While we read about the future, we do not take advantage of the great technology that’s here today. Many existing facilities are running on automation that’s hardly been touched since it was started up years ago. New facilities shy away from “new” technology and take the safe route.
Did we become automation engineers to rubber stamp what was done 10 or 20 years ago? I hope not. Here are some great technologies that are here today which deliver proven value.
The number one issue in process plants is the disruptive cost of an unplanned shutdown. And the number one cause of unplanned shutdowns is rotating equipment. Amazingly, most compressors and pumps are not monitored. Today, there are great sensors for monitoring machinery health, either wired or wireless. The sensors feed a machinery health application that predicts failures before they happen. I know of one North American refiner that loses 7 percent of capacity to unplanned shutdowns. Think of the savings and increased safety that result from knocking a point or two off this loss.
The fundamental job of an automation engineer is to use automation to reduce variation. This typically leads to thoughts about the control system. In practice, however, control loop performance is impacted much more by the quality of the control valve. You’d be surprised at how many sloppy valves are out there trying valiantly to control things. If you are not using digital valve controllers on your valves, you really have no idea which valves are bad actors.
Digital valve controllers improve dynamic performance, but also incorporate incredible diagnostics that watch the valve over time and predict impending problems such as sticking, hysteresis, or even the loss of air to move the valve. These valve controllers get the information back either over existing wires or wirelessly using wireless adapters that mount right on the controller, get their power from the loop, and send back that valuable information.
The first controller I ever worked on was a Foxboro Model 40 large case pneumatic device. This is certainly something that time has passed by (the controller, not me!). Yet, the basic fact is that the control algorithms used in that controller—the famous proportional-integral-derivative (PID) control—are identical to what is being used today in digital control systems. These algorithms work for many loops, but there are always rogue loops that defy good control. Unfortunately, these loops usually are critical ones that can rob plant performance. Advanced control concepts like model predictive control are here today and they put the rogue loops in their place. Also, PID itself can be improved using one of the many auto-tuning packages that are available.
Left stranded
Last but not least is what I call stranded diagnostics. For the last 20-plus years, suppliers have provided HART communication in their standard 4-20 mA devices. The devices have diagnostics that help immensely during start-up and commissioning as well as when operating. Yet, millions of these devices are rendered useless because nothing is receiving the diagnostics. Again, this can be done either over existing wires or wirelessly.
So, the next time you read a futurist article, remember that yesterday’s future is here today. There are so many great things that can be done today, and most of them can be accomplished as small victories rather than big bangs. Small victories add up to dramatic performance improvement which, after all, is what our profession is about.
>> John Berra, [email protected], is the retired Chairman of Emerson Process Management
