Asset Management

Incremental Thinking Won't Solve Automation Challenge

Automation's Off the Critical Project Path, but More Work Remains

By Keith Larson

HUG2014 header

As the project execution arm of the global energy giant, ExxonMobil Development Company typically is involved with more than 100 active projects around the world at any given time. And while automation engineering is a critical aspect of them all, it's an "infrastructure" activity that is wholly dependent on other engineering disciplines for its design inputs. Consequently, it often rests on the critical path for project completion. Sandy Vasser wants that to change.

On behalf of ExxonMobil and the rest of the user community, Vasser several years ago began challenging the company's automation suppliers to throw away traditional thinking in order to address its most pressing project needs. "Because automation hardware and software design is totally dependent on plant design data from other disciplines, input data frequently changes throughout the project design cycle," Vasser explained in his closing keynote address to the Honeywell Users Group (HUG) Americas Symposium this week in San Antonio, Texas. "Late input changes put us in continuous recycle mode on our engineering designs and drawings."

Vasser leads a group of some 120 design engineers responsible for the electrical, instrumentation and automation aspects of ExxonMobil's increasingly large "mega" projects. "Despite increasing scale and complexity, little time has been added to project schedules,” Vasser said. "Whereas things used to be done more sequentially, now all disciplines are working full steam ahead from the start."

He described the complexity and inflexibility of time-honored I/O design and commissioning practice, such as engineered junction boxes and marshalling and controller cabinets that result in 15 to 25 wire terminations for each instrument. Weight, floor space, cooling load and tedious factory acceptance tests (FATs) are also the target of ExxonMobil's campaign to simplify, streamline and, where possible, eliminate system components and steps along the way. "We realized we couldn't continue to do automation this way and be successful," Vasser added.

A Wish List

To begin to address these issues, the company brought its suppliers together and challenged them with a 13-item wish list of priorities:

  1. Eliminate, simplify and/or automate steps in the automation execution process;
  2. Minimize custom engineering;
  3. Shift custom engineering to software and rely on standardized hardware components;
  4. Use virtualization to separate hardware from software; validate software independent of hardware. Eliminate hardware FAT;
  5. Prevent design recycle and hardware/software rework;
  6. Eliminate components not necessary in the system architecture and standardize those that remain;
  7. Eliminate or minimize physical, data and schedule dependencies with other disciplines;
  8. Simplify the configuration of interfaces with third-party packages;
  9. More easily accommodate even very late changes;
  10. Mitigate the effects of software and hardware version changes;
  11. Eliminate, simplify and/or automate generation of required documentation;
  12. Manage alarms and ensure cybersecurity by design;
  13. Challenge traditional approaches and solutions.

Vasser further noted the persistent barriers to achieving these goals, including the impulse to focus on "perfecting" current processes. "Improvement only results in incremental change," he said. "We have to think differently if we're to transform the way we do things."

Progress to Date

Although work remains to be done, progress toward ExxonMobil's goals has been made. New I/O technology, such as Honeywell Process Solutions' Universal I/O, together with its cloud engineering environment and lean engineering for automation projects (LEAP) methodology showcased at this week's HUG Americas event, are making a difference.

"So far we've eliminated 66% of the components in the system and 60 to 70% of the wire terminations. We've eliminated marshalling cabinets altogether and have fewer controller-only standard cabinets," Vasser said.

Other forward progress includes:

  • Reduced heat load, space and weight;
  • Reduced quantity of spares;
  • All system design customization now in the software;
  • Standard junction boxes, with easily configurable I/O, ordered by part number; and,
  • No hardware FAT.

Key items that have now bubbled to the top of Vasser's wish-list include:

  • "DICED" I/O that allows systems and instruments to auto-detect, auto-interrogate, auto-configure, auto-enable and auto-document;
  • A standardized and simplified interface for electrical systems;
  • Multivariable transmitters that, for example, convey flow, pressure and temperature data over a single cable via a single process penetration;
  • The direct programming of safety system logic by translation from cause-and-effects tools;
  • DC power to eliminate inverters.

"It's the packaging of many technologies that will achieve the necessary outcomes," Vasser said.

Meanwhile, the company is benefiting from many strides forward in technology and project methodology. "We've taken automation off the critical path."