September, 1993
INTRODUCTION
In chemical engineering, it has long been the custom to require undergraduates to prepare the design of a chemical plant or some similar entity. Such a requirement has at least two purposes: First, to impose upon the students the need to utilize the theoretical knowledge to which they have been exposed in their course work, in a more nearly practical setting than is usual in the normal course of study. Second, to acclimate them to the kinds of designs and economic analyses which many of them will be called on to perform when they enter industry.
A third purpose is particularly important, in view of the current emphasis upon engineering science in the curriculum. Many students choose to study engineering because they seek "hands-on" exposure to practical problems, in contrast to the idealized versions which scientists often solve. Because of the great mass of information which the students must master, the curriculum tends to reinforce the need for generalization, and hence for the mathematical expression and manipulation of that information. Inadvertently, this draws the students away from the practical problems that attracted them into engineering in the first place. It is very difficult to strike a satisfactory balance between a thorough grounding in the basics - physics, chemistry, mathematics, and the scientific disciplines derived therefrom - on the one hand, and descriptive material concerning filters, pumps, boilers, tanks, reactors, towers, heat exchangers, and the myriad objects which make up the engineer's world, on the other. This is an important justification for our attempt to have the plant design course make up, in part, for the "hands-on" courses - machine shop, engineering laboratories, plant visits - which have necessarily been curtailed or dropped entirely from the curriculum.
The emphasis of most of current education is properly upon the work of the individual. Yet much of modern industry functions in teams, and only rarely is an individual alone in working on a project. To prepare students for this fact of industrial life, design projects are assigned to groups of students, two or three at most, who must organize the job, subdivide the effort among themselves, function effectively as a team to execute the design, prepare the written report, and deliver the oral presentation. On a few rare occasions, this has required one or two members of a team to take over the responsibilities previously assigned to others who have fallen short or dropped out. This is a requirement recognized by any engineer who has been part of an industrial organization; just as the theater requires that "The show must go on!", a working engineer knows that the job must be done, by whoever is around to do it.
Thus, the design project is more than simply another course offering. It is the logical conclusion of the education of chemical engineers at the undergraduate level, embodying a major part of the material covered in all previous chemical engineering courses, and demanding (and, hopefully, inculcating) skills and disciplines which the students have rarely needed previously. At Penn, as at many other schools, the reports both written and oral are regarded as if they were industrial reports: in effect, the results of the students' first job in "industry".
With the recent ABET decision to provide flexibility in design instruction, beginning in 1993 many curricula can be expected to shift emphasis toward a more comprehensive design experience at the Senior level. Furthermore, as the computer enables students to solve more open-ended problems throughout the curriculum, it should be possible to provide a more formal treatment of the design approach at the Senior level. At Penn, along with a few other schools, a two-course sequence has been offered at the Senior level in chemical engineering for many years, and other departments are likely to consider such a sequence in the near future.
The objective of the Fall lecture course is to provide a smooth transition into the design project in the Spring. In their previous courses, which emphasize the engineering sciences, the students have had some exposure to design techniques through the solution of several open-ended problems, often using the computer. However, they have not received training in a systematic approach to process synthesis, the use of flowsheet simulators in process synthesis, and the application of economic principles in venture analysis. These and related subjects are covered in lectures, accompanied by numerous homework problems, as summarized in Table 1.
The course begins with an introduction to process synthesis as described by Seider (1984). To summarize briefly, through a case study the students are introduced to the synthesis of reaction paths, the distribution of chemicals, the synthesis of separation trains, the synthesis of networks of heat exchangers, the insertion of power-related units (pumps, compressors, and turbines), and task integration. Then the ASPEN PLUS simulator is introduced, with emphasis on the synthesis of the reactor section of a chemical plant, followed by a separation train. Here also the approach described by Seider (1984) is used.
With one-third of the semester completed, including the solution of three problems with ASPEN PLUS, a more formal coverage of process synthesis is undertaken. Heuristics for the design of individual separators are presented together with the tree of separation-train alternatives. Then the ordered-branch search strategy of Rodrigo and Seader (1975) is described, and an illustrative problem is solved.
Next, the concepts of thermodynamic availability are reviewed according to Chapter 1 of an excellent monograph by Sussman (1980) entitled Availability (Exergy) Analysis - A Self-Instruction Manual. Then the concepts of thermodynamic efficiency and lost-work analysis are covered, via another excellent monograph by Seader (1982) entitled Thermodynamic Efficiency of Chemical Processes. The latter concentrates on refrigeration cycles (which most students do not study in their thermodynamics courses) as well as distillation. The principal sources of lost work are identified, and the students design a refrigerator that reduces significantly the sources of lost work.
