Printer Friendly

Process control education: are we teaching the right stuff?

Process Control Education: Are We Teaching the Right Stuff?

The following text is taken from a letter by W.L. (Bill) Bialkowski, Chairman of the Process Control Subcommittee on Education, Canadian Pulp and Paper Association (CPPA), written to all deans of Engineering and Applied Science across Canada in the spring of this year. He is also president of Entech Control Engineering, Inc., Toronto, which provides consulting services to the pulp and paper industry in addition to an industrial training programme aimed at improving skills of working engineers in the area of process control. Bailkowski is very concerned that the process control material being taught to engineering students does not provide the necessary theoretical basis for meeting the demands of control engineering in industry. Read what Bailkowski has to say and I will close with some thoughts of my own and a suggestion for the next step.

"I am writing to you as the newly appointed chairman of the Canadian Pulp and Paper Association (CPPA), Process Control Subcommittee on Education in order to first of all apprise you of a perceived problem in the quality of engineering education as it relates to control engineering and process control, with the additional hope that this letter may lead to further constructive dialogue on the subject.

Background

The Canadian Pulp and Paper Industry is Canada's largest resource based industry, contributing over 3% of Canada's GNP and some $15-billion a year in export trade.

More and more our competitiveness is threatened by offshore competition. The industry has become a heavy user of process control technology in order to maintain its competitive position in the world marketplace. However, in spite of having installed modern process control systems in many of our 180 or so pulp and paper mills, evidence over the last decade has indicated the benefits of improved product uniformity, quality and cost have not been achieved in a sustained and consistent manner. It has long been believed that the root cause was related to staffing and training in the control engineering discipline.

Industrial Training

In 1984 the CPPA commissioned an industrial training programme with the aim of improving the skills of our engineering graduates in the area of process control, process dynamics, control strategy design, control loop tuning and the impact of control system performance on final product uniformity and quality.

In 1984 I was given the responsibility of designing, organizing and delivering the first course. I also have been associated with the delivery of the subsequent courses every year since. These were run as CPPA-sponsored courses through to 1987 and given under contract by my company EnTech Control Engineering. Since 1988 they have been given twice a year by EnTech as CPPA considered the programme to be mature and wished to initiate new training initiatives in other directions. To date, over 200 participants have taken the programme, of which about 70% were Canadian, the remainder coming from the US, Europe, Japan and New Zealand.

CPPA Course Content

The course was intially given at McMaster University as a two-week summer course.

McMaster was chosen because professors John MacGregor and Paul Taylor (Chemical Engineering Department -- control engineering professors) are widely regarded as having an excellent control engineering programme relating to industrial process control. Their expertise was used extensively in the design of the 1984 course. The course design included a full treatment of the PID, feedforward, decouping and deadtime compensation algorithms, with a particular emphasis on their practical use on a simulated paper machine. The special focus of the course was not just how to tune these loops in the perfect world of the laboratory with ideal measurements, but in the real world of the pulp and paper mill where the intent should be to minimize the impact of raw material and other induced variability on the final product through better control tuning. The control tuning objectives of the course spanned the classical quarter amplitude methods dating back to 1942 through to the most recent minimum variance control concept which has its roots in the evolution of stochastic control theory from 1965 through to 1980.

To cover the theoretical side of this material, we selected key topics from four university courses including: 1) linear dynamics systems (this is usually the only course taught to undergrads). 2) nonlinear control systems (our most difficult industrial control problems are due to nonlinearities -- this course is usually a graduate course). 3) sampled date control (all digital modern control systems sample analog signals hence their behaviour is governed by sampled data control theory -- this course is usually a graduate course). 4) stochastic control (understanding of random variables is critical to minimizing variability -- this is only given as a graduate course).

The subject material covered was thought to be the smallest `subset' necessary to adequately describe the operation of continuous industrial processes. This material is required to provide a scientific foundation for understanding process dynamics and control. Without it the subject can only be approached as a set of arbitrary `rules-of-thumb'.

Conclusions Based on Five Years of Experience

The majority of practitioners in the process control field in the pulp and paper industry (pulp and paper companies, consultants, process control suppliers) are graduates in chemical, electrical and mechanical engineering with a 4 year bachelors degree.

