An often debated subject on maintenance rebuilds is exactly which parts or components should be replaced. There are two extremes when making this decision. The low-risk option is to replace everything with new each time, however, the extreme high cost is not justified in most business models. Some sort of condition-based maintenance inspection would be more ideal for replacing components or signaling major rebuilds. Infant mortality of installing new parts may even make that option high risk. The high-risk option for maintenance rebuilds is reuse and run till failure strategy. There is a balance in reliability and cost goals, as shown in Figure 1. There is also a balance in replace or reuse to achieve the proper reliability and cost balance. This article will give some considerations around decisions to reuse or replace when executing maintenance.
Figure 1: Reliability cost curves
As a typical example, consider this real-life lawn mower situation. A lawn mower is basically run to failure, with the exception of some basic preventive maintenance (PM) measures, such as lubrication (e.g., grease and oil change), sharpening blades, etc. But in this situation, the drive sheave assembly bearing failed. It is a one piece sealed bearing (i.e., lubed for life), as shown in Figure 2. Since it is necessary to disconnect the deck and roll it out from the mower to change this assembly, should other components being changed as well? The entire drive sheave assembly costs around $300 (which is the original dealer’s selling price), while bearings cost only $24. Since the mower was partially torn down, changing the blade bearings was considered. They are only sold as assemblies and cost $150 each, so $450 total. Of course, this investment needs consideration. The mower is 25 years old and is serviced on a regular basis by greasing the bearings. Although it still runs and works great, spending $450 to replace blade bearing assemblies and $300 to replace the drive bearing sheave assembly could not be justified due to its age. Since the sheave itself was in good shape, new bearings were purchased for $24. The belts were also inspected and they were in good shape. So, with free labor, the mower was fixed for $24 versus $750 for a partial overhaul. If all new belts, blades, springs, etc., were added, well over $1,200 would have been put into a 25-year-old mower.
Figure 2: Lawnmower deck assembly
The decisions about the scope of what to replace should be vetted since there are many factors to consider and rarely does a one-size-fits-all approach apply. A reliability organization adds value in these decisions by taking known information about the equipment so an educated decision with calculated risk can match the business model. This information comes from various sources, such as vibration analysis, oil analysis, infrared, preventive maintenance, failure history, past root cause failure analyses (RCFAs), etc. Criticality of equipment is also an important factor in risk evaluation.
Evaluating risk is a key area for determining reuse or replace. Consider some statistical learnings from the reliability bathtub curve shown in Figure 3. In the infant mortality zone, which is the time after installation and start-up, the risk is high, so a maintenance focus is necessary to manage risk. Normal preventive maintenance will not be effective here or in the wear-out zone. Some sort of vigorous checkout and commissioning or load testing is best to prevent failure in this zone. Also, special attention to installation detail for maintenance is critical in the infant mortality zone.
Figure 3: Bathtub curve
The normal maintenance zone is the middle portion of the bathtub curve. The failure rate is basically random, as this is the bulk of the operating time of the life of the equipment. Choosing to replace a part too soon increases the failure rate on the equipment as the infant mortality drives up the failure rate. The optimum time to change the part is once the failure rate exceeds that of infant mortality risk.
Remaining Useful Life
For components that can have a remaining useful life (RUL), the P-F curve may give some insight into decision-making. In Figure 4, two scenarios are shown. After a PM, a component is shown to have five months RUL. If the P-F interval is six months for the typical failure mode, then the PM is executed at around three months or one-half the P-F interval. So, with five months RUL, there would be another PM before replacement was necessary. For a safety factor, one might use 75 percent of RUL as a rule of thumb, which is 3.75 months. This is still greater than the PM frequency of three months. In this case, do not replace the component since RUL>PM frequency.
Figure 4: P-F curve reuse or replace example for a brake pad
In the other scenario, the RUL is only two months, so 75 percent of RUL is less than the PM frequency and the component needs to be replaced. Note that this is just a hypothetical situation to illustrate the thought process. It may very well be wise and feasible to go ahead and change the component with other circumstances factored into the equation. Let’s take a look at a few practical examples of replace and reuse.
A common job on industrial equipment is a bearing change. Whenever a bearing has a defect, always replace the defective bearing unless one isn’t available, then some creative alternatives may need to be considered until the bearing can be replaced. If the bearing is an adapter sleeve mount, it is good to have one on hand in case needed, but many times this can be reused. A bearing nut also can be reused many times, but have one on hand. Generally speaking, it is good to have a lock washer because reusing one can lead to failure as the tabs can fail from fatigue of repeated use or easily become damaged upon reinstallation.
Infant Mortality Tip: It is best to mount the bearing without the lock washer to avoid damage from torsional stress. A lock washer should be installed after bearing clearance is set.
Replace any contact seals in bearing housings. Contact seals usually last between six to 12 months depending on application conditions. Labyrinth seals typically should be reused. Bearing housings also are typically reused once inspected for wear and fit. If a housing is changed, do not mix two different split housings. For larger, more involved rebuilds, small bearings (i.e., less than 180mm bore) are generally replaced.
Another example is changing an end suction centrifugal pump rotating assembly. Most items on a centrifugal pump depend on the wear rate of key components, such as the impeller, suction wear plate, impeller nut, and casing. Replacement of these components is determined by a combined result of visual inspection and pump hydraulic performance. Impeller clearances can be reset to restore hydraulic performance for a time, but other negative side effects will creep in. Wear and impeller adjustments eventually lead to increased thrust loads on the pump and bearing life will diminish. These decisions may require more information about the application.
If the pump uses packing, then it should be replaced. The lantern ring can be reused unless damaged. Many times, the packing shaft sleeve will need to be replaced, but with a mechanical seal, the sleeve should be able to be reused.
Infant Mortality Tip: For many pumps, the shaft sleeve is not the same for seals and packing. All O-rings on the impeller or nut should be replaced. These are critical sealing areas on the liquid end.
One last example is belt drive equipment. There are not a lot of components, so there is not as much to consider. Belt life is typically three to five years depending on application conditions, such as loading, temperature, start/stops, shock loading, etc. Drive belts are likely replaced during each equipment rebuild as they are the low life wear element of the drive. The smaller, higher speed sheave would be another high wear component. The low arc contact makes it a high wear component, like the one shown in Figure 5 on an agitator drive. Likewise, the larger sheave can be typically reused. Bushings, keys and other small components can be typically reused.
. 
Figure 5: Belt drive sheave wear
In all these examples, PM inspections and condition evaluation can give the necessary information to determine RUL. The decision to reuse or replace seems like a small one, but can have a big impact on the plant, both financially and in terms of reliability. This decision is an area where maintenance reliability professionals can add real value to an organization. The evaluation of these details by maintenance reliability professionals can elevate maintenance performance.
For success, it starts with good inspections by the technicians and then follow-up by management. Wherever your organization is in its reliability journey, there is tremendous potential in the decisions on whether to reuse or rebuild.
. 
