Printer Friendly

A new old look at computerized process monitoring.

A few days ago, my old friend Paul Moravan retired from the position of technical supervisor, responsible for chemical control, at Ontario Hydro's Pickering nuclear generating station. At the retirement party, I had an opportunity to reminisce with many of my old friends from Pickering and in the process had the inspiration for this month's column.

In the early 1970s, I worked for Ontario Hydro at the Douglas Point and Bruce 'A' nuclear generating stations. I had many opportunities to visit Paul and see the workings of the chemical laboratory at Pickering. This laboratory operates around the clock with 4 to 5 technicians per shift. In a typical week, well over 1,000 data analyses were performed.

Wayne Clarke, who was senior chemical technician, looked for methods to improve the usefulness of the laboratory data. He installed a large free-standing cruciform-shaped (similar to shape of Royal Bank building in Place Ville Marie, Montreal) cork board and had a variety of system-specific graph papers printed. The plots were organized on the boards in positions related to the reactor units and systems. As data was produced, the chemical technicians would update the graphs.

In the early 70s, this cork board produced a real breakthrough as a method to make the data visible. The station managers, shift supervisors, operators and engineers all had to make at least one daily visit to the chemical laboratory. This was the place to find out such information as the |D.sub.2~O isotopic content for the unit-2 moderator or the 3H concentration in the unit-3 heat transport system. All you had to do was come to the chemical laboratory and the data were posted on the cork board. One important thing to remember is that each and every datum point on these graphs was plotted by hand. The chemical technicians who plotted them had to be aware of the current status of each and every system. It was these people who flagged problems ... and, over the years, the station staff learned to depend upon them.

Today, in many plant laboratories, the data are entered into a computer which then takes over the job of getting the calculations and plots to those who need them. There is no need for the shift supervisor or plant managers to come to the laboratory. They can call up a display on their own terminals and get whatever information they want whenever they want it.

After looking carefully at this improvement in a variety of plants, I am not so sure that we have really gained better control. The chemical laboratory war room atmosphere has vanished. A computer can flag out-of-specification values ... or can it? Does it really have the intuition to see a trend before the system reaches the specification limit?

The computer can plot very nice graphs. They are very neat and sometimes also very colourful, but do they really provide something better than the chemical technicians at Pickering did with the aid of Wayne's graphs?

I always enjoyed my visits with Paul and Wayne at the Pickering laboratory. Every technician could tell me the exact state of any system. Today, I visit many plant laboratories, but rarely get that same feeling. Something is missing. The first thing I look for should be on the walls. Are there graphs attached to the walls showing some current data that are being particularly carefully followed? When you ask about the current state of a system, do you get a quick and direct answer or does someone automatically go to a computer for a print-out before attempting an answer? Who's really in charge, a human or a machine?

Am I a Luddite who really wants to go back to the age of graph paper and manual plotting? Definitely not. My concern is that the process of monitoring has, all too frequently, become a task of storing data into a system that can retrieve the recorded numbers either as a hard copy or a simple runs chart. This is really no more than what was being done in the 70s. The graphs may look a little tidier as they are prepared by a machine rather than by half a dozen people on shift with a variety of pens and writing styles. The computer can, at best, allow you to see one or two plots at a time rather than spread them out in a logical fashion, as did the cork boards. As the data is stored somewhere on a hard disk, does anyone really have that feel for the big picture?

The 70s are long gone and we have computers at our disposal in almost every plant situation. The methods needed to apply statistical and time-series techniques to the data are readily available and have been for many years. Using them could provide a wealth of information regarding minor trends and variations making it possible to catch small trends before they have a chance to develop into full-fledged problems.

In the past, these techniques were rarely applied to routine process data as manual calculations and plotting took too much time. By the time enough data were obtained to consider going ahead with the calculations, the data had become history. If it was too late to do anything with it, why bother doing the calculations at all.

Even with the simplest of today's computer systems, meaningful analyses can be done in real time with all graphs updated as each new data point is entered. This capability enables us to fully interact with the process and maintain much tighter control. We can do considerably more with data today than we could in the 70s, but are we demanding this? Our equipment can do it.

Good luck on your retirement Paul. You and Wayne accomplished something very worthwhile before we had computers. Now that we have them, you left us with a real challenge to use them properly rather than go back and re-invent the wheel.

Note: Other than the direct references to Pickering generating station in the 70s, the comments made do not apply to any particular plant or laboratory. They are a composite collection of comments from a large number of industrial plants across the country, and covering a wide spectrum of processes.

Marvin D. Silbert, FCIC, is Computer Review Editor, ACCN.
COPYRIGHT 1993 Chemical Institute of Canada
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1993 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:Chemputing
Author:Silbert, Marvin D.
Publication:Canadian Chemical News
Article Type:Column
Date:May 1, 1993
Words:1048
Previous Article:The basic research predicament in Canada: how to create a Polanyi in a Lindros culture.
Next Article:Leadership in management - Canada's problem.
Topics:

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