Sampling every lubricated asset within a process is impractical, and some of the data is likely to have very little value. However, choosing which assets should be sampled for an in-service oil analysis program can be a daunting task. With a straightforward approach to auditing, one can arrive at a reasonable shortlist. The audit should address four factors: criticality, environment, safety and cost.
Any asset that is critical to overall operation must be sampled. Critical to operation can be defined as any asset whose slow-down or stoppage causes a ripple-effect throughout the rest of operations. Even assets that are duplicated, but whose output is required continuously must be sampled.
Any asset that operates in a significantly contaminated environment, whether moisture, particulate or both, is a strong contender for the shortlist. Even systems equipped with filters and breathers are not exempt, as these protection devices may fail unexpectedly.
Any asset whose failure to operate may cause injury to personnel must be sampled, e.g. a backstop on a conveyor belt or fluid power speed controls. If the failure of the asset itself is likely to cause injury, and can be mitigated by good fluid health, then it should be sampled, e.g. high-pressure hydraulic system.
Any asset whose failure or loss of usefulness carries significant cost, whether in downtime, parts or labor, should be sampled. Occasionally, simply the ease of repair due to remoteness and accessibility is strong enough to warrant sampling. Assets with low-cost, yet chronic, issues can be considered as well.
Assets that can be exempted from sampling include any asset that operates with predictable failure modes for acceptable durations of service with minimal maintenance, i.e. assets operating with under a reactive maintenance philosophy like oil lubricated electric motors or centrifugal pumps that fail due to volute wear.
The need for sampling should be examined on any small, reliable asset where pulling an oil sample requires topping up afterwards or virtually constitutes a complete oil drain. A similar concession can be made to reliable assets with regular oil drains which are difficult to sample, e.g. cooling tower gearboxes that require shutdown before access can be granted.
The value of oil analysis does not necessarily come from the outright prevention of a failure, but rather from the reduction in severity of the failure. After all, the fluid will always degrade over time and so it must be monitored to ensure it occurs predictably and that the lubricant is changed or conditioned accordingly. Contamination and wear will always occur, yet action is only taken when a trend is abnormal or a threshold is crossed. Therefore samples must be taken from assets at a frequency where this is possible.
Assets that experience repeated failures between samples should be sampled more frequently, and inversely assets that achieve full drain or rebuild intervals without issue can be sampled less frequently, even to the point of not being sampled. The purpose of any tool is to employ it when and where it is effective.
Establishing a proper shortlist ensures oil analysis is employed effectively and efficiently to help improve asset reliability.