While field device data, unified networks and effective visualization are powerful enabling technologies, plant optimization is more about the synergistic effects that the combination of the three can bring.
That's the collective opinion of three keynote presenters at the Rockwell Automation Process Solutions Users Group meeting this morning in Houston: Todd Lucey, general manager of Endress+Hauser USA; Barry Johnson, director of sales for Rockwell Automation's information software business; and Lisa Garrison, North American managing director of Acuite, an engineering consultancy focused on HMI development.
"Information in the field device is an enabling technology,” Lucey said. “Our Coriolis mass flowmeter, for example, provides over 100 variables, and those variables map to six basic buckets: process variables, safety, logistics, diagnostic information and asset lifecycle management, yield data and energy management, and quality and metrology data about the instrument." This digital data, however, is not always used to deliver value, Lucey said. "There are 37 million HART devices installed, yet on 80% of them we access no information other than the 4-20mA signal. But with that data, we know that we can reduce time to production and ensure and improve asset performance in operation."
"We were given a challenge by a customer in the front-end engineering design [FEED] stage of a project," Lucy explained. "They wanted to shorten the project cycle, reduce project costs and reduce risk. They said that their research showed that typical delivery times for flowmeters were about eight weeks, but with drawings, approvals, calibration testing, witnessed tests, and documentation delivery, the schedule would blow out to 27 to 30 weeks. We were able to leverage the instruments' digital data, and by being the main instrument vendor (MIV) were able to help bring the plant up one year early. That meant a huge cost savings."
"How do we do that? We partnered with Thomas Enterprise Solutions to make device data available to any authorized subscriber,” Lucey said. “As the instrument configurator builds the model number, the configurator also builds 2D and 3D drawings automatically. We leveraged that data into Intergraph SmartPlant 3D, which produces the 3D drawings as part of the entire plant documentation and checks the data against the specifications. If the device meets spec, it is colored green; if not, red."
Lucey continued: "We also worked with a major chlorine producer to reduce calibration and device failure problems. Using digital instrument data, our eight embedded engineers were able to determine what the failure points were, and we found that we could either extend calibration time or make the accuracy band tighter. We were able to diagnose problems before they happened and improved uptime. We were able to reduce electrode wastage by two-thirds, drive sensor failures to near zero and drive unscheduled work orders to almost none.
"What these examples show is that field data and associated analytics provide the platform for improved business decisions," Lucey said.
Rockwell Automation's Barry Johnson echoed Lucey and placed his examples in context. "What we have here is a huge amount of raw data, big data, and we can contextualize the data and feed the relevant information into what Rockwell Automation is calling the Connected Enterprise. For example, we know we can meet high-velocity business requirements with information that enables demand-responsive shipment and flexibility that allows decreased inventories and systems to ensure product quality. But how do you deliver?"
"What's needed to make this happen," Johnson said, "is a single industrial network technology, a single integrated architecture for plant-wide control and industry-focused software applications."
"With the data turned into information, you need to present it so that your operators can make good decisions with it," Lisa Garrison of Acuite added to the discussion. "To do that, you have to harness the power of your HMI.”
“Bad design is everywhere, Garrison added, yet operators keep operating plants. “Human operators are the most powerful and capable controllers anywhere in the plant, but any controller is only as good as the feedback it gets."
"In the 1960s, the Apollo program used lists of variables, and the Apollo 13 disaster happened because nobody noticed one of those variables changing until the oxygen tank exploded. In the 1970s, we went to minimal graphics in the first DCS systems, and it didn't help process safety. Automobile manufacturers tried digital displays, but they didn't provide the information that drivers needed. People do pattern recognition far better than computation.
"We can do far better than we have," Garrison concluded. "We can give operators the information they need in a way they can use. We can build HMIs that harness the power of data and enable optimal plant performance."