“People often maintain something because they think that is what they need to do, but they don’t understand the function and they don’t understand that sometimes when they maintain something, they are probably restarting a potential failure mode,” says Ring.

One problem is that too many people place too much stock in the original equipment manufacturer (OEM), and blindly input the manufacturer’s maintenance data into their CMMS. Granted, the OEM recommendations are necessary for the warranty period, but after the warranty expires the owner should carefully analyze the use and application of the equipment and develop a more efficient and cost effective maintenance program. Once the warranties expire, it is up to the owner to develop a new, more efficient maintenance plan.

The recommended maintenance actions in the warranty are meant to protect the manufacturer at least as much as the owner, says Ring, and they are not written in context. The manufacturer has not thought through the type of role that the equipment is going to play: Is it at a high reliability data center, or is it just providing air conditioning or a pump to an office facility? Is it being used in steady state or in a cyclic state?

“Too often we are over-maintaining the equipment,” he says. “You need to revalidate that maintenance information because when you take over a site, you are probably assuming a lot of legacy data.”

“Maintenance” can best be defined as maintaining the status quo or functionality of equipment, which is a critical function in important research facilities or data centers, but as much as one-third of the maintenance done is unnecessary, he says.

Maintenance programs are inefficient for a number of reasons. For example, only about one-third of the companies in North America conduct a root-cause analysis when they have a failure, says Ring, and this leads to short-term repairs due to repeated failures. Pumps have bearings. If a pump fails because the bearings fail, most maintenance people will simply replace the bearings. However, if they do a root cause analysis, they might discover that the shaft is misaligned, putting uneven pressure on the bearings. If left undetected, the misalignment will continue to damage the pump until the failure is more extensive than simply the bearings.

“All too often we just fix it and go on with our business,” says Ring. “You have to teach your people that’s not the way to think.”

Another helpful strategy is to implement a predictive maintenance program that relies on sensors to detect when a component is about to fail. The system could include, for example, a wireless sensor that monitors the temperature and pressure of lubricating oil, and sends a radio frequency signal when the numbers fall below normal set-points. That would obviate the need for a maintenance person to regularly check the oil.

It also is important for maintenance experts to be part of the engineering and construction design teams. This puts the maintenance professional in a position to point out if the design, materials, or layout of a building will make it more difficult or costly to clean and maintain. Maintenance is a long-term, life-cycle cost, Ring argues. It also is a profit center because proper maintenance is what keeps a plant up and operating and lowers the cost of manufacturing a product.

If It Isn’t Broken, Don’t Fix It

A plant can run at peak efficiency if 80 percent of its maintenance energy is focused on 20 percent of its equipment. This is the basis of Reliability Centered Maintenance (RCM). It is based on the realization that different kinds of components exhibit different failure modes; over-maintaining them can be tantamount to tinkering with perfection, and can make a piece of equipment less reliable than if it is left alone.

The RCM concept was born in the 1960s with the development of the Boeing 747. The Federal Aviation Administration (FAA) certifies a unique maintenance program for each new aircraft. When Boeing first developed a maintenance program for the 747, it was so expensive that no one would have bought the plane. Boeing and its buyer, United Airlines, worked with the FAA to determine the statistical probability of component failures. That work resulted in the 80/20 idea and RCM, which have since been adopted throughout the airline industry and in other industries, such as nuclear power plants.

Non-structural equipment and systems fall into six categories of reliability, with failure rates that can be placed on a series of “maintenance bathtub curves,” so named because the first curve looks like the cross section of a bathtub. The curves demonstrate that only 11 percent of these systems and products would benefit from a limit on their operating life. A full 89 percent could theoretically last forever—notwithstanding random failures—if they make it past a certain early benchmark.

The first curve represents the life of electronic/mechanical devices, such as a car engine. The electronic parts suffer from “infant mortality”: If they are going to fail, they will fail early as a result of faulty manufacturing. If they work well from the beginning, they are likely to last indefinitely. Mechanical parts, however, eventually wear out, particularly if they are poorly maintained, which shows up as the flat failure curve turns upward again as a function of time.

The second curve shows a constant failure rate with an upturn as the equipment ages. This represents the life of things like a reciprocating aircraft engine.

The third line shows how a turbine jet engine would have an increasing failure rate over time, but no identifiable wear-out age.

The next curve shows a low probability of failure when the item is new or fresh out of the shop, followed by a quick increase to a constant level. This happens with a simple mechanical pump or valve after repacking, or a hydraulic unit or system.

The fifth line represents a constant failure mode, in which failures are random and unpredictable, for example with single-function mechanical devices where there is little heat, resistance, or metal fatigue.

Finally, there are electronic devices with solid state circuitry, which suffer only from infant mortality.

If you perform unnecessary maintenance on the last three categories, representing 89 percent of equipment and components, you are “resetting the clock” on its failure probability and placing it at greater risk of failure each time you tinker with it.

The RCM has four features: preserve function; identify failure modes; prioritize function need (via failure modes); and select applicable and effective preventive maintenance tasks for the high failure modes.

It can be costly to implement this system. For a simple plant with about 30 systems, for example a fossil-fuel power plant, implementation could cost $200,000 to $300,000 or more over one to two years. The cost jumps to more than $1.2 million over three to five years in a more complex plant with 100 systems, such as a nuclear power plant.

“You do not want to apply RCM to every system in a plant,” says Ring. “Using the 80/20 rule, apply classical RCM to 20 percent of the plant systems.”

A more simplified and less costly approach, which Ring calls “RCM Lite,” can also be applied. Under this system, a company will categorize its equipment by mission criticality and analyze it for reliability, risk, and failure only in the mission critical systems and components. The rest can be judged using the maintenance bathtub curves and manufacturer’s data.

Preventive Maintenance and Management

The foundation of maintenance is preventive, says Ring. If that is properly in place, other, more proactive, functions become possible: stores and procurement, work flow system, RCM, and, ultimately, continuous improvement.

“If you don’t have an effective maintenance program, all of this other stuff has a very unstable platform,” he says.

The person in charge of that program should be a maintenance planner, a position that simply doesn’t exist at many companies where maintenance is considered a cost center and not a profit center.

“A maintenance planner isn’t an administrative person or a secretary,” he explains. “It is your most senior technician, a person who understands the work and the process, a true maintenance professional.”

Ring recommends that companies nurture their best maintenance technicians and train them to be planners.

A properly utilized maintenance planner can increase the wrench time efficiency of the technicians, who are making from $16 to $35 an hour.

“Wrench time efficiency at most companies we analyze is between 15 and 35 percent,” says Ring. “World class wrench time efficiency is between 56 and 65 percent. If you achieve that, you are not wasting anyone’s time: As soon as the technicians show up, they have their tools and required materials, they are in their little golf cart, and are off doing work.”

A maintenance planner also can keep track of results by comparing results to work orders and determining if the job took longer than it should have.

“Use of maintenance planners is one of the largest potential areas for cost savings,” says Ring. “Planned vs. unplanned work may have a cost ratio as high as 1:5.”

By Lisa Wesel

We welcome your Questions and Comments

Copyright 2008 Tradeline Inc.
All Rights Reserved
ISSN: 1096-4894