Why PdM programs fail and how to keep yours alive
In brief:
- Many a PdM program falls flat after the initial excitement diminishes due to many reasons.
- The key to a greater degree of success lies in how your PdM program is defined, created, and launched.
- Using all the tools at your disposal, CMMS, work management process, and diagramming software often and early in the PdM process is important for attaining the program’s goals.
Making predictive maintenance (PdM) an important part of an overall maintenance and reliability strategy enhances uptime. The decision to proceed with such a program is easy. How one places this program into effect and perpetuates it is the challenge.
SMRP Conference Ed Espinosa, program manager, CMMS, at Puget Sound Energy, will present “How Predictive Maintenance Work Gets Done” at the Society for Maintenance & Reliability Professionals Annual Conference in Indianapolis on Oct. 16 at 9:45 AM. The presentation will provide a high-level overview of a process for executing a practical and successful PdM program. It will contain essential components, as well as potential pitfalls to avoid to increase the chance of achieving a robust PdM program. Learn more about the SMRP Conference. |
Integrating technologies into a facility’s reliability strategy requires commitment and follow-through to attain the benefits of PdM. Expending scarce resources to perform PdM for the sake of meeting a corporate objective is meaningless, unless the data collected from PdM is used to make business decisions to optimize plant reliability and availability, the ultimate goal. The crux of this decision-making isn’t knowing whether to conduct the maintenance, but knowing when to do the maintenance. Many a PdM program falls flat after the initial excitement diminishes due to many reasons. PdM programs fail because:
- best PdM practices consistent with mainstream MRO trends are not recognized in multi-facility organizations
- PdM technology and practice consistency is lacking
- there’s no organization-wide central function for reporting, analysis, and results-archiving whereby plant personnel can view their data, as well as comparative data from other plants
- the PdM process is not well-documented and supported by various software titles and internal corporate processes, and as a result there is reliance on a key individual for continued program success
- PdM tasks are performed outside of the CMMS
- PdM tasks aren’t subject to a maintenance management process whereby planning and scheduling increase PdM success.
The key to a greater degree of success lies in how your PdM program is defined, created, and launched. First, an organization has to decide whether to seek the assistance of outside subject matter experts well-versed in state-of-the-art PdM technologies. There are advantages to doing this, such as knowing that the experts bear the burden of developing and implementing a viable turnkey program. But this convenience comes with a high price tag. The other option is to create a homegrown, self-managed PdM program. Depending on the individual situation, one approach will be better than the other. For the sake of this discussion, our organization decided to develop a homegrown version.
Next, an organization needs to decide which technologies to implement. Should a facility implement all available PdM technologies? With tightening maintenance budgets, PdM technologies need to be ranked and selected based on relevance and cost-effectiveness. The resulting PdM technology set from this selection process was compared to the company’s most reliability-centric facility. Site leadership and plant technical services (PTS) had a dialog surrounding what this facility presently does, as well as what they would like to do for the purpose of coming up with a final agreed-upon set of PdM technologies going forward.
Fourteen PdM technologies were identified as appropriate for our facility. One known constraint before the onset of this project was limited resources within plant technical services. This in turn influenced how the program would be rolled out. In-house subject matter experts in each of these technologies performed intensive research probing known expert sources in these areas to formulate and write the PdM technology standards, the technical cornerstone of the project.
These many technologies each required its own ranking system to determine a timeline and overall schedule for implementation. This list was divided into two phases, with the first phase consisting of six technologies and the second phase, eight. Each phase was allocated one year to implement. The implementation consisted of creating the supporting CMMS functional location — equipment hierarchy containing equipment records if not already there, assuming a prior CMMS implementation. A good set of equipment candidates lending themselves to PdM technologies would be RCM-defined critical equipment, depending on failure mode.
The first phase of PdM in the plant rollout consisted of technologies in which plant technical services had existing expertise and could implement and support more easily. The Phase I technologies chosen were:
- balance-of-plant vibration analysis
- infrared thermography
- lube oil fluid analysis
- transformer fluid analysis
- generator partial discharge analysis
- rotor flux analysis.
The Phase II technologies were:
- on-line motor analysis
- motor current evaluation
- high-energy pipe monitoring
- plant information (PI) trending
- ultrasonic pipe wall thickness
- polarization index — Megger, combustion turbine generator vibration analysis
- transformer partial discharge analysis.
Phase I represented the “low-hanging fruit” and could, upon proper execution, materialize into scoring a first “win” with site personnel and senior management — a crucial step in establishing a foundation of trust and credibility. Phase II PdM technologies required more research, manpower resources, and costly specialized equipment.
