Editor's Note: This article is the first in a series about machinery health. To read the next article, click here.
Determining the correct maintenance strategy for industrial production assets poses a problem for many plant asset managers and reliability and maintenance engineers.
The cost and effectiveness of applicable maintenance strategy options varies and needs to be carefully considered before being enacted.
This article is the first in a series of articles describing asset criticality and applicable maintenance strategies. In this initial article, we will describe various maintenance strategies. The next article will describe proactive centered maintenance (PCM) and asset criticality in more detail. The final articles in the series will describe the advent of wireless technologies as part of the evolution of condition monitoring technologies.
Figure 1 shows an example of a Failure (P-F) curve, with P representing the point in time when the potential failure can be detected, and F representing the point in time the asset reaches functional failure. While not all failures manifest themselves in this manner, most failure modes do have technologies that can detect failures early in their failure cycle. The intent is to manage assets at the top of this curve.
While most companies strive toward managing their assets proactively, many plants often find themselves managing assets in a reactive mode. They are continually reacting to assets reaching functional failure without warning. For example, a plant may experience spare parts shortages due to limited planning time, increased overtime and callouts, and poorer quality repairs and documentation. All of this can inhibit the plant from having the time and resources to complete the repertoire of maintenance routines and move into a more predictive mode.
Best in Class Performers
Best in class maintenance and reliability performers typically manage the majority of assets as far up the P-F curve as possible. This results in strong planning and scheduling programs, with condition monitoring technologies as one of the key work identification systems driving those programs. The maintenance and reliability teams spend the right amount of time identifying failures earlier for each asset, enabling better coordination to better plan and schedule maintenance activities. Today's challenging economic environment, combined with the costs associated with HSE and regulatory compliance, makes it an imperative to optimize the return on investment for maintenance activities.
Figure 1. Potential to Failure (P-F) Curve
Maintenance Strategies
Some production assets require a combination of different maintenance strategies based on their failure modes and the consequences of those failures. Determining the correct maintenance approach for various assets can help optimize the life cycle costs of your assets.
A preventive maintenance (PM) strategy is often based on OEM recommendations for specific production assets. Preventive maintenance is performed on a specific time-based interval derived from OEM recommendations typically based on historical failure time cycles. PM includes time-based inspections for assets whose consequence of failure presents enough of a risk to warrant an intrusive inspection. Any intrusive maintenance has the potential of imparting maintenance-induced failures, often called infant mortality. Since asset failures can happen between scheduled maintenance intervals, a strictly time-based strategy may not be right for many assets and certain failure patterns.
Quantitative preventive maintenance incorporates non-intrusive predictive maintenance inspections (such as measuring belt tension) into a preventive maintenance program. It is non-intrusive, but it is also time-based and cannot be performed on an asset unless the asset is shut down for testing. Therefore, while it does not result in the typical problems inherent in intrusive inspections, it still results in a loss of production while the asset is shut down and may result in reduced product life cycle, based on wear and tear from starting and shutting down certain assets.
Predictive maintenance (PdM) and condition-based maintenance (CBM) are synonymous. A CBM program evaluates machinery via instrumentation, either periodically or continuously, to determine its condition, usually through a condition monitoring (CM) program. It allows planners and schedulers to schedule maintenance when it is most cost-effective and precedes functional failure. A condition monitoring program helps identify the required maintenance work and ideally allows reliability and maintenance professionals to program all required maintenance on a given piece of equipment to occur simultaneously, saving costly downtime.
CM programs have often been characterized as "vibration monitoring" programs. While vibration monitoring is a critical aspect of most CM programs, there are more than 75 different types of non-intrusive CM technologies, such as oil particulate analysis, temperature, thermography and ultrasonics. Asset failure modes and asset criticality will drive the determination of which technologies and what frequency data needs to be collected and analyzed on an asset-by-asset basis.
The goal of predictive maintenance is to use condition monitoring technology to detect future failures through the evaluation of early warning indicators. A robust PdM program has many benefits over other maintenance strategies (see Table 1).
Table 1
Once a robust PdM program is in place for an asset, PM routines can be reviewed and in many cases optimized or eliminated. As an example, assume that a particular OEM recommendation is to change a bearing after 30,000 hours of operation. With proper application of predictive technologies, the health of the bearing can be monitored and managed to a high degree of reliability. By managing the health and condition of the asset, time-based maintenance intervals can be extended or eliminated in their entirety.
Reactive maintenance (RM) may consume up to 80 percent of the total time and budget of companies stuck in this mode. In a reactive mode, daily maintenance activities are driven by unforeseen problems from assets breaking down without detection of the pending failure. RM, "living life at the bottom of the P-F interval curve," is maintenance performed after a failure or after an obvious, unforeseen threat of immediate failure. Running machines in run-to-failure (RTF) mode is a strategy for assets where the consequence of failure (including cost to replace), is so low that it makes sense to allow the asset to fail in lieu of spending value maintenance time doing PM or PdM tasks. Companies who do not have good CM and PdM programs often react to failures and "fighting fires," hindering the transition to a more proactive approach to managing assets.
Figure 2 shows the cost per horsepower/year for the maintenance strategies above. As the graph demonstrates, PdM is not only more effective in driving early warning and the ability to plan and schedule properly, but it is also less costly than reactive and time-based strategies.
Figure 2
Asset Criticality
Assets with a high consequence of failure are considered highly critical assets. In most cases, a continuous, on-line condition-monitoring program is warranted on these assets to protect the machine and to predict failures. Assets in the low- to mid criticality level predominantly require essential measurements (see sidebar) to monitor the health of the asset. Some low- to mid-level criticality machines also have some critical measurements. We label low- to mid-criticality assets that require essential measurements as essential assets. These assets are usually monitored with a scanning system (wired or wireless) or are monitored with a portable data collection system. By definition, assets where a run-to-failure strategy is ideal due to the low consequence of failure are not considered essential assets.
Conclusion
Most plants want to make the right measurements on the right assets at the right time. Most companies use a combination of maintenance strategies, depending on the failure modes and the consequence of failure. Best in class performers typically use more predictive maintenance than other customers.
Pumps and Systems, May 2009