This leads naturally into the synthesis of networks of heat exchangers as well as heat and power integration. First, the methods that minimize the use of external utilities are discussed, including the temperature-interval method (Linnhoff and Turner, 1981) and the graphical approach for identifying the "pinch" temperatures. A problem is solved using the TARGET II program (1987). Then the methods of stream-matching are covered, beginning at the pinch temperatures, as recommended by Linnhoff and Hindmarsh (1983). Finally, the heat loops are broken and the effect of heat being exchanged across the pinch temperatures is examined. Here also the students design a network of heat exchangers.
Since, in the synthesis of a process, the analysis of individual units often involves approximations (e.g., an overall heat-transfer coefficient), for costly units, it is important to check the approximations by the development of a more rigorous model. This procedure is demonstrated for the design of a shell-and-tube heat exchanger for which the heat transfer resistances and pressures drops are adjusted through the details of the tube bundle and the baffle spacing. Chapter 14 of Plant Design and Economics for Chemical Engineers by Peters and Timmerhaus (1991) provides excellent coverage of the design procedures. These are used by the students to design a multi-pass heat exchanger.
Throughout the course, there is the need to estimate capital and operating costs, as well as the simpler measures of profitability such as the venture profit and the "annualized" cost. Detailed cost and profitability calculations, however, are saved for later, until the topics on process synthesis have been completed, approximately two-thirds into the semester. At this point, the factored methods of capital cost estimation are described as per Chapter 5 of A Guide to Chemical Engineering Process Design and Economics by Ulrich (1984). The students are also introduced to the implementation of these methods in ASPEN PLUS. Then, in a sequence of four lectures by Adjunct Professor Robert M. Busche, the students learn the principles of venture analysis. For a fermentation flowsheet, they estimate the fixed capital investment and a cost sheet, and compute the cash flows, as well as the net present value and the internal return on investment. Dr. Busche also introduces the students to his CASH'92 spreadsheet program, which they may use to carry out similar calculations for their design projects during the Spring semester.
The Fall lecture course concludes with the scheduling of the Senior design projects, and the presentation of instructions for executing them during the Spring. The nature of the design projects and the format of the Spring course are discussed in the sections that follow.
The students are not required to purchase a textbook for the lecture course, because no existing text follows the sequence in which process synthesis and flowsheet simulation are intertwined. The Conceptual Design of Chemical Processes by Douglas (1988) is excellent in its presentation of a hierarchical design strategy using many heuristics, but it does not readily accommodate the sequence in Table 1. The heuristics are very helpful and are shared with the students throughout the Fall semester.
SUBJECTS FOR DESIGN PROJECTS
In the Fall of each year, the consultants are invited to suggest ideas for projects to be undertaken during the following Spring; interested faculty members and the students themselves suggest projects occasionally. The processes are expected to be timely, challenging, with reasonable likelihood that the final designs will be economically attractive. Project originators are reminded that student motivation and faculty enthusiasm are directly related to the feasibility and potential impact of the final designs. Potential problems must be workable by Seniors without unduly gross assumptions, and good sources of data should exist for the reaction kinetics and thermophysical and transport properties. Pertinent references should be provided. In a recent project involving the reactive distillation of mixtures with many azeotropes, ARCO provided the thermophysical property data for the ASPEN PLUS simulator. With the approval of the course organizers, the students signed a non-disclosure agreement not to share the data with others.
After a process of winnowing, an approved list is prepared, which includes one or two projects more than the required number. In making its selection, each team rates each project on the list as a first through fourth choice. Wherever possible, a team is given its first or second choice. Often, however, if none of its choices are available, a team is simply assigned a topic by the professor in charge of the course. The pedagogical justification behind this practice is that no junior engineer in industry has the luxury of picking the jobs to which he or she is assigned; jobs come up, and engineers are assigned to do the best they can with the assignments they are given.
The design projects reflect the current interests of the people who suggest them. In some cases, the projects do not involve the design of a chemical plant (e.g., the design of a heat exchange system for a fast-breeder nuclear reactor, or of a heart-lung machine). Such projects demand assistance from consultants with specific experience in the pertinent field; obviously, such problems cannot be assigned unless consultants can be found who have that experience.
Part of every design problem is the requirement to take into account environmental and safety issues. Each waste material is carefully noted, and the means and cost of disposal investigated and presented. Increased emphasis is now placed on the cost of energy, designs which avoid or minimize the handling of hazardous chemicals, and protection against processing accidents. Increasingly, projects are directly related to environmental issues; e.g., the design of a tetrahydrofuran plant to achieve "zero emissions", the reduction of NOX in boiler stack discharges, and the partial recovery of the carbon content of the CO2 from power-plant off-gases.
