Intersections of Manufacturing and Process Control
Have you ever thought about how many average families have been changed by having GPS devices in our cars? Just think back to the (not so) good ol’ pre-GPS days when you pretended you knew where you were going, only to find out you didn’t, and you never heard the end of it. When those wonderful gadgets came along, they put a stop to all the arguments about directions. Yet the change snuck up on us so quietly, it went mostly unnoticed.
I see the same thing happening in our industry with variable speed drives (VSDs), including variable frequency drives (VFDs) and inverters. Yes, we all know you can reduce energy consumption while improving efficiency and process control by using VSDs to drive pumps and blowers instead of using flow restriction devices. But that’s not what I want to talk about. What I want to focus on is how much the cost of modern VSDs has dropped in recent years. Now, instead of being a rarity, it feels like they’re everywhere. Some would even say they’re where they don’t belong.
For example, let’s look at a feed conveyor, say a belt or screw type, driven by a motor with a VSD. The material being transported is not a fluid, but something like powder, granules, pellets, or even some sort of chips. This is a very common application in manufacturing.
Since the speed of the conveyor can be modulated, the feed rate could be controlled in the same fashion as liquid flow in a pipe with a valve as the final element. It would be impossible even to count the number of algorithms developed by very clever engineers to control fluid flow in every imaginable process arrangement.
Yet, controlling flows and levels of non-fluid (let’s call it “particulate”) material is commonly done with custom algorithms designed without the benefit of process control theory. Why? Because it’s a discrete manufacturing facility, and they don’t think in terms of process control concepts. It doesn’t have to be that way. I argue that the advent of VSDs — which make it possible to modulate the material flow —blurs the line between process and discrete manufacturing, allowing for the application of control theory in manufacturing plants.
Is it done, though? Generally, no. In some complicated cases, the process overwhelms the ability of plain PID control, requiring plant engineers to develop additional custom logic which may or may not work well, if at all. When a process control engineer stumbles upon a discrete manufacturing line or cell struggling with control issues, a closer look often reveals that the problem would have been readily solved at the process control level. When such a connection is made, a tremendous return on investment can follow.
Here is a real-life case: An OSB (oriented strand board) plant in northern Minnesota had been struggling with wood chip bin level control, which was affecting board quality (weight) and causing frequent line shutdowns due to level interlocks. The problem was the five-minute delay between the VSD-driven conveyor speed adjustment and the effect on the target bin level.
Application of the venerable Smith Predictor algorithm (routinely used for gas composition control with gas chromatography for PV measurement) resulted in a 7% saving in raw materials (roughly $1.5 million annually per production line). The implementation required only two weeks of the process consultant’s time.
A two-fold message
If you’re a discrete manufacturer dealing with a control problem, step back for a moment and ask yourself if an engineer from the adjacent field carrying odd-looking tools could come to your rescue. Give us a chance to look at it. The results may exceed your expectations.
For you process control professionals: Discrete manufacturing facilities present a number of interesting and rewarding process control opportunities. The recent advent of VSDs makes this statement even more relevant.