A&M University
University of Texas at Austin
Texas Technological University
University of Toledo
Tri-State University
Tufts University
Tulane University
University of Tulsa
Tuskegee Institute
University of Utah
Vanderbilt University
Villanova University
University of Virginia
Virginia Polytechnic Institute
Washington State University
University of Washington
Washington University
University of Waterloo
Wayne State University
West Virginia College of Grad Studies
West Virginia Institute Technology
West Virginia University
University of Western Ontario
Widener University
University of Windsor
University of Wisconsin (Madison)
Worcester Polytechnic Institute
University of Wyoming
Yale University
Youngstown State University

If your department is not a contributor, write to CHEMICAL ENGINEERING EDUCATION,
do Chemical Engineering Dept., University of Florida, Gainesville FL 32611, for information on bulk subscriptions.

Union Carbide

Union Carbide Corporation is one of the nation's major
industrial companies. We employ more than 50,000
people at 700 plants and facilities in 37 countries, and
market our products in about 130 countries.

Over the past few years, we have become a more
streamlined, simplified and focused organization,
structured to respond effectively to rapidly changing
economic and market conditions.

for Success

Our principal businesses include chemicals and plastics,
industrial gases, and carbon products.

Union Carbide is one of the world's leading producers
of ethylene oxide/glycol and polyethylene. We have
the widest range of solvents of any U.S. chemical
company and are the largest domestic supplier of
industrial gases. We're also the world's largest producer
of carbon and graphite electrodes for steel-making.


S. .*
*y I/ t *


~*: U


Department of

39 Old

Danbury, CT

An Equal


Full Text