Their exposure to control engineering or control systems includes an introductory course in linear control theory. Electrical students may get sample data theory. Virtually no undergraduates get any nonlinear control theory or stochastic control theory.

Most of the course participants agreed they had forgotten all the control material once they passed the exam. They considered it to be impractical.

The present course programme is seen as a remedial course designed to overcome the apparent short fall in teaching control engineering at most of our universities at the undergraduate level. It appears incongruous that an industrial course should be attempting to teach theory for the first time. If this theory was well placed in the minds of the participants then industry could concentrate on achieving higher level objectives.

It appears quite feasible to modify and update existing control engineering curricula in order to present the required material without necessarily requiring an increase in time allotment.

Two professors in chemical engineering have attended the programme to date. They were impressed by the following:

-- the degree of academic unpreparedness of the participants,

-- the heavy use of simulation laboratory as a teaching aid,

-- the complexity of some of the commonly occurring industrial control problems and the need for the inclusion of sampled, nonlinear and stochastic control theory.

One of the above professors has rewritten his course curriculum based on this experience.

I have attempted to describe the problem as we see it from the standpoint of the pulp and paper industry in as succinct a manner as possible. I would appreciate your advice on this matter and in particular on concrete steps which might be taken to bring about an improvement."

Bialkowski's concerns are very real and should not be dismissed lightly. When you consider the contribution of not just the pulp and paper industry, but all of the process industries in Canada to our GNP, and the impact of process control on the quality and hence the competitiveness of these industries' products on a world scale, I think that the emphasis and content of process control in our Canadian engineering curriculum needs to be examined.

When I joined the Department of Chemical Engineering at the University of Toronto in 1986, there was one course in process control offered in the 3rd year. This course covered the linear dynamic systems and basic feedback control material which is listed in Bialkowski's letter as the only key topic usually covered in an undergraduate programme. I recognized the need for more control engineering material, and in 1987 I introduced a 4th-year elective. I was one of the professors given the opportunity last summer to attend the EnTech course on a complimentary basis to observe the level of understanding in this field among engineers working in the pulp and paper industry, and to see what material is presented in this course. In short, this was a great learning experience for me. The reactions that are listed in Bialkowski's letter are accurate. I would only add that my reaction was not that I was teaching the wrong material in my courses, but that the emphasis on its application and perhaps importance (ie. product quality) was not coming through as clearly as it should. I subsequently revised by course structures and I think I am now addressing the majority of the key topics which form the foundation for understanding process dynamics and control.

I would like to raise one other issue related to this topic of control engineering education. As Bialkowski points out, the majority of engineers who apply control engineering in industry are graduates of chemical, electrical or mechanical programmes. When I look at the undergraduate courses in control engineering at the University of Toronto, there is an incredible amount of overlap. This redundancy combined with industries needs suggest that control engineering should be taught more on an interdisciplinary basis. In fact, several European universities have departments dedicated to automatic control engineering. I think Canadian engineering schools as a whole must respond to the current and future needs of industry, of which control engineering is only one example, and perhaps be willing to re-examine our traditional divisions.

One role of the new Systems and Control Division of the Canadian Society for Chemical Engineering is to encourage appropriate education in control engineering and to provide a forum for the discussion of systems and control topics. I hope Bialkowski's letter will provoke those of us involved with control engineering education and also the process industry at large to think about this issue. I also would like to suggest that the Systems and Control Division sponsor a panel discussion at the 40th Canadian Chemical Engineering Conference with participants from both the academic and industrial community to examine this issue more closely.

Will Cluett, MCIC 1st Vice-Chair, Systems and Control Division, CSChE
COPYRIGHT 1989 Chemical Institute of Canada
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1989 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Cluett, Will
Publication:Canadian Chemical News
Date:Sep 1, 1989
Words:1673
Previous Article:General amendments to the Patent Act.
Next Article:Low-priced package for simple flowsheets.
Topics:


Related Articles
How can I afford not to teach?
Division Report: Technology Education Division).
EDITORIAL : UNFINISHED BUSINESS THE SCHOOL BOARD HAS TO PLAY CATCH-UP AFTER THE ELECTION.
The core connection: CTE and academics: a perfect fit.
End-of-year assessment.
Making teaching work; 'teaching smarter' in the post-compulsory education.
Hooray for the democrats and more accountability! NCLB is barely understood.

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