Table 2 provides a selection of the project titles from 1960 through 1992. The time-dependent interest in space exploration, nuclear power generation, medical technology, ecology, and improved energy efficiency, as well as a variety of chemical or petrochemical processes, will be immediately evident. A report entitled Possible Design Projects has been prepared in which over one hundred project descriptions (about one page per project) presented to the Penn Seniors over twelve years, are included. This report is available from the authors, as well as many of the design reports.
INDUSTRIAL CONSULTANTS: WHO THEY ARE, WHAT THEY DO
No chemical engineering department contains persons expert in every aspect of plant design. The progenitors of the plant design course at Penn, the late Profs. Melvin C. Molstad and A. Norman Hixson, both had ample industrial experience before and during their academic careers; still, it was obvious to them that the students' efforts would be greatly enhanced by exposure to other engineers, in addition to the faculty, with widely varied experience. Because the Delaware valley is home to many companies in the chemical processing industries, and to the consulting engineers, contractors, and equipment vendors who serve them, Penn has been able to secure the volunteer services of a body of experienced and competent engineers to serve as a source of vicarious experience for the students.
Each consultant usually spends two to four hours, one afternoon per week, on alternate weeks throughout the spring semester. Over the length of the semester, every consultant meets with several of the design groups three or four times. They provide specific answers for those students who know enough to ask meaningful questions, and offer guidance and suggestions to those whose progress leaves something to be desired. They are particularly effective in providing advice on the best choice of processing equipment (e.g., in selecting from among vacuum filters, centrifuges, and hydroclones), materials of construction, plant capacities, and start-up strategies. In the past five years, the Department has added, as an adjunct professor, Dr. Arnold Kivnick, a retired engineer who had served for over thirty years as one of the consultants. His job is to be available as a resident consultant for two days each week during the spring semester. Over the years the relationship between the consultants and the plant design course has developed to the point that the students feel free, within reasonable limits, to call upon the consultants outside of the scheduled sessions, when the need arises. The students learn that equally competent people, with different experiences, often reach quite disparate opinions on the basis of the same information. They also learn how competent people reach conclusions in the face of inconsistent data, and when insufficient information is available.
A faculty advisor is assigned to each design team. That person's experience in the specific area of the team's problem may be very limited. All Departmental faculty members have worked in this capacity at one time or another, with several serving almost every year. The faculty advisors, of course, provide their own expertise; they also provide continuity and general supervision from week to week throughout the term. Further, they use their knowledge of the interests and strengths of their colleagues inside the Department and elsewhere in the University, to direct the students to sources of information and advice best suited to their needs. As a result of having advised design teams, all of our faculty have a better appreciation of the important prerequisites to be covered in their own courses, for better utilization by the students' design groups.
One of the indirect objectives of the plant design course is to teach the students the need for information networks in the development of their projects, how to set up and be part of such networks, and how to persevere in the face of indifference or non-cooperation from potential sources of information. An experienced design engineer is well aware of the assistance provided by sales representatives from equipment and material vendors; he or she also knows which colleagues have expertise in areas of importance to the project, and is not shy about picking their brains, if necessary, to further that project. This course aims to provide the Seniors, who have worked individually throughout most of their academic lives, a taste of the experience of professional teamwork. The cooperation among people - students, faculty, consultants (and sometimes their colleagues), and sales representatives - motivated only by the need to solve design problems (within reasonable limits to the time available and the sensitivity of the often proprietary technical information sought) helps to build within the students a camaraderie with other members of their chosen profession, as well as a sense of the value of their own efforts.
The Department is gratified that several people who were students at Penn in years past, some of whom have received graduate degrees elsewhere, serve in the consulting capacity, with the permission and encouragement of their employers. Table 3 lists the consultants currently working with the plant design course, the companies which contribute their services, and the number of years they have guided our Seniors. It has long been recognized that most human achievements result from each generation standing on the shoulders of its predecessors to reach new objectives. Penn's chemical engineering Seniors know upon whose shoulders they stand.
Penn is, of course, fortunate to be located in an area in which the process industries are very active. Many other schools of chemical engineering located near major industrial centers can enjoy a similar advantage. However, any school located in an area served by a local section of the A.I.Ch.E. should be able to get help of this kind. Even one consultant from outside academic circles should bring a worthwhile broadening of exposure to the undergraduate engineering students.
