Chemical Industry Must Adopt Higher Automation Levels

Certainly, lots of technologies are available, observes Thomas J. Fiske, a senior analyst with ARC Advisory Group Inc. (www.arcweb.com), Dedham, Mass.

A “game-changer” Fiske mentions is advanced analytics, through which data are gathered, then visualized. This will help companies focus on significant improvement, “particularly with their ability to look at unstructured data, data mining, ad hoc analysis and predictive capability. That’s going to really help—especially in this economy,” he asserts.

Better enterprise connectivity can also help. Sasol South Africa discovered this at a coal-to-liquids (CTL) facility. There, the company used a dashboard system to connect, in near-real-time, control and business systems.

When producing CTL, the plant was a “huge consumer of steam,” states Martin Turk, director of global industry solutions for automation supplier Invensys Process Systems’ (IPS, www.ips.invensys.com) hydrocarbon-processing-industry group, in Plano, Texas. But when the plant switched production to gas-to-liquids, the steam plant’s original base load was no longer needed. “They put in place a series of dynamic performance indicators. From the top down to operators, everyone was getting information pertinent to their jobs.” The result was that, after three months with the dashboard system, “they basically flat-lined their cost of steam production,” Turk notes.

Advanced controls refined

Upgrading anything can be a major endeavor, though, so better manipulation of existing technologies such as advanced process control (APC) makes sense. An important recent advance, according to Chris Hamlin, is “the ability to embed ‘advanced’ control in the execution layer of the distributed control system.” That fundamentally shifts the role and function of APC, says Hamlin, director of global chemical industries for Austin, Texas-headquartered vendor Emerson Process Management (www.emerson.com). And that, he adds, eliminates the architectural and interface complexity usually associated with APC.

Aspen Technology Inc. (AspenTech, www.aspentech.com), a Burlington, Mass.-based automation supplier, developed an adaptive-modeling technology for APC systems. It uses automation to detect mismatches between plant control models and actual plant behavior, explains Robert Golightly, an AspenTech product marketing manager. “One primary benefit is focusing scarce engineering resources on the most pressing problems.” Once those have been identified, he notes, the technology isolates the areas within the control models that are at the root of the mismatch. Then, “the technology automates the plant, step-testing to generate accurate data, and automatically generates a new control model.”

Besides prediction and optimization, good simulation technology also can improve operator training, comments Louis Meyer, IPS director of global industry solutions, chemicals. He stresses how simulation will soften the shortage of experienced operators: It will “make sure the shrinking workforce does not hamper the ability to operate a plant safely and profitably.”

But, as Fiske also notes, “There are advances in software for monitoring the performance of regulatory [control] loops and identifying the ones that are not functioning properly.” Even so, he emphasizes that “there is still no substitute for knowledgeable, well-trained, experienced technical personnel.” The loss of their knowledge and skill can represent an enormous risk, suggests ARC Vice President Sid Snitkin. Both men are right.

However, there is also no substitute for better asset management. And the asset-lifecycle-management model holds “a lot of potential” for the chemical process industry, Fiske says. Assets must be cared for throughout the operating lifetime to ensure that expectations envisioned for the investment are fulfilled, Snitkin counsels.

Caring for physical assets necessarily means enhanced maintenance. Calling unscheduled downtime a “crippler,” Fiske suggests that companies consider using advanced tools to identify maintenance criticalities, and then characterize failure modes based on such criticalities. That allows end-users to “focus on the important things,” he says. Caring for those will improve profitability, won’t it?

C. Kenna Amos, [email protected], is an Automation World Contributing Editor.