Printer Friendly

Flowsheet simulation in undergraduate chemical engineering education: now there's a controversy.

I read, first with interest and then with surprise the article by John Shaw, MCIC and D.D. Jacas, MCIC in the March issue of Canadian Chemical News/L'Actualite chimique canadienne. The use of "controversy" in the title had caught my eye and it was obvious, from the introduction, that the authors were referring to views expressed at the 39th Canadian Chemical Engineering Conference in Hamilton last October. I had attended the conference where Shaw and myself had presented papers in the same session on the use of professional flowsheet simulators in undergraduate chemical engineering symposium. I had not perceived any controversy, at least during that session, hence my interest in their obviously different perspective.

As had seemed clear at the conference, a large number of chemical engineering departments in the country use professional flowsheet simulators in their senior level process design courses. So, is not the purported controversy whether we should or not use such simulators a thing of the past? I thought so. Actually I felt that the debate (and not the controversy) had moved on to new concerns: how could we now best use flowsheet simulators for the benefit of our students? A healthy and constructive debate it was, I think. The debate was focussed, from the outset, by the introductory remarks by the session chairman, Terry Hoffman. Quoting from memory, some highlights (from Hoffman's remarks or subsequent presentations) were:

- flowsheet simulators are the professional's tools for process design and graduating engineers should have some exposure to them.

- their use expends the scope and realism of the design projects that can be tackled by students.

- freed from the tedious, repetitive and very time consuming computations, the students can invest their time and energy into the research of innovative approaches and designs.

- in a controlled course environment flowsheet simulators can be used to illustrate, investigate, experiment very effectively and at little expense.

- they can be used to help students comprehend fundamental phenomena, by a process very similar to laboratory experimentation (to which however it should never be substituted)

But (for there is a but!), there are challenges to the instructor, pitfalls to be avoided:

- time spent acquiring appropriate utilisation skills must be kept under control. Usage instruction should not be done at the expense of the study of fundamental phenomena.

-flowsheet simulators belong in advanced process design courses, not in unit operation and other courses where the students master the basic principles of operations and physico-chemical phenomena.

-simulators should not be used as black boxes; underlying phenomena should always be stressed by the instructor, the students' understanding should be verified.

- avoid wasted time and money on unnecessary or useless computations by using the simulators only in controlled situations.

- the students must recognize that the simulator is not a substitute for comprehension and judgment.

Some of those points also appear in Shaw and Jacas' list of pros and cons. Then, there may be a difference in perspective, in emphasis, 'mais il n'y a pas la de quoi fouetter un chat'. No controversy there.

However, reading further, I realised the second part was a partisan comparison of some simulators (Aspen, Process, Flowtran and Hysim) on the basis of very partial data and dubious arguments. The authors obviously consider that Hysim is a more appropriate software for undergraduate design instruction.

Having acquired and used myself all of those simulators at Ecole Polytechnique I agree Hysim is the simplest to master and use, at least at a primary level, a point which has been stated on other occasions and without ambiguityl, [1,2]. There is no doubt the developers of Hysim have done an excellent job.

The most seriously biased point made by Shaw and Jacas is that they only need "two to three hours of tutorial instruction" to bring the students to the level where they can do "further independent explorations", while they quote one instructor who "devotes 30 lecture hours for the instruction of Aspen". I believe that this is based on a statement made during a panel discussion held after the formal paper presentations and which, unfortunately, I was unable to attend. Thirty hours of instruction is far too much. It is far more than devoted to the subject in a standard introductory workshop for any of those simulation softwares (incidently, judging from the announcements I receive regularly, introductory workshops are approximately of the same length for Hysim and the other simulators).

The approach by the instructor is in my opinion what is at fault here, not the simulator. In the process design course at Ecole Polytechnique I give a three-hour tutorial lecture on Aspen before the students perform their first exercise. After that, it's pretty much learning by doing with occasional tutoring on specific points and help by assistants who are far from knowledgeable' (they are generally graduates from the previous course and have had no additional experience themselves), as deemed necessary by the quoted Aspen instructor, according to the article.

