While digitalization allows incremental innovation, the goal should be prioritized multi-year programs, approved at a high level, that allow steady levels of achievement, said McNabb’s co-presenter, Don Salyer, technical associate, corporate automation council, Eastman Chemical Co.

The two had their facts to hand.

About 10% of chemical and petrochemical assets are industrial control system- (ICS) related, yet about 30 to 50% of technical support is devoted to automation assets. Most project deliverables are automation related, and demand for automation skills is increasing within the industry.

“Following a period of heavy investment in the 1980s and 1990s, capital reinvestment today is less than 20% of depreciation,” said Salyer, “with a significant portion of the asset base more than 25 years old.”

Companies push replacement projects down the road to avoid the engineering and construction costs associated with ICS modernization, costs driven by the scope of facility, network, wiring, and instrumentation work involved. It’s also the case that ICS acquisition costs are dwarfed by annual support costs.

Pinpointing difficulties

At the event, Rick Howell, manager automation, E&P data and analytics, Devon Energy Corp., pointed out it’s increasingly difficult to pinpoint the borderline between operations technology (OT) and information technology (IT).

In 2011, Howell said, the company began an initiative meant to increase management trust in data they were presented with, including strategic use of process historian and distributed control system (DCS) technology. “We collect 6.5 million data points today,” said Howell, “but we don’t necessarily save it all.”

When it comes to specifications and standards relevant to drilling, completions, and production, Howell mentioned two: the Wellsite Information Transfer Standard Mark-up Language (WITSML), for transferring technical data; and the message queuing telemetry transport (MQTT).

MQTT is an ISO standard publish-subscribe-based messaging protocol that works on top of TCP/IP. It has tremendous acceptance, extending far outside the world of industrial automation and is used for remote connections where a small footprint is needed, or network bandwidth is limited. Howell mentioned that WITSML use can be problematic because of the various flavors and variations of it.

Keeping standards just that

At the Inductive Automation Forum, recently held in Folsom, Calif., the challenges of standards integrity were illustrated in a presentation by the two inventors of the MQTT protocol. These are Arlen Nipper, president of Cirrus Link Solutions, and a regular panelist for CFE Media webinars on digitalization topics, and professor Andy Stanford-Clark, IBM CTO for all the United Kingdom and Ireland. (Stanford-Clark’s business card also notes that he is a “distinguished engineer and master inventor.”)

This year is the 20th anniversary of MQTT’s genesis. When AT&T was deregulated, remote telemetry became a custom project, said Nipper. The question was how could message-oriented middleware be applied to operational technology? MQTT was built on TCP/IP, already known in the IT world. In the last 10 years it’s developed into a major success.

Working within IBM, Stanford Clark mentioned how challenged he was to keep others from adding things into the specification, which might have ruined its wide-ranging applicability.

Kevin Parker, senior contributing editor, Oil & Gas Engineering, CFE Media, kparker@cfemedia.com.

View the original article and related content on Control Engineering