Most companies have figured out by now that reactive maintenance just doesn’t cut it for mission-critical assets. Reactive maintenance, or a mentality of “fix it when it breaks”, is a sure-fire way to ensure poor equipment reliability, while increasing maintenance and downtime costs.
As such, your company, like most, may have invested in a condition-based maintenance program. By definition, condition-based maintenance is the act of performing a maintenance task based on some measurable parameter, whether due to a simple visual inspection, or a more sophisticated tool such as oil analysis or vibration analysis.
But are you receiving the optimal benefits from your CBM program?
Figure 1. Two Main Categories of ConditionBased Maintenance
Figure 1 outlines the two categories of condition-based maintenance: predictive and proactive maintenance. Predictive maintenance uses tools such as wear debris analysis, vibration analysis, thermography and motor current analysis to diagnose fault conditions, to allow us to (hopefully) correct the problem when it is the least severe, such that the problem has yet to cause collateral damage, and when the repair can be effected at the least inconvenient time to the organization (such as during a scheduled production shutdown or when the hardware, tools and resources are available).
However, what many exponents of condition-based maintenance fail to realize is that predictive condition-based maintenance does not alter the failure rate, it simply changes the balance of which problems become severe versus those that can be corrected while the problem is still a moderate or minor issue (Figure 2).
Put simply, the plant could deploy a team of world-class vibration analysts (or oil analysts) to take predictive data hourly or even continuously, and it would still not change the fact that the bearing, gearset, etc. is going to fail.
Predictive maintenance does not alter the total number of failure events, but rather changes the balance of problems that become severe, verus those that can be addressed when the problem is either moderate or minor in consequence (Figure 2).
In my experience, this approach to maintenance has by default become popular in many industries, particularly those where production occurs 24-7, when just a few minutes of downtime represents huge dollar losses. Industries such as pulp and paper, oil and gas production and refining have invested millions of dollars in predictive maintenance in an attempt to eliminate unscheduled downtime.
So what is driving this predictive approach to maintenance at the board-room level? The answer of course is dollars. Managers are accountable to bottom-line performance.
They look at key performance indicators such as availability and overall equipment effectiveness (OEE - defined as the product of availability, quality and yield) to gauge how successful the organization has been in eliminating unscheduled downtime; but at what cost? Let’s look at one other commonly used metric, return on net assets (RONA). RONA is defined as:
By reducing unscheduled downtime through predictive maintenance, the objective is to increase revenue through increased equipment availability. But again, at what cost? I remember one extreme example of unscheduled downtime avoidance gone wild in a plant, which called for the removal and rebuild of a mission-critical gearbox during the annual (scheduled) maintenance shutdown, at a cost of $100,000, because at $30,000 per plant, this was considered by management to be cheap insurance.
Anyone who has attended any form of Noria training will know that we are strong proponents of a proactive approach to condition-based maintenance (Figure 1). Rather than waiting for equipment to fail, hoping that the predictive program finds the problem in time, we promote proactive maintenance - which targets the fundamental root causes of premature equipment failure; therefore, we can address and eliminate the root cause of the problem before failure occurs.
While there are too many lubrication (and other) root causes to explore here, let’s focus on one of the most common, particle contamination. Many organizations have recognized that depending on operating environment, particle contamination contributes up to 80 percent of all lubrication-related wear problems.
Because of this, many organizations have adopted an aggressive program to monitor particle contamination, using either laboratory or onsite particle counting. These organizations have truly evolved to a proactive condition-based maintenance organization, right? But wait … let’s take a closer look.
Proactive maintenance is not about measuring causative factors like particle contamination. It’s about recognizing why equipment is prone to failure, and putting a structured program in place to eliminate, or at least control the causative agent to acceptable levels.
To illustrate the point, consider this example: Company X has learned that in its industry, particle contamination is a leading cause of premature failure. Armed with this knowledge, a condition-based maintenance program that includes regular ISO particle counting is implemented.
To the dismay of company X, every time the sample reports come back for critical gear drives, the lab indicates that fluid cleanliness levels are at ISO 25/25/25 (usually indicative of sensor overload due to extreme levels of contamination) rather than the recommended ISO 18/16/13. Based on this, a work order is generated to change the oil. A new oil barrel is brought to the shop from the outside storage area. The oil is dispensed using a hand-cranked barrel pump (no filter) into a homemade oil can.
The oil can is then carried across the plant during a (scheduled) outage (remember – our mandate is to make sure all work is completed during scheduled downtime), ready to change oil.
When the machine becomes available, the lube tech commences to change the oil. After draining the old contaminated oil, the tech refills the gearbox. First, he removes the combination fill cap/breather (not too difficult because the simple bayonet fitting is loose because the OEM seal has long-since disintegrated).
After adding the correct amount of makeup oil, the diligent lube tech carefully removes the dipstick, wipes the level mark with a rag, then reinserts the dipstick to check the level. Satisfied, he continues with his next task.
But what a surprise when the next sample is taken: the particle count is still ISO 25/25/25. How could this be, did the lube tech forget to change the oil? The answer is obvious: The bulk of this contamination comes from poor contamination control practices related to storage and handling, dispensing practices coupled with a failure to adequately equip the gearbox with the correct hardware (external level gauge instead of the dipstick, proper breather and quick-connect to add oil, quick-connect on the drain to allow for periodic decontamination with a filter cart etc.).
Proactive maintenance is about changing old practices – the way things have always been done - to achieve a new “business as usual”. Unless this is achieved, you will slowly go insane as you expect to get a different result (improved fluid cleanliness) but continue with the old “business as usual” (poor contamination control practices). Proactive condition-based maintenance is about a lifestyle change - creating a new maintenance strategy.
It is not about particle counting, measuring viscosity or acid number, or looking for bearing defect frequencies – these are merely tools that tell us if our strategy has the desired effect.