Comparing Maintenance Strategies Based on Cost and Availability

[Please note that the following article — while it has been updated from our newsletter archives — may not reflect the latest software interface and plot graphics, but the original methodology and analysis steps remain applicable.]

 

Reliability Centered Maintenance (RCM) analysis provides a structured framework for analyzing the functions and potential failure modes for a physical asset (such as an airplane, a manufacturing production line, etc.) in order to develop a scheduled maintenance plan that will provide an acceptable level of operability, with an acceptable level of risk, in an efficient and cost-effective manner.

 

RCM techniques often utilize a logic diagram approach for evaluating the potential effects of failure and selecting the appropriate maintenance strategy. As an example, Figure 1 shows a portion of one of the decision-making flowcharts presented in the SAE JA1012 document, A Guide to the Reliability-Centered Maintenance (RCM) Standard. Similar diagrams are provided in other published RCM guidelines. (Some of the major RCM publications are listed in the References section of this article.)

 

In addition to, or instead of, a logic diagram approach, the RCM analyst may wish to use cost- and availability-based comparisons of potential maintenance strategies when selecting and assigning maintenance tasks. This article provides an overview of these comparison techniques along with a couple of demonstration examples.

 

Types of Maintenance Strategies to Consider

Although there is variation among practitioners regarding the terminology used to describe the available maintenance techniques, in general, the RCM analyst may consider any of the following maintenance strategies to address a potential failure mechanism:

 

• Run-to-Failure - fix the equipment when it fails but do not perform any scheduled maintenance.
• Scheduled Inspections
• Failure Finding Inspections - inspect the equipment on a scheduled basis to discover hidden failures. If the equipment is found to be failed, initiate corrective maintenance.
• On-Condition Inspections - inspect the equipment on a scheduled or ongoing basis to discover conditions indicating that a failure is about to occur. If the equipment is found to be about to fail, initiate preventive maintenance
• Service - perform lubrication or other servicing actions on a scheduled basis.
• Repair - repair or overhaul the equipment on a scheduled basis.
• Replace - replace the equipment on a scheduled basis.

• Scheduled Preventive Maintenance
• Design Change - Re-design the equipment, select different equipment or make some other one-time change to improve the reliability/availability of the equipment.

 

Using Simulation to Compare Maintenance Strategies

Given certain information about how the equipment will be operated, the probability of occurrence for the failure mode and the maintenance characteristics, the analyst can use simulation to estimate the cost and average availability that can be expected over the operational life of the equipment when a particular maintenance strategy is employed. The calculations can then be used to compare available maintenance strategies so that the analyst can select the most cost-effective strategy that provides an acceptable level of performance.

 

Run-to-Failure (Corrective Maintenance Only)

 

To estimate the cost and average availability that can be expected for a run-to-failure (corrective maintenance only) maintenance strategy, the analyst must provide the following information:

 

• The amount of time that the equipment will operate.
• The probability density function (pdf) that describes the probability that the equipment will fail due to a particular failure cause.
• An indication of whether the failure is detectable during normal operation.
• The amount of time that the equipment is expected to be down each time corrective maintenance is required. This can include the time to perform the maintenance as well as any logistical delays (i.e., waiting for labor and/or materials required).
• The cost each time corrective maintenance is required, including the downtime, labor, materials and other costs.
• The degree to which the equipment will be restored by corrective maintenance (e.g., "as good as new," "as bad as old," etc.).

 

The analyst can then simulate the operation of the equipment for the specified operating time, given the specified reliability/maintainability characteristics, in order to estimate 1) the expected number of corrective maintenance actions that will be performed and 2) the amount of time that the equipment is expected to be operating (uptime) over the specified time. These estimates can then be used to calculate the total operating cost, cost per uptime and average availability, as follows:

 

 

Scheduled Repair/Replacement

To calculate the cost and availability that can be expected from a maintenance strategy that involves preventive repair/replacement of the equipment, the following information is required (in addition to the inputs described previously):

 

• The time interval at which the preventive maintenance will be performed.
• The amount of time that the equipment is expected to be down each time preventive maintenance is performed.
• The cost each time preventive maintenance is performed.
• The degree to which the equipment will be restored by preventive maintenance.

With this additional information, simulation can be used to estimate the expected number of corrective maintenance (CM) and preventive maintenance (PM) actions, along with the uptime. The total operating cost for this maintenance strategy includes the cost of all CMs plus the cost of all PMs, as shown next. Note that the Cost per Uptime and Average Availability calculations are the same, regardless of task type.

 

 

Calculations for Service and Failure Finding tasks are performed in a similar manner except that the assumptions of the simulation will vary to fit the conditions of the task. For example, if the failure is undetectable during normal operation and the equipment is found to be failed during a scheduled Service task, then the simulation will assume that corrective maintenance will be initiated. Likewise, a Failure Finding task can initiate corrective action if the equipment is found to be failed but does not restore the equipment to any degree if it is found to be running.

 

On-Condition Inspections

On-Condition Inspection tasks (which are designed to monitor the equipment at scheduled intervals or on an ongoing basis and initiate preventive maintenance only if a specific condition is detected) require additional information and a more complex simulation/calculation method. In addition to operating life, probability of failure and corrective maintenance characteristics, the analyst must describe the characteristics of the scheduled inspections that will be performed:

 

• The time interval at which the inspection will be performed.
• The amount of time that the equipment is expected to be down each time an inspection is performed.
• The cost each time an inspection is performed.
• An indication of when the approaching failure will become detectable during inspection (which could be described as a percentage of item life or as a fixed time interval).