As far as the quality of flowsheet error diagnostics, here again, I am not sure that there is such an advantage for Hysim. I have done a parallel experiment which, I think, is quite revealing. Working with a fairly homogeneous group of students with no previous process design experience (this was done at Ecole Nationale Superieure de Chimie de Mulhouse, in France, during a sabbatical leave) and after about the same formal instruction on both softwares and some practice, the students were asked to propose a separation scheme for a hydrocarbon mixture (which could be done by flashes plus distillation and appropriate materials recycling). The students working with Hysim made faster progress at first and could literally play with remarkable facility with various arrangements of flash separations while their friends working with Aspen were painfully learning the syntax (but with help from adequate built-in diagnostics). But things eventually changed when the students tried to set up a distillation, got into tight specifications and.... the simulations stopped. Nobody, including the instructors could figure out why and restart the simulation (there was no diagnostics). Our only way out was to start again from scratch and rebuild the flowsheet carefully step by step. In the end, as I recall, all groups arrived at some reasonable solutions within about the same time: the Aspen groups had spent more time learning, the Hysim groups had been victims of the apparent facility ! Aspen, Process and Flowtran, are declared by the authors, as "unforgiving" programs written for "specialized practitioners" and suggest our students should only be asked to handle programs written for "neophytes". Are we to presume that easy to use, recommended Hysim is one of those? This is not at all the case: the difference between those softwares is not a matter of hierarchical complexity but of specialization. Hysim is geared toward the hydrocarbons processing industry (ie. phase equilibrium separations) and as such it is a very excellent tool. Companies operating in the field should consider its acquisition. There are other specialized softwares for example, for the pulp and paper industry, which is of great importance to Canada; one can mention Papmod the program developed by the Pulp and Paper Research Institute of Canada (PAPRICAN) and which is itself evolved from the pioneering Gemcs of McMaster; there is also Massbal and other products developed by SACDA. All are Canadian products. Larger programs can treat a broader variety of problems. For example, in addition to standard hydrocarbon processing capabilities, Aspen has extensive reactor models, the possibility to treat electrolytes and solids, to cost and dimension equipments and to perform a full project economic analysis. Some features should be of interest to some key Canadian industries such as the metallurgical and allied industries and the inorganic and specialty chemicals industries. Those features are also of interest to process design instruction. I believe that we should expend our design projects beyond the traditional scope of the petrochemical industry as more of our students will enter other fields. I find it well worthwhile to be able to use a single software which offers me a broad variety of project possibilities even if it comes with a price tag in learning time. Naturally, introducing the students to all those features takes more than a three-hour tutorial class. Additional instructions are given as they progress throughout the two-semester design course, when needs arise. Yet, including equipment dimensioning and costing I think that the amount of formal instruction has rarely exceeded nine hours for any given project. The breadth of capabilities and the flexibility which is valuable to us could be a deterrent for an industrial user with more targetted needs because of purchase cost. An interesting solution would be to offer a software by the menu, a solution which should be reasonably easy to implement since most of the softwares mentioned here (Massbal is an exception) are of the modular type. I know only one flowsheet simulator marketed on that basis, the Belgian Belsim. We have limited experience with Belsim and plan to study it in more depth since it has other interesting features (such as data reconciliation)

Shaw and Jacas have indeed written a controversial piece on what was a non-controversial issue. I do not think that they have done a service to our colleagues who may be considering the acquisition of a flowsheet simulator for their own institution by over-simplifying the matter and implying that, there is, for the time being, only a single option. The selection of a simulator should take into account a number of factors which, in the context of educational purpose, should include: equipments available for its implementation, size of user groups, approach anticipated and previous experiences by the instructors.

Factors in favour of Aspen at the time that I made the decision were its broad capabilities, my previous experience with the somewhat related Flowtran and the fact that it was operational on main-frame which made it easily accessible on the large network of terminals already installed. The major problem with Aspen on main-frame is excessive queuing times because of system overload, by far the major and almost only gripe from our students. but here again the simulator is not at fault. I had considered Hysim which is operational on microcomputers, with a hard disk but could not, at the time, be networked. So my students will continue to moan and groan until I can acquire software with capabilities comparable to those of Aspen but operational on a microcomputer network. Such a software would be ideal, considering the state of development and current needs at Ecole Polytechnique. Another solution could be more appropriate for another institution.


1. J. Paris Et A. Noel, "Experience d'un logiciel interactif de simulation sur micro-ordinateur pour l'enseignement du design en genie chimique", Congres annual ACFAS, Montreal (1986).

2. Paris, J., "La conception des procedes: logiciels sur ordinateur grand-cadre vs. logiciels sur micro-ordinateur", Congres international sur l'informatique dans les industries chimiques, Comptes-rendus, 181-6 (1987).
COPYRIGHT 1990 Chemical Institute of Canada
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1990 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Paris, Jean
Publication:Canadian Chemical News
Date:Jun 1, 1990
Previous Article:Creative and innovative problem solving using flowsheet simulators.
Next Article:An intelligent spectrophotometer for monitoring chemical vapours.

Related Articles
Flowsheet simulation in undergraduate chemical engineering education: why the controversy?
Creative and innovative problem solving using flowsheet simulators.
Chemical engineering education at TUNS.
Chemical engineering education at UNB.
Chemical engineering students at the University of Ottawa enjoy the best of all worlds.
Computing in chemical engineering.
Control performance as a unifying concept for process control teaching and learning.
On ranking chemical engineering departments.
Another look at ranking Canadian chemical engineering departments.
Multimedia courseware for chemical engineering.

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters