Analysis and Modeling of Grouped and Opportunistic Preventive Maintenance
[Editor's Note: This article has been updated since its original publication to reflect a more recent version of the software interface.]
When planning preventive maintenance strategies, considerations of the overall benefit for the whole system should supersede the optimum plan for each component separately. Maintenance plans should be addressed in a holistic way to optimize the cost of ownership of the system. For example, sometimes it is less expensive and more convenient to perform preventive maintenance on a component in a system when performing a repair action on another component in the system rather than at the optimum time for performing the preventive maintenance for that component. A repairable system reliability and availability modeling and simulation tool, such as BlockSim, can help model your system and assess different maintenance strategies to determine what will be best for the overall system. This article discusses how to model grouped and opportunistic maintenance strategies in BlockSim.
What is Meant by Grouped and Opportunistic Maintenance?
Suppose that you take your car to the mechanic shop to replace a timing belt. You then get a call from the mechanic asking you whether you would like to have the water pump replaced (preventively) because it is a good opportunity to do it now rather than later. The water pump is located close to the timing belt. It might be cheaper and more convenient to change the water pump now, since most of the required labor is complete, rather than waiting for the recommended preventive replacement time for the water pump, taking your car into the shop again and having parts removed from the engine to be able to access the water pump. You consider all costs and inconveniences related to the labor and downtime caused by having to take your car to the mechanic again, in addition to the increased risk of failure. You also consider utilizing the additional life you can get from the existing water pump. You might find that it is in your benefit to accept the mechanic's recommendation. Was this just a mechanics' gimmick? Or is this a viable preventive maintenance strategy? It may be more of the latter.
The purpose of grouped and opportunistic preventive maintenance is to take advantage of the resources, efforts and time already dedicated to the maintenance of another part in the system in order to cut cost. It is typically possible and more beneficial when these parts are located close to each other in the system and no significant additional work is needed to perform the second maintenance activity. It is also more commonly applied in cases where significant travel cost is incurred or major disassembly of the system is required or, when the parts are not too expensive, they may be disposed of even before they are due for maintenance. It can also be applied in situations in which waiting beyond the recommended PM time to actually perform the maintenance (i.e. waiting for the next opportunity) does not present high risks to the system. This concept may be expanded by performing the preventive maintenance every time the system is down, not just when certain parts are being repaired.
Such maintenance strategies are performed in many areas and industries, including manufacturing, military, automotive, infrastructure, capital equipment, etc. The approach has many benefits in addition to being convenient and making fewer interruptions to the system. It could also reduce the chances of human error and other problems that arise from frequent intrusions and disruptions to the system.
Let us consider the following repairable system's reliability block diagram (RBD).
The owner of the system performs repairs and preventive maintenance actions to extend the system's life. The company sends a repair crew to perform maintenance (corrective and preventive) and is concerned about its significant traveling costs ($400 per incident).
The next table displays the failure and repair properties of the system's components.
|Parts||Failure Distribution||Replacement Labor Duration||Part Cost|
|A||Exponential (Mean = 900 hours)||2 hours||$100|
|B(*)||Weibull (β = 2, η = 500 hours)||1 hour||$20|
|C||Exponential (Mean = 500 hours)||2 hours||$240|
Parts A and C follow an exponential distribution, which makes preventive maintenance for these parts useless (more on this can be found here.)
(*): The duration given is for the replacement of all four B parts.
The optimum replacement time for each of the B components can be calculated using the corrective (unplanned) and preventive (planned) replacement costs and the failure distribution. (This analysis is not the focus of this article, more on this computation can be found here.) The unplanned cost would include part cost and travel cost, whereas the planned cost would include the part cost (since preventive maintenance will be performed when the crew is already at the system, as we will see later in this example). The optimum replacement time for each of the B components is calculated to be every 112 hours.
The optimum replacement time of 112 hours is calculated with the goal of optimizing the maintenance cost of the B parts, independent of the rest of the system. The grouped and opportunistic maintenance concept, on the other hand, tries to optimize the cost of maintaining the whole system. To illustrate the application of this concept, we compare three different maintenance strategies:
Strategy 1: Perform preventive replacements for the B parts upon a fixed time interval calculated based on the component's optimum replacement time (i.e., every time a B part accumulates 112 hours of operation, it gets preventively replaced).
