The 16th annual Reliable Plant conference took place recently in Cleveland, coinciding with both the 2015 Rock and Roll Hall of Fame induction ceremony and the start of LeBron James' first playoff series as a Cleveland Cavaliers in five years.
Rising to that level of competition, Reliable Plant delivered a terrific week of workshops and conference sessions for attendees. Sessions were loosely broken into two tracks, one for lubrication professionals that centered on program management and best practices, and one for reliability professionals that stretched from leadership and planning to preventative maintenance (PM) and predictive maintenance (PdM) in the era of big data.
Of the presentations I was able to attend, one of the standouts was delivered by Russell Flagg of Duke Energy, who described how an extensive condition monitoring program was implemented at their Smith Energy Complex. The scale of the project was enormous: more than 30,000 sensors were deployed on more than 10,000 assets, along with dedicated analytics software, at a cost of about $112 million. Phase 1 of the project involves identifying early warning of vibration or temperature changes, using advanced pattern recognition capabilities to continuously monitor the plant's rotating equipment.
Flagg, the condition-based maintenance (CBM) program owner at the complex, spoke passionately about the ways that smart monitoring and diagnostics programs could alter traditional PdM/CBM roles and responsibilities at his location. These changes include the transition from monitoring-based programs to diagnostic-based approaches, and driving collaboration among remote teams to solve complex problems. Flagg also mentioned the importance of translating and scaling project successes to other facilities, looking at these kinds of projects not just as single-plant initiatives, but as models for potential fleet-wide implementation.
The question of why the Smith Energy Complex project was a success may have been answered by Burt Hurlock and Tyler Pietri of Azima DLI. In their presentation "Using Big Data to Improve Your Uptime," Hurlock and Pietri challenged conventional thinking on what a successful PdM project really tells you about an organization. Noting that the three most commonly cited hurdles to PdM success include undefined financial benefits, undefined operational benefits, and lack of support and resources, they observed that the failure of some PdM programs might be more cultural than mechanical, stemming in part from self-inflicted wounds in the form of undefined goals and resistance to change.
"PdM persists because the value is so easily captured," Hurlock said. "The problem is not want of benefits. The problem is failure to pursue and support change with well-communicated metrics and reporting. These are cultural issues based on deeply rooted beliefs that have to be confronted before the easy work of measuring and improving begins. Without commitment to building awareness, setting goals, and encouraging action, benefits will remain elusive."