One roadblock PTS team had to maneuver around was the procurement of portable as well as permanently installed PdM instrumentation. PTS had to get creative in this effort. This cost was not budgeted for, originally. Instead, the equipment was justified individually and was not considered part of the larger PdM program. This tactic was adopted because PTS didn’t have the reputation built up to spend such a large amount of money on such a large initiative. All of the PdM program development and half of the instrumentation were covered under plant technical services overhead. Each major purchase required a justification via a written business case. And each business case was followed by an ROI analysis. This is how a $60,000 motor tester was acquired. The belief was held by PTS leadership that this program was building for the future by painstakingly and methodically creating a culture of reliability.
The PdM program’s appeal and success were dependent upon the sharing of reference documents, technology training material, white papers, conference literature, developed standards, raw test data, analysis and conclusion, and proprietary test equipment vendor reporting software with PTS subject matter experts and plant personnel. The PTS supervisor placed these documents on a shared drive with security access for key individuals in order to improve data management and increase visibility providing line-of-sight overview in one location.
The PdM data was archived on a dedicated shared drive entitled, “Operation_Services_Predicitive_Maintenance.” Within this folder, there are subfolders drilling down multiple generations of reporting data. It is organized by month, facility, and equipment within each unit, and it has analysis results. This is a dynamic operation with a constantly growing knowledge base with each passing month. This folder structure was one of three communication enhancements on the shared drive.
The second and third communication enhancements consisted of a dedicated calendar of events where PTS engineers and plant personnel could see which dates the PdM tasks have been scheduled for performance based on the multiple-year future scheduling of maintenance plan order calls. A dedicated email address was also created to communicate notices of upcoming PdM tasks from the work management process planning and scheduling modules from the CMMS. This email address also serves as a place where transfer of data takes place between PTS and the plants, as well as transfer of data between the CMMS and PTS.
Figure 1. The PdM program was designed to be a living program, residing within the CMMS and administered by planners/schedulers.
Corporate leadership buy-in to the PdM program had been dragging since, in the past, others had tried similar efforts resulting in program failure after a key individual left. To address this flaw, the PdM program was designed to be a living program residing within the CMMS and administered by planners/schedulers via the work management process — a system for managing CMMS-created work orders (Figure 1).
The desire to perform PdM is much more convincing when real-life functional failure detection events are quoted, rather than a PdM proponent trying to sell management on a PdM program by presenting graphs and statistics. This information is particularly relevant when business decisions are made based on actual equipment material condition, rather than future equipment failure probabilities.
Even before the commencement of a formalized PdM project, the benefits of PdM technology were being realized. Several more maintenance events provided overwhelming reasons to pursue a corporate-wide PdM program. Post-outage gas turbine vibration issues were identified and documented with sufficient validity to support a vendor warranty claim. A large motor driving a pump used in the water plant was subjected to a Megger and polarization index. The results revealed eminent failure. Upon further investigation, a failed mechanical pump seal was the source of the flooded motor. Consider what would have happened had that motor been placed in operation not knowing its true condition. Partial discharge (PD) analysis technology was used at a hydroelectric generating facility. The engineer in-charge with this project felt he was very familiar with PD technology and knew the benefits the technology could deliver. The analysis of measuring PD on these generators resulted in a decision to defer corrective maintenance by five years backed up with supporting and undisputed data. The decision to expand the program across the thermal fleet can be attributed to this event.
The PdM program is supported by a three-legged stool comprising of the CMMS, the work management process, and business process diagramming. The CMMS functionality is leveraged to facilitate the performance of PdM work. The CMMS provides the digital framework and is the vehicle to further the PdM program. The CMMS contains the applications to perform, document, report on, and archive the work. CMMS is the mechanism to do the PdM work — a means of making it visible, elevating it up to the forefront of maintenance dialogue amongst management and crew. The work management process complements the CMMS, It defines the work centers, special tooling, spare parts requirements, contractors, and work center capacity, and it uses scheduling and planning as the centerpiece. The work management process takes the data within the CMMS and fine tunes what is truly relevant and presents the same information in a focused way. The business process is documented by diagramming software and is used as a process development tool tailored for each plant based on its unique manpower and logistical situation.
The CMMS and work management process are two proven attributes of modern work management and execution that are mainstream and part of any maintenance manager’s toolbox. But these two systems working together were not enough to ensure PdM work success. Work management root-cause failure analysis was needed to find out why important PdM work was sometimes getting done but quite often, not.