For the cases in which the inspection detects that a failure is approaching, the analysis also requires the downtime, cost and restoration factor associated with the preventive maintenance that will be initiated.

Simulation of this scenario will return 1) the expected number of corrective maintenance actions, 2) the expected number of inspections, 3) the expected number of preventive maintenance actions and 4) the amount of uptime. The total operating cost then includes the cost of all CMs plus all inspections plus all PMs, as shown next.

This total operating cost is then used to calculate cost per uptime and average availability as described previously.

 

Example 1 - Mechanical Component with Wearout

Consider an RCM analysis for a large truck that is intended to operate for 120,000 miles per year. A critical failure mode has been identified for a mechanical component and reliability analysis indicates that the failure behavior follows a Weibull distribution with beta = 2.3 and eta = 72,000 miles. Considering logistical factors, downtime penalties and the actual repair resources, it takes 7 work days (3,500 miles of lost "production") and costs $4,650 each time the component must be replaced when it fails. The component will be "as good as new" after the maintenance action. The RCM analysis team is considering whether to incorporate a scheduled preventive replacement task into the maintenance plan. Because there are no additional logistical delays/costs for a planned replacement, the PM task will take only 1 work day and cost $2,050.

Using the RCM++ software, the team can first estimate the optimum preventive replacement time for the component and then simulate the operation of the equipment for 120,000 miles to estimate the cost and average availability that can be expected in a year from the two maintenance strategies that are under consideration. By entering the cost of corrective maintenance (CM), the cost of preventive maintenance (PM) and the probability of failure into the following equation, the optimum PM interval is determined to be 60,330.25 miles.

 

 

Rounding to 60,000 miles and performing the simulation yields the following results per vehicle per year:

Run-to-Failure

 

• 1.43 CMs and Uptime = 115,114.20
• Total Operating Cost = $6,626.25
• Cost per Uptime = 0.058 per mile
• Average Availability = 95.93%

Preventive Replacement

 

• .98 CMs, .79 PMs and Uptime = 116,246.80
• Total Operating Cost = $6,188.95
• Cost per Uptime = 0.053 per mile
• Average Availability = 96.87%

 

In RCM++, the analysis indicates that the scheduled replacement strategy provides both lower cost and better availability. Note that the differences between the two strategies will be even greater when applied to the entire fleet of vehicles over multiple years of operation.

 

Example 2 - Electrical Component with Infant Mortality

Another critical failure mode has been identified for an electrical component of the truck described in Example 1. This follows a Weibull distribution with beta = .76 and eta = 100,000 miles. The RCM analysis team is considering a planned replacement for this component at 60,000 miles to coincide with the PM for the mechanical component. For this failure mode, the CM downtime is 4 work days; the CM cost is $2,800; the PM cost would be $1,200 and there would be no additional PM downtime because the equipment is already down for the other maintenance task. The analysis yields the following results:

Run-to-Failure

 

• 1.21 CMs and Uptime = 117,617.74
• Total Operating Cost = $3,374.00
• Cost per Uptime = 0.029 per mile
• Average Availability = 98.02%

Preventive Replacement

 

• 1.39 CMs, .88 PMs and Uptime = 117,252.30
• Total Operating Cost = $4,934.80
• Cost per Uptime = 0.042 per mile
• Average Availability = 97.71%

In this case, the analysis indicates that a run-to-failure maintenance strategy will be more cost-effective and provide better availability. In fact, since the beta parameter of the failure distribution is less than 1, this indicates that the equipment does not experience wearout and there is no optimum preventive replacement time. The team could repeat the analysis for other maintenance intervals and would always determine that run-to-failure is more cost-effective.

 

Conclusion

As this article demonstrates, cost-based comparisons can be very useful to help RCM analysts to select the most appropriate maintenance strategy for a particular piece of equipment/failure mode. ReliaSoft RCM++ software automatically performs the maintenance task cost calculations described here. This functionality relies on the same powerful simulation engine available in ReliaSoft BlockSim software, which can also be used for maintenance planning and other more complex system reliability, maintainability and availability analyses. For more information, see /content/hbkworld/global/en/products/software/analysis-simulation/reliability/rcm-reliability-centered-maintenance-software.html or /content/hbkworld/global/en/products/software/analysis-simulation/reliability/blocksim-system-reliability-availability-maintainability-ram-analysis-software.html.

 

References

ATA MSG-3 "Operator/Manufacturer Scheduled Maintenance Development," updated in March 2003.

NAVAIR 00-25-403 "Guidelines for the Naval Aviation Reliability-Centered Maintenance Process," issued in February 2001.

Nowlan, F. Stanley and Howard F. Heap, Reliability-Centered Maintenance. Issued in December, 1978.

SAE JA1011 "Evaluation Criteria for Reliability-Centered Maintenance (RCM) Processes," issued in August 1999.

SAE JA1012 "A Guide to the Reliability-Centered Maintenance (RCM) Standard," issued in January 2002.