EFFECTS OF THE SIMULATOR ON THE PLANT DESIGN COURSE
In bygone years, the assignment of each plant design project led to a design which satisfied the problem statement - one design. The development and availability of design simulators and the computer spread-sheet have changed this. Those tools have so accelerated the design process that it is now reasonable to require the design teams to choose from among two or more alternative designs (with the need to study all of them and to justify the choice), and to optimize the design ultimately chosen with respect to energy utilization and choice of operating conditions. In some cases, the simulator has enabled the students to arrive at more effective processes, designs that would not have been possible otherwise, with much improved profitability. Recent cases have been the reactive distillation of azeotropic mixtures, and the recovery of krypton and xenon from air in thermally-coupled distillation towers.
There is, however, a tendency for students in the 1990's to depend entirely on the simulator, sometimes without understanding exactly what it is doing. Students are urged to perform manually crucial parts of the design study; this may provide approximate results which serve as initial estimates for the simulator calculations. Occasionally, especially in fractionation calculations, the simulations take so long to converge that manual approximations - such as McCabe-Thiele plots based on key binaries, or the sketching of residue-curve maps and simple distillation boundaries - can rapidly provide useful insights into the problem, permitting the simulators to achieve more rapid convergence. More often, the manual procedures increase the students' awareness of the process details; e.g., whether more distillation trays are needed above the feed tray or below, or where phase changes are occurring. Once convergence has been achieved, it is a legitimate use of the simulator to study the effects of adding trays at various locations, and of changing the reflux ratios.
THE INFORMATION NETWORK
Throughout much of their prior course work, the students have used textbooks which present new concepts with many examples and homework exercises. In the design lecture course, as mentioned previously, individual chapters from several books are utilized to present the concepts in the sequence shown in Table 1. This helps the students to become accustomed to working with many sources of information, but does not involve them in the gathering of information from the vast literature.
At the outset of the Spring project course, the students quickly learn to access such well-known sources as the Kirk-Othmer Encyclopedia of Chemical Technology and the Encyclopedia of Chemical Processing edited by McKetta and Cunningham. More importantly, they are introduced, by our Librarian, Gretchen Sneff, to the electronic media and the data bases available, including the Science Citation Index, the Engineering Index, and Chemical Abstracts. Examples of search procedures are provided, and sources of assistance in the library system are introduced. Furthermore, the students learn that library resources at other universities can be searched through electronic mail, and interlibrary loans used to obtain sources not available locally. This relative ease of information access has a major impact on the quality of the designs.
THE WRITTEN REPORT
Since one of the objectives of this design course is to introduce the students to some of the requirements of the profession, the design report must be prepared as if it were written for an industrial supervisor (for transmittal to his superiors) by a junior engineer assigned to study a potential project. The required form is that of a typical industrial report, beginning with the letter of transmittal. The usual sections are required: abstract, introduction, process flowsheet (including a material balance block), process description, unit descriptions, energy balance, specification sheets, equipment cost summary, fixed capital summary, economic analysis, conclusions and recommendations. One specific requirement is that the report be so organized that a conscientious industrial supervisor can check the design of any particular item of equipment, from its functions in the unit descriptions, to its details in the specification sheets and its purchase price in the equipment cost summary, down to the detailed design calculations, in the form of Xerox copies of reasonably legible calculation sheets, in the Appendix.
It is expected that the preparation of the report will take a great deal of time; students are encouraged to start to write the descriptive portions while the design computations are still under way. With the advent of the word processor and the personal computer, the appearance of all the students' reports has reached a uniformly high plane. That report adjudged to be the best in the class is awarded the Molstad prize (a non-negligible cash award), and is often submitted for the prestigious Zeisberg Award, administered by the Delaware Valley Section of the A.I.Ch.E., in competition with other schools in the area.
THE ORAL PRESENTATION
A lucky junior engineer may get the opportunity to attend the meeting where his or her work and ideas are presented to the decision-makers among his or her employers, but it is rare that he or she is required to make the presentation in person. The experience of making such a presentation has been a part of the plant-design course at Penn since its inception. Each design team must present its report in public, before their classmates and as many as possible of the faculty and consultants. All team members are required to take part in the oral presentation, with each team having about forty minutes, including five or ten minutes for questions. (To set the appropriate atmosphere, the students attend in clothes suitable for a business meeting.) The salient factors in the design are discussed, including the pertinent chemistry, design problems and their solutions, equipment costs, and project economics. The use of audio-visual aids is encouraged, including transparencies and slides, with suitable projectors, and more recently, computer-screen projectors.
The oral presentations are weighted in the students' grade, and in the considerations for the Molstad prize. All faculty members and consultants present at the sessions contribute to the evaluation of the presentations.
