Imagine it’s that time to proactively schedule your annual physical exam with your doctor. During the appointment, after asking a series of questions about your habits and lifestyle, the doctor conducts the exam and gives you a clean bill of health. A few months later, however, you realize something is not quite right. You contact the doctor’s office and make an appointment. After performing thorough medical tests, the doctor reveals you have a serious, life-threatening, medical condition. Blindsided by this diagnosis, racing thoughts overtake your mind. You start asking yourself, “How could this have happened? Why wasn’t this identified during my annual physical? How did the doctor miss this?”

Similarly, many organizations are blindsided when asset failures occur, along with their unanticipated repair costs and downtime. These failures not only disrupt efficient and effective operations, but also significantly erode profit margins. The unplanned downtime can turn minutes into hours, hours into days and days into weeks. This results in missed productivity targets and, most importantly, an inability to meet customers’ commitments.

Preventive Maintenance

The primary objective of maintenance is to provide the maximum availability and efficiency of company assets throughout their expected life cycles at the lowest possible cost. Preventive maintenance (PM) is routine maintenance performed at set intervals to minimize the potential of unplanned downtime. This includes general upkeep of equipment through the replenishment of consumables, such as oil and lubrication, or the replacement of components subject to wear and tear.

Many organizations defer to original equipment manufacturer (OEM) recommendations as the basis of their PM strategy. These OEM “suggestions” are usually highly conservative and, in many cases, not based on operational and historical data. They are estimates based on the design and recommended operation of that equipment.

This “cut and paste” approach may actually have the opposite effect and may be giving organizations a false sense of security. They may be performing premature maintenance, which results in the commitment of time, money and resources that is truly not needed. They also may be performing intrusive maintenance, which has the potential for imparting maintenance induced failures or infant mortality. If asset failures are experienced between scheduled maintenance intervals, a strictly time-based strategy may not be the right approach for many assets and certain failure patterns.

Challenge the PM Strategy

Many organizations conduct a daily schedule deviation meeting to verify that the scheduled PMs were performed (a good rule of thumb for PM compliance is +/- 10 percent of the due date) and/or determine why there were deviations to the schedule. Yet, some of these same organizations do not take the time to review and assess the effectiveness of their PMs. These are the organizations that have become complacent and unwilling to objectively poke holes in their strategies and strengthen them.

Organizations should implement formal review processes to periodically analyze and evaluate maintenance programs to ensure these are, in fact, achieving the primary objective of minimizing unanticipated asset failures. PM activities and procedures should be reviewed to ensure they are not only accurate, but also still relevant and scheduled at the proper frequency. This is paramount to ensuring that asset maintenance activities are value-added and cost-effective.

PM Optimization

PM optimization (PMO) is a structured, continuous improvement process with the objective of balancing maintenance costs and the risk of failure. PMO evaluates and refines existing PM activities to verify that they are truly adding value by identifying potential gaps in PM performance and frequency. PMs may be eliminated or their set intervals revised to positively influence cost-effectiveness and asset reliability.

PMO is essentially a performance evaluation of PM tasks using historical data to assess whether the tasks eradicate failures as originally intended. The process also addresses whether the failure actually needs to be prevented and if there are better techniques to perform the same tasks now. This comprehensive approach may result in the elimination of redundant tasks, optimization of existing tasks, or the addition of tasks that were lacking from the original set. As a result, there should be no ambiguity in the PM program.

Data-Driven Approach

An effective PMO process demands the evaluation of historical and real-time performance data on assets. Organizations can achieve maximum value with precise and consistent data derived from their computerized maintenance management system (CMMS). Granted, it is critical that organizations employ remediation exercises to ensure the information entered is accurate, consistent and complete. Otherwise, they will not reap the full benefits of the process.

Analyzing the performance data in context with the PMs currently being performed should be done while knowing the equipment’s failure modes. If a PM does not apply to any failure modes, then it can be deemed redundant. Similarly, a task also can be considered redundant if it isn’t feasible to perform due to equipment design or other factors. If the failure modes of a system are unknown, then a failure mode and effects analysis (FMEA) should be performed. FMEA is a methodology used to synthesize potential failure modes based on severity, expected frequency and likelihood of detection. This categorizes the different ways in which assets fail, along with their respective impact on the process.

There are varying opinions based on experience for optimizing the multitude of preventive and predictive tasks. Greasing, for example, can be condition-based or time-based, with intervals calculated using complex formulas. Cleaning tasks can be revised based on observations from plant personnel. In the absence of good data, a 6:1 ratio can be applied to most inspection tasks. This is sometimes called the golden ratio of maintenance or the John Day ratio. John Day, a manager of engineering and maintenance, proved this ratio at a certified world-class plant. In essence, his experience found that there should be one follow-up task for every seven inspections. Assuming an exponential distribution, this yields approximately 85 percent effectiveness.

Figure 1: P-F interval approach (Source: Troyer, D. Plant Reliability in Dollar & $ense Course Book, page 330. Sigma Reliability Solutions, Tulsa, OK.)

Another approach is the P-F interval. Provided the data is available, the P-F interval can be used to challenge and revise PMs by illustrating deterioration over time. In Figure 1, the horizontal axis represents time, while the vertical axis represents condition. As an asset gradually deteriorates, it surpasses several points before total failure. “P” is the point on this curve when it is first possible to detect that a loss of function is imminent. “I” is the point of failure inception. A failure cannot be detected prior to “I,” so all maintenance activities prior to this reside in the proactive domain. If a flaw remains undetected and unmitigated through P, the deterioration process accelerates until the asset experiences functional failure (F). The time range between P and F is referred to as the P-F interval.

The general rule of thumb is to inspect at 1/5 the P-F interval for critical equipment and 1/3 the interval for the balance of the plant. Again, assuming exponential distribution and, for example, a three month P-F interval, the effectiveness is:

a.Yearly inspection < 1%;

b.Six monthly = 13.5%;

c.Quarterly = 36.8%;

d.Monthly = 71.6%;

e.Twice Monthly = 85% (correct for critical equipment – may warrant online monitoring if a sensor is available).

Mean time between failures (MTBF) and mean time to repair (MTTR) are other useful metrics to help establish PM frequencies. MTBF defines the average time an asset is operating between breakdowns or stoppages. The equation is:

This duration only includes operational time between failures and does not include repair time. MTTR is the average time it takes to troubleshoot and repair a failed component or asset and return it to operational condition. The equation is:

It’s also critical for organizations to supplement this data with the legacy knowledge of the operators and maintenance technicians. By engaging the individuals that operate and maintain the assets, they can validate the data, identify improvement opportunities, and provide a complete and accurate picture of what asset failures are occurring.

Figure 2: Formal PM tasks review process

With this historical performance data, a formal PM tasks review process can be implemented, as shown in Figure 2. Starting with all existing PMs, each PM should be screened to eliminate redundant tasks. Each PM should be reviewed to ensure that the executing work centers are accurate. With a final verification of work instructions against the OEM’s procedure, the work orders should be enhanced to remove ambiguity. The finalized revisions, once implemented, should be routinely assessed to ensure the anticipated improvements are achieved.

Conclusion

The majority of organizations have recognized the need for a proactive PM strategy to increase the overall performance of their plants. However, once implemented, organizations must review, reassess and validate that the PMs are adding value and truly preventing unanticipated failures. Demanding continuous refinement, PMO ensures the right work is performed at the right frequency in the most cost-effective manner.