I heard from a neighboring plant, a refinery whose controls infrastructure is dominated by a legacy system that reached maturity 15 years ago, that it was looking at fieldbus for a new desulfurization unit. The plant's favorite supplier's present-generation I/O no longer supports the aging proprietary standard for digital integration of their own field devices. The plant has an option to use legacy I/O and stick with the old standard, or switch to the new I/O and install either HART or Foundation fieldbus (FF) field devices. "HART sounds like the easier choice to me," he writes, "only a one- (transmitter) card different from our old model transmitter."
I helped to install a number of his legacy, proprietary, digitally integrated transmitters back in the early 90s, and spent a few years in the previous decade trying to lead the instrument maintenance department there. It was probably more of an education for me than for them. One of the key operating principles was, "We are like a fire department. If we're all sitting around the shop, that's a good thing," i.e. no fires are good, so be happy.
Things must have been good a lot back then, because the shop always seemed to have more seats filled than empty—much to the dismay of my boss, who grew up in some Franco-Prussian military family where, I'm pretty sure, idle people were flogged. And real fire departments, I thought, spend a lot of effort preventing fires through education, inspection and enforcement of codes and standards. Then they sit around the station waiting for the bell to ring.
An interesting thing happened when we completed installation of the first wave of digitally integrated transmitters. The instrument maintenance department manpower dwindled—through attrition—from 15 technicians and five supervisors to seven plus one. Most of them had been around 20 years, so between vacations, illness and injuries, we were at least another person short about every day. Nonetheless, refined products continued to flow, and instrument reliability was arguably better.
The new digitally integrated transmitters had nothing in common with their predecessors. Having spent a few years getting the guys to use multimeters, they were now useless—at least for the new transmitters that modulated milliamps in a way very similar to fieldbus. The proprietary handhelds for configuration and re-ranging were not especially user-friendly, but the group embraced them and learned what they needed to know to get the job done.
Unlike the revamps of the 90s, the current project will add less than 10% to the population of instruments in the plant, so any choice other than the legacy standard—whether it be HART or FF—threatens to make the new plant the "island of misfit instruments." This is a potential place of mystery, where every call-out response is, "Sorry I have no experience with the new stuff," and therefore a guaranteed after-hours plant visit for the engineers who installed it. Since the plant already has some HART-smart valve positioners, you could argue that HART is a safe choice, but I'm going to speculate that the HART communicator is in a cabinet somewhere in need of new batteries. At least HART, in this case, could get a few more people proficient and comfortable in using the communicator. But what's the value in that?
The question is what's the goal? As at our factory, the overriding goal is to make lots of high-value products safely and efficiently. There's an excellent chance my neighbors can achieve this goal, even with the old-timers' legacy I/O and legacy proprietary protocol transmitters. They can also achieve it with new HART or FF devices, but perhaps with a little more pain.
The reasons to endure this pain is learning. The plants that capitalize on intelligent device diagnostics are thrilled about it and aren't likely to revert to 4-20 mA or proprietary protocols. The island of misfit instruments could become a great place for learning how the rest of the plant could be—and help shape the future when the aging systems will be replaced.