Distributed Control

Can improved usability shift the image that "Fieldbus is too hard?"

Bubba and Mongo have enough skills to commission most devices without calling out an engineer.

By John Rezabek, process control specialist, ISP Corp.

Our instrument department experienced some struggles a few weeks ago, when it was called out to repair or replace a 1999-vintage vortex-shedding meter. The DCS wouldn't allow the engineers to update the old device's tag to reflect the more recent device's revision and capabilities.

I remember the old days (sonny), when this instrument was commissioned. Some tweaks were required for some obscure timing parameters in the H1 interface card before this generation of vortex flowmeters would work. It really wasn't a huge deal in 2000; during commissioning one has the full complement of technical resources from owner, contractor, consultant and corporate. A knowledgeable individual was quickly found at the manufacturer to guide us through this subtle customization. It didn't hurt that we were already ahead of the game in terms of loop checking and commissioning: we were right on the heels of the electricians, terminating devices. It was helpful as well to be in a part of the world where the skills and expertise of the trades were high, and a years-long project helped attract many of the most capable craftspeople in the region.

The little vortex meter had been relentlessly measuring the flow of water to the seals of some critical positive-displacement pumps ever since startup. And when instruments are reliable, you can forget little tweaks that made them work. The ones who do remember have often moved on or retired.

There comes a time when the systems (DCS) people want to go home in the evening, and adrenaline-powered 14-hour days are no longer the norm.

Fieldbus was shaped by end users who expected/demanded extraordinary reliability, and one of the ways that original H1 designs delivered that reliability was through a regimented commissioning process. You didn't just slap a device on a segment in an operating unit and expect it to automatically start jabbering on the segment. There might be some closed-loop control going on, and the introduction of a new device to the network had to be closely managed.

By design, instruments show up as "blank slates" of capabilities with little configuration. All of the function blocks and parameters defining how a device will behave on a segment are configured and stored in the host system. Devices are welcomed onto the segment  through a commissioning process where the host sort of says, "I hereby christen thee FT-84292," after which function blocks like analog input, analog output or any of the 60-odd standard function blocks in the fieldbus specification can be downloaded. The device then becomes an unrelenting vessel of process control.

There comes a time when the systems (DCS) people want to go home in the evening, and adrenaline-powered 14-hour days are no longer the norm. High-dollar corporate, consultant and contractor resources go on to the next job. And this is where fieldbus' regimented commissioning process becomes a challenge. Everyone doesn't train or even want their instrument technicians to click the mouse on a DCS engineering interface. My personal experience: Even Bubba and Mongo can be trained to be respectful of the DCS, and have enough skills to commission most devices without calling out an engineer. But many plant cultures forbid this or abhor the very thought, and the ensuing struggle and confusion around a simple device replacement has, in numerous instances, caused them to dismiss fieldbus as "too hard," fleeing back to analog technology of the 1960s.

Can these plants be brought back to the 21st century by recent fieldbus usability improvements? "Compatibility_rev," a parameter that's been around for a number of years, has been incorporated as a "mandatory" part of the most recent interoperability test kit (ITK), so in theory, in the like-device-unlike-revision case with which our techs were struggling, a "new" device can be commissioned as an "old" revision. Whether such a replacement still needs a systems engineer to click a mouse may depend on the specific host and its software revision level, so if you want it, test for it.

Beyond that, devices certified as ITK 7 and hosts registered to "Host Profile D" will someday—perhaps in a year—be tested to support a screwdriver-only fieldbus device replacement. No systems person necessary. So if the facility requires the boy or girl who mows the lawn to replace a device, that could be a possibility. Our techs were able to get their vortex meter replaced with the aid of a few phone calls to the DCS specialist. But where old plant cultures forbid techs to touch the DCS, resilient dummy-friendly help is on the way.