It was found that PdM work completion was lagging behind other scheduled work. The reasons that were given varied:
- “We don’t have the funds for doing that kind of work.”
- “The plant conditions are not met since we are a peaking plant and our operating schedule is anybody’s guess.”
- “We don’t have the people available to do this.”
- “We don’t have the equipment to do this.”
- “We have never done this in the past. Why start now?”
- “We once did this, but not anymore.”
This is hit-or-miss sporadic work planning, scheduling, execution, and completion at best. These responses from the plant personnel left too many simple questions unanswered — how, what, when, who, where?
The PdM work order process was closely observed at a high level. To get to the heart of the matter, work process questions were asked and directed at the “boots on the ground” journeymen, plant management, PTS engineers (subject matter experts), PTS engineering supervisor, and facility planners/schedulers. The following is what was discovered:
- plant technical services not familiar with CMMS
- plant personnel not familiar with PTS capability
- plant personnel not trusting “corporate” to follow-through
- plant personnel view PdM as “flavor of the month”
- ambiguity over what the expectation is.
These disconnects foster confusion and apathy. The logical way of capturing what was transpiring in order to expand upon, improve, and document was using diagramming software. It was developed to incorporate processes outside and beyond the work management process realm so that the big picture is truly recorded. Diagramming software communicates the process between all involved parties and is the core of discussion for change or process improvement. It is used to segregate roles and identify responsibilities for each process at each plant. The primary reason it is most useful for troubleshooting process problems is identifying disconnects that otherwise would continue to go unnoticed. Diagramming software is the road map to reach your destination by providing clear, thought-out, direction by “writing it down on paper” — making it visible. Together with CMMS and work management process, diagramming software adds the necessary impetus to drive PdM to reaching its stated goal.
The lubricating oil analysis process was diagramed and presented to all involved for discussion. One work order that didn’t get done was chosen for further scrutiny. It was dissected into its components. The task-list long description contains the steps for performing the work. It was discovered that the procedure lacked the sufficient detail in the chronological order necessary for successful completion. Another aspect of the work order was identifying the contributors to its success and how and when all parties involved interacted with one another. Task lists, from the plant point of view, need to identify those parties external to the plant essential for the successful completion of the work. Oil testing laboratories need to be included in the task-list details, identified by name, phone number, email address, and point of contact in some cases. If, in the future, this information changes due to a vendor change, the task list must be updated and a provision must be made in the system’s management-of-change protocol to address this. And on a much broader scale, the PTS subject matter expert also needed to be identified even if outside of CMMS.
Programs, processes, and software are inanimate objects void of human character and personality. The most important aspect of this process relative to the rest is the human interaction. We speak to and dialogue with people daily in a never-ending exchange of written and spoken communication. It is easy to succumb to tunnel vision and forget who your customers are in the heat of battle when driving to the finish line. This reckless approach results in failing to connect effectively. Keep in mind, you are not a one-person operation and therefore should never do anything that affects customer data in an information vacuum. Constantly assess how your customer is interpreting your actions so that unwarranted surprises do not occur. Keeping your direct contact in the loop of what you are doing and when you are doing it minimizes these chances. This could be a challenge connecting with people who have a dozen other things going on.
Your customer’s perception of this process may require you to create the framework for effective communication pursuant to the demands of the process changes. This framework may take the form of a site visit with a known itinerary disseminated ahead of time, weekly email status report, or just picking up the phone and chatting with the site’s stakeholder to inform them of what you’ve done so far, what you are currently doing, and what you plan to do. Don’t be careless by keeping your customer unaware of your next move. Ensure agreement of change before the change. Here are a several points to ponder before beginning and during the project:
- take adequate time to develop the diagramming process
- keep in mind the diagramming process is iterative and evolving
- know your customers
- minimize touch points, especially giving the plants “one more thing to do”
- bring in snacks
- do what you say you are going to do
- make your presence known by being a change agent
- solicit everyone’s input — don’t operate in an information vacuum
- gain confidence and trust
- find a path moving forward by being a consensus builder
- promote spirited two- and three-way dialogue.
In summary, using all the tools at your disposal, CMMS, work management process, and diagramming software often and early in the PdM process is important for attaining the program’s goals. Review your site’s work completion report keying in on PdM work to see if it is being accomplished. If not, determine why not, find a cause and a solution to remedy the cause. Using these tools will close the loop and increase the chances of all scheduled work and, above all, PdM work getting done.