CONCLUSIONS
The plant design course at Penn has come to be regarded (by the students and faculty alike) as the culmination of the Senior's efforts. Since the B.S. degree is still considered the professional degree in engineering, this course is so designed and conducted that the students utilize much of what they have formally learned during their years of study, in preparation for what will be, for many of them, the start of their working lives. With few exceptions, the students will have put more concerted effort into the design, the written report, and the oral presentation than into any other single event in their lives up to that time. It is considered as a kind of final exam, not in a particular course offering, but in the whole chemical engineering undergraduate curriculum. In recognition of that fact, the Department customarily invites the members of the graduating class, along with as many of the faculty and consultants as can be present, to have lunch together during the mid-day break in the presentations, to celebrate the students' success and hard-won maturity.
Linnhoff, B., and E. Hindmarsh, "The Pinch Design Method for Heat Exchanger Networks," Chem. Eng. Sci., 38 745-763 (1983).
Linnhoff, B., and J.A. Turner, "Heat Recovery Networks: New Insights Yield Big Savings," Chem. Eng., 56-70, November 2, 1981.
Peters, M., and K. Timmerhaus, Plant Design and Economics for Chemical Engineers, Fourth Edition, McGraw-Hill, 1991.
Rodrigo, B.F.R., and J.D. Seader, "Synthesis of Separation Sequences By Ordered Branch Search," AIChE Jour., 21, 885-894 (1975).
Seader, J.D., Thermodynamic Efficiency of Chemical Processes, The MIT Press, Cambridge (1982).
Seider, W.D., "The Process Design Course at Pennsylvania: Impact of Process Simulators," Chem. Eng. Educ., Winter, 1984.
Sussman, M.V., Availability (Exergy) Analysis - A Self-Instruction Manual, Milliken House, Massachusetts (1980).
Target II - User's Guide, Linnhoff March Process Integration Consultants, distributed by the CACHE Corporation, Austin, Texas (1987).
Ulrich, G.D., A Guide to Chemical Engineering Process Design and Economics, Wiley, 1984.
Table 1. Outline of Topics in the Fall Lecture Course
Lecture Hours
Introduction to process synthesis 3
Flowsheet simulation using ASPEN PLUS 11
Synthesis of separation trains 2
Thermodynamic efficiency and lost work 5
Heat and power integration 4
Heat exchanger design 2
Capital cost estimation 1
Profitability analysis 6
Selection of design projects 2
(for Spring project course)
36
Subject Year Liquid-Metal Heat Exchanger: Fast Breeder Nuclear Reactor < 1970 Recovery of Minerals from a Lunar Station < 1970 Design of a Heart-Lung Machine < 1970 Recovery of Solar Energy by Hot-Water Solar Panels < 1970 Conversion of Methanol to Gasoline in a Fluid Bed Reactor 1979 Manufacture of MTBE Anti-Knock Additive 1981 Pressure-Swing Adsorption for Separation of Air 1982 Heat Pump for Ethane-Ethylene Split 1983 Heat and Power Integration for Manufacture of Propylene Oxide 1984 Scleroglucan Biopolymer for Enhanced Oil Recovery 1985 Helium Recovery from Natural Gas 1985 Cogeneration Flue Gas Cleanup 1986 Groundwater Clean-Up and Organics Incineration 1987 Thermally-Stable Amylase Enzymes 1988 Gas Processing for Ethane Recovery 1989 Ammonia Purification by Refrigeration and Membrane Processing 1990 Zero Emissions from a Tetrahydrofuran Plant 1991 Itaconic Acid by Fermentation 1992 Ultra-High Purity Oxygen Manufacture 1992 MTBE Manufacture 1993
Name Company Years Service Dr. Rakesh Agrawal Air Products and Chemicals 3 Dr. E. Robert Becker Environex, Inc. 12 Dr. David D. Brengel* Air Products and Chemicals 1 Dr. Robert M. Busche BIO EN-GENE-ER Associates 10 Mr. Leonard A. Fabiano ARCO Chemical Co. 15 Dr. Brian E. Farrell* Air Products and Chemicals 4 Mr. F. Miles Julian E.I. DuPont de Nemours 10 Dr. Grant G. Karsner* Mobil Research and Development 5 Dr. Frank Kelly* Mobil Research and Development 2 Dr. Donald J. Klocke* Mobil Research and Development 15 Dr. Jack McWilliams* Mobil Research and Development 15 Dr. Mark R. Pillarella* Air Products and Chemicals 4 Dr. William B. Retallick Consultant 13 Dr. Henry M. Sandler Consultant 1 Mr. Andrew Savo* Rohm and Haas 2 Mr. Peter Schmeidler Rohm and Haas 15 * U. of Pennsylvania alumnus