Strategy 2: Replace all the B parts whenever any of them fails or whenever part A fails.
Strategy 3: Perform preventive replacements for all of the B parts whenever the system experiences a downing failure (i.e., when all of the B parts fail, when A fails or when C fails).
Note: In this example, for the sake of simplicity, we limited the maintenance strategy options to these three. Readers are welcome to try other variations or combinations of the plans listed above.
For this example, we are interested in the system's operation for 10,000 hours.
Maintenance Strategy 1:
We set up this maintenance strategy in BlockSim by creating a Preventive Maintenance Policy and using the fixed preventive replacement time of 112 hours for each item, as shown next.
This policy is then linked to the B part by selecting this as the Preventive Maintenance Policy in the Preventive tab of the block's property window as shown next (a similar step is to be repeated for all the B parts and in the assessment of Strategies 2 and 3).
We assume that each time the B parts needs to be preventively replaced, only one trip is required by the crew. The B parts are replaced one part at a time to avoid bringing the system down.
Maintenance Strategy 2:
We set up this maintenance strategy by creating a Preventive Maintenance Policy and using the option of performing maintenance Upon Maintenance of another Group Item (part A), as shown next.
The B parts are grouped with part A using the Item Group # property, by giving the A and all B parts the same number (using a number other than 0, since 0 means no groups).
In this scenario, the crew proceeds to replacing the B parts (one part at a time) right after completing the maintenance of A. No second trip is required.
Maintenance Strategy 3:
We set up this maintenance strategy by creating a Preventive Maintenance Policy and using the option of performing maintenance Upon System Down, as shown next.
In this scenario, the crew proceeds to replacing the B parts (one part at a time) right after completing the maintenance of the part that brought the system down. A second trip is not required.
We simulate the system for 10,000 hours. In this example, we will use maintenance cost (parts and travel cost) in addition to downtime cost ($500/hour) as our metric to be used for assessing the maintenance strategies.
|Total Parts Failures||Total Number of Corrective Replacement Trips||Total Number of Preventive Replacements||Total System Corrective Replacement Downtime||Total Corrective / Preventive Replacement Trips||Total Cost|
|Strategy 1||$13,009.33||48.28||338.54||62.33 hr||386.82||$198,900.66|
|Strategy 2||$10,012.47||71.72||41.67||61.70 hr||71.72||$69,546.75|
|Strategy 3||$9,645.53||87.63||31.53||63.12 hr||87.63||$76,255.71|
The values in the table above were taken from the System Overview and Block Summary reports obtained as part of BlockSim's simulation results.
The total cost is calculated by multiplying the number of "trips" by the travel cost and adding the parts cost and downtime cost. In the case of Strategy 1, the number of trips is the number of corrective replacements added to the number of preventive replacement trips (because in this strategy, the preventive maintenance gets a dedicated trip). Also in the case of Strategy 1, the number of preventive maintenance trips is the sum of the number of preventive replacements performed for each of the B parts (because in this strategy, each B part gets a preventive replacement scheduled independently from the others). In the case of Strategies 2 and 3, the number of trips is the number of corrective replacements only (because in these strategies, the preventive maintenance is performed along with the corrective maintenance). Strategy 2 emerges as the most economical strategy for the system overall.
The system's incurred downtime is that of corrective maintenance only since the preventive replacement of the B parts can be performed in a sequential fashion and the system won't be downed.
Note: This article does not intend to imply that the grouped and opportunistic maintenance is always a preferred maintenance strategy. In other scenarios, different types of strategies might be more appropriate. Our goal in this article was to illustrate how such grouped and opportunistic maintenance options can be modeled and analyzed using BlockSim.
In other cases, the grouped and opportunistic preventive replacement or maintenance might be improved with inspection activities to achieve even more flexibility in maintenance plans and cost savings. For example, in the scenario of the timing belt and water pump, the mechanic could first check if the water pump is getting close to failure before recommending the preventive maintenance. To model such a situation, we need a way to set up an additional condition (i.e., checking whether a failure is imminent in the water pump). This can be performed in BlockSim by performing on-condition maintenance (on-condition maintenance modeling and analysis was covered in a Hot Topics article in the June 2007 issue of Reliability HotWire).