Why automation isn't process control

Aug. 14, 2015
Products shaped by the needs of expedient automation may get us bitten by the wild process beast
About the author
John Rezabek is a process control specialist for ISP Corp., Lima, Ohio. Email him at [email protected]When exactly did the "a" in ISA change from "America" to "Automation"? It seems like 10 to 20 years ago our premier community for instrumentation and control began seeking a bigger, more inclusive tent that might expand the scope of membership and perhaps dissuade big-budget exhibitors from retreating back to their respective end-user groups. The big-spending end-user companies—primarily from large process industries like refining and petrochemical—became weary of sending their employees to what seemed like a week-long boondoggle, which was a perception not totally unfounded in reality. Attendees' propensity for food feasts, parties and swag was perhaps too big of a distraction from the biggest benefit of this community: furthering meaningful, useful and user-guided products and standards for process measurement and control systems.

In recent years, "upstream," the sector of our industry that explores, drills and delivers a vast variety of hydrocarbons, has come to the fore in the procurement of instrumentation and controls. When oil and gas prices were high, projects that stood between discovery and delivery couldn't be delivered fast enough, and while there may have been some deference to "low bidder," a lot of money was being spent. Most of the same products used in "downstream," such as control valves, pressure transmitters and control systems are employed in upstream, but I'd argue that the discipline as practiced leans more toward SCADA and "automation" than "process control."

What's the difference and why should we care? To me, the epitome of automation is a bottling line, where thousands of containers are cleaned, filled with a delicious beverage and capped at a dizzying pace that is marvelous to behold. Now I've performed the same thing manually—clean, fill, cap, put in a case—and the bottling line is pretty much doing exactly what I do manually. Every aspect of it can be precisely controlled because every aspect is precisely known, down to the beverage itself. There is "feedback" control in the servos that perform the various tasks, but they can be cranked up to blinding speeds because the motors and solenoids are engineered for a specific task. Because the things being "controlled" are engineered and manufactured, automation functions largely in an "open loop." It's just robotically doing a manual task, albeit in an amazing fashion.

See also: Process control innovations: What future process plants can be like

What distinguishes process control? In the process industries, we do our share of automation. How many motor-operated valves (MOVs) do you have? Every control valve "positioner" is a "servo." But whether you're manipulating a control valve in a closed-loop PID scheme or manually turning a globe valve, chances are you're interacting with a process, manipulating a flow, temperature, pressure or level that will have some effect on, for example, a distillation column or a chemical reaction. Boiling hydrocarbons and reacting chemicals may obey well-known properties of physics, but they're rarely engineered or consistent enough to the extent one can rely solely on robotic automation.

Control systems and their human operators have to react to measurements in real time because processes obeying physics and physical chemistry are wild by nature. The molecules we're tasked with herding through processes to produce a saleable product are mostly invisible. Unforeseen bad actors—co-products, contaminants or corrosive agents—show up all the time. While automation can do some amazing things, process control engineers are tasked with taming a wild beast.

When you're in the ring with a lion, the importance of precise, timely and reliable information—"measurements" if you will—is tantamount to not being eaten. The whole science and mathematics of process control we learned in chemical engineering classes is predicated on the regular and relentless delivery of live, truthful process data. Processes are wild creatures by nature—we need to know, our control systems need to measure: where the beast is (now), and where do I need to move to remain whole?

"A" may stand for "automation" now, and automation is part of the infrastructure we deliver as process control professionals. ISA remains an entity that delivers useful and meaningful standards, such as safety instrumented systems, alarm management, network security and intelligent device management, for the process industries. However, products shaped by the needs of expedient automation may get us bitten by the wild process beast. Let's hope our control systems suppliers ensure their core products aren't diluted or compromised in the delivery of the relentless determinism required for process control.

Controllers and Control Systems
As we look to the future, we see rapid changes coming in operator interface, computing, data analytics and virtual reality. This state of technology report shows the breadth of application of our existing control systems and demonstrates the strength of the foundation they offer for future developments. Download the report.
About the Author

John Rezabek | Contributing Editor

John Rezabek is a contributing editor to Control