Of note in the top performers is the depth into the asset population to which these multiple PdM technologies are applied. For example:
- From 63% to 95% of rotating machines (depending on the industry) are included in a robust vibration analysis program – not just the critical equipment!
- 91% to 100% of electrical equipment is included in a robust thermography program (incidentally, 58% to 79% of mechanical equipment is also included in the thermography program at top performers, particularly smaller motors and gearboxes in packaging and similar operations)
- Lubrication analysis and contamination control practices are extensive and comprehensive
- Use of Motor Circuit/Current Evaluation (MCE) technology for drivers is extensive
- Extensive use of ultrasonics (airborne and contact) and various nondestructive testing (NDT) technologies for piping and pressurized assets is also present at top performers
- And again, only 20% to 25% of the equipment population in a top performing plant is covered by traditional, time-based, invasive PM
This last bullet is worth emphasizing a bit. After World War II, it was believed in general industry (despite knowledge to the contrary in the aerospace and airline industries and some branches of the United States military) that most equipment behaved in a time-based predictable pattern – that-is, that the probability of failure was relatively low and constant until a so-called “wear-out” zone was reached, at which time rapid and exponential increase in failure probability occurred. Traditional time-based Preventive Maintenance was designed to intervene into the equipment right before the wear-out zone was reached. See the diagram below taken from RCMII by John Mobray:
In reality, a precious small percentage of equipment actually behaves in this fashion. In fact, as the following diagram below (again taken from RCMII by John Mobray) shows, there are many failure patterns of machinery behavior, and only about 11% of the equipment in a typical industrial plant has a time-based predictable “wear-out” zone:
Note the phenomenon of infant mortality, depicted above by the initial high probability of failure upon commissioning an asset into service. About 72% of equipment in a typical industrial plant (this of course varies by industry) experiences infant mortality, while, again, only about 11% has a time-based predictable wear-out pattern. By relying predominantly on PM as a maintenance strategy for most of our assets, we are potentially adding value on a small percentage of equipment, and potentially introducing infant mortality on a high percentage of our assets – unnecessarily – doing more harm than good. I remember coming out of engineering school in the mid-1970’s and arriving at a nuclear power plant full of vigor with great ideas, and being confronted by a school of thought that held “if it ain’t broke, don’t fix it”. At the time, I thought these folks were unaware of the science of machinery behavior. It turns out that I was the one that was uninformed. They knew intuitively and based on their experience that machine failure was very likely shortly after doing work on that machine. They may not have known the engineering behind the experience, but they were right.
We are not saying that we shouldn’t do anything to our machines until they fail. We are saying that while most of our machines do not have the time-based predictable wear-out pattern, failure is predictable on a large percentage of our equipment using predictive maintenance and condition monitoring. Eliminating the unnecessary PMs and introducing PdM enhances our ability to proactively manage our assets to be more reliable, and reduces the cost of maintenance at the same time!
At the top performers, these PdM technologies are the primary work identification system. These PdM technologies are actually driving about 80% of the daily work. Again, the performance characteristics at top performers are remarkably similar regardless of industry. Here are some highlights of top-quartile work-flow:
- Over 50% of the daily work order hours are related to the PdM program
- 15% Collecting and Analyzing Condition Information
- 35% Performing PdM “Results” Corrective Work (PdMr)
- About 30% of the daily work order hours are related to the PM program
- 15% Collecting and Analyzing Condition/Operating Parameters
- 15% Performing PM “Results” Corrective Work (PMr)
- Less than 20% of the daily work orders were initiated via a traditional work request from equipment operators
Keep in mind that the use of the PdM technologies objectively identifies corrective work based on real science and real data, and the early and objective identification of machine faults, if acted upon properly, should avoid catastrophic failure and collateral damage, meaning that the repairs that are made are typically less extensive, using less labor and less parts. This all drives costs down.
Traditional work identification based largely on the “five-senses” of the equipment operators provides inadequate time to effectively plan corrective work, which handicaps schedule compliance, which undermines the credibility of and trust in maintenance on the part of the operators, and so on. The domino effect is clearly present here if the root cause of the problem – work identification - is not addressed. A top-quartile objective work identification system, based on comprehensive PdM, allows the Planners to plan the “PMr” and “PdMr” Corrective work orders. By virtue of early and objective machine fault identification, these work orders can be effectively planned because we have ample time. Once planned these work orders can be advanced to a ready-to schedule status – feeding a more effective scheduling process. This in turn allows wrench-time f the maintenance workers to approach (and in some cases exceed) 50% (note that the average wrench-time in the United States industrial plant is about 28%). This also eventually will allow the equipment operators to trust the schedule and actually prepare the work-site and the equipment for the scheduled repair.
No comments:
Post a Comment