Joanie sat in the morning operations Tier 2 daily meeting waiting to share her report. She engaged with her pump supplier to deploy a pilot project, a few dozen hazardous area-capable wireless vibration monitors. Her facility was not huge, like the neighboring plant, but reliability was a key objective, as it is for most of us whose office is “close to the iron”, as the saying goes.

The vendor’s gathering points retrieved data from the sensors, which gathered relatively high-resolution vibration data in three axes, as well as temperature. The gathering points utilized a cellular provider to “phone home,” so that each pump’s health could be observed on the internet. As a rotating machinery specialist, she wasn’t especially tuned in to any security concerns. “What could an intruder derive from this read-only dashboard?” The vendor assured her everything was secure.

When it was her turn to share her screen, there were a lot of green indications—no issues had been detected by the sensors. There were no vibrations issues, impending bearing failures, excessive temperature, or any concerning indications of wear or imbalance. Owing largely to the alignment tools, training and skills of her maintenance counterparts, all these second-tier pumps, fans and blowers were running with zero evidence of defects or impending breakdown. While it was good news for the reliability team and operations, the novel wireless sensor network was still fishing for its first good catch after over a year in the field.

Jake, who looked after instrumentation reliability, had been casting his own nets, too, but mostly hauling in seaweed. While tools existed to facilitate alarm management from instrumentation, gathering the team to review thousands of measurement applications seemed elusive. Each measurement device could have more than a dozen alerts, and his positioners had several dozen. Some of the more modern devices supported NAMUR NE 107 categories of diagnostics, but typically each alert was configurable. He could disable an alert or choose in which category the alert belonged. Was a valve’s travel deviation out of specification or  was maintenance needed? It depended on the application. Could he really do this for thousands of devices?

Get your subscription to Control's tri-weekly newsletter.

Jake recalled a near miss from a prior employer that might have been a good catch, if only the plant was enduring an unusual period of extreme cold and frozen level indications effectively defeated automatic low-low level interlocks. The weather-challenged crew believed the frozen devices, even when the other independent vote of low-low level came in from the level switch. Fortunately, only a single superheat tube ruptured when the level became perilously low. There were no injuries, only an extended unplanned outage in the dead of winter. What, if any, diagnostics would have spared the crew from believing the frozen devices?

His most modern valve positioners had diagnostics for too hot and too cold, but these were meaningless in his climate. Meanwhile his powerful pressure and differential pressure transmitters had no configurable alert of any variety for too cold or too hot. Both would be useful. There remained a plethora of other alerts and diagnostics cluttering the asset management screen, and certainly a multitude could be rationalized without engaging a multi-disciplined team like they did for process alarms.

What if we only saw the diagnostics that truly mattered?

In alarm management, we use the term consequence of inaction to prioritize alarms. Reliability specialists like Jake don’t need a meeting or committee to determine his plant’s steam drum level instruments are high on the criticality list. NE107 may tell us that a device needs maintenance, but it does not tell us whether ignoring that maintenance risks a nuisance work order, a lost batch, or a ruptured boiler tube. If you have some scars like Jake, you can likely sort out the truly crucial services whose consequence of inaction is dire. Even if you’re in a new role and not as knowledgeable about all aspects of the enterprise, you can surmise that a low battery alert is less important than plugged impulse lines.

Operators only see alarms that require them to take remedial action. Why should our asset management alerts be any different?