Competitive Maintenance Strategies

Many consultants are prospering today by creating "new" maintenance strategies. What they’re really creating is new buzzwords. At a recent conference, for example, speakers debated the pros and cons of reliability centered maintenance, profit-centered maintenance, equipment-based maintenance, total productive maintenance, and proactive maintenance. But though the titles were slightly different, the overall concepts were amazingly similar: Competitive maintenance, the experts all seemed to agree, uses an analytical process and condition-monitoring tools to select the optimum maintenance strategy for each plant system. Note that the optimum strategy for the plant’s single steam turbine will be different than that for its four chemical-injection pumps. When the dust settles from all the discussions, there are only three ways to maintain a piece of equipment:

  • Corrective maintenance. This is often called "crisis management" or "run-to-failure," because maintenance or equipment repairs are scheduled only when there is a noticeable deterioration in machine condition — like the shaft just separated from the coupling and went flying through the air. Corrective maintenance is marked by a high percentage of unplanned maintenance activities, high replacement-part inventories, and inefficient utilization of maintenance personnel, but it is still the right choice for certain pieces of equipment.

  • Preventive maintenance. Often called "time-based maintenance," preventive maintenance (PM) represents a shift away from unplanned to planned maintenance activities. Periodic inspection and maintenance are scheduled at prescribed intervals in an attempt to reduce or eliminate equipment failures. Depending on the intervals set, this can represent a significant increase in inspections and routine maintenance; however, it should also reduce the frequency and seriousness of unplanned machine failures.

  • Predictive maintenance. Also called "condition-based" maintenance, predictive maintenance (PdM) replaces the arbitrary time intervals with maintenance scheduled only when the condition of the equipment warrants it. The increase in cost of equipment monitoring should be more than offset by a reduction in unnecessary PM inspections and unplanned machine failures. But it takes time and money to properly perform PdM maintenance, and not every piece of equipment is worth that investment.

Unlike these three methods of maintenance, reliability-centered maintenance (RCM) is a process — a process by which the appropriate method, or combination of methods, is selected. RCM is a highly organized, statistical approach to equipment maintenance. Maintenance activities are determined for a specific machine or machine type based on the goal of having it operate in the most reliable way possible at the lowest cost. For example, a chemical-injection pump might be run to failure because it is not critical to production and cost-effective simply to replace; a chain drive may be placed on a PM program with daily lubrication; and a steam turbine will be monitored with sophisticated diagnostic tools.

Another Link in the RCM Chain

Failure mode and effects analysis (FMEA) is a key step in implementing RCM. It sounds complicated — which is what the consultants want you to think — but it’s simply a tool to quantify what you know and assemble it in a logical manner. Perhaps without realizing it, you’ve done many FMEAs whenever you had a plant incident. You analyzed why something failed, looked for the root causes, and made the changes necessary to prevent it from happening again — or at least minimize the chances. An FMEA walks through the same exercise, using hypothetical failures of plant functions, which then allows you to make smarter maintenance plans.

The table included here can be used to perform FMEA. As you follow across the table, from system to failure modes and causes to failure effects, you understand the relative importance of each plant system or piece of equipment. This understanding helps you decide how much preventive and predictive maintenance each system warrants. When developing the FMEA, don’t get too imaginative with possible failure modes; consider only those that have a high probability of occurrence.

Suggested tabulation used in a failure modes and effects analysis.
Source: POWER

Failure description

Failure effects


Failure mode

Failure cause





Strategic exercises like FMEA can produce dramatic improvements in your maintenance plan, but don’t stop there. The execution of the plan is just as critical.

Consider this story: The new maintenance chief at a plant in Virginia, in order to understand the work management process he was now responsible for, drew up a block diagram showing the functional process of taking a maintenance task from beginning to end. He was startled when the diagram ended up being 17 feet long. His effort was not in vain, because he promptly used the diagram to convince his staff that the existing system was way too cumbersome and that urgent changes were needed.

In response, the plant implemented what it calls a "10-4-2" process. Ten days before a maintenance task is scheduled to be performed, a work order is delivered from planning to the maintenance workers. That allows the workers time to review the job and make some preparations. Four days prior, the maintenance workers draw all required materials from the warehouse and stage them in their shop. Two days before the work begins, the work order is delivered to the operations crew to make sure the work is compatible with plant status and to issue lockout/tagout packages.

Regardless of the size of your plant, the 10-4-2 process — or something like it — can improve your maintenance productivity. Or if a formal process like this isn’t needed, look at ways to reduce setup and traveling time. Say your plant has an air-cooled condenser and frequent work is needed on the upper deck. Why bother dragging tools and supplies up and down the stairwells, when you can inexpensively set up a toolbox on the deck? With the input of the crew, consider all the places where time can be saved, such as:

  • Fully equipped golf carts or work trucks.

  • Electric hoists mounted on high work platforms.

  • Maximum use of hydraulic, pneumatic, electric, and cordless tools.

  • Portable tool boxes packed and ready for recurring maintenance tasks.

  • Remote computer terminals that can access the computer-based maintenance management system.

Kick-Start RCM with a Pilot Program

RCM began with the airlines in the late 1960s, when wide-body jets were introduced to commercial service. Then it spread to countless other industries. Now, virtually all power producers are walking the enlightened path toward RCM.

A key benefit of the process is that it identifies and eliminates unnecessary maintenance actions, which may have been called out in overly conservative OEM schedules. Not only is excess maintenance costly, but it can actually harm operations if employees are literally tripping over each other or tinkering with machinery until it breaks. But in starting the movement toward RCM, airline companies had an advantage: They launched their programs on a new generation of equipment, a "blank sheet of paper," if you will. Most power producers, on the other hand, have had to fit the strategy into existing plants and long-established maintenance practices. As a result, plant managers striving to implement RCM often are frustrated by high start-up costs and organizational resistance.

One way around these hurdles is to implement RCM one plant system at a time; as management gurus say, eat the elephant one bite at a time. That’s what one Rocky Mountain utility did at a 540-MW, coal-fired station. Rather than force a disruptive, plantwide conversion to RCM, the station focused only on its coal-handling system. The station then implemented RCM in three phases:

  • Analyzing failures and developing appropriate maintenance tasks.

  • Packaging the maintenance tasks into concise work orders.

  • Measuring the results.

Engineers at the station prepared Pareto charts of coal-handling system failures — bar graphs showing the frequency of failures by category, listed in descending order of occurrence — and discussed the results with plant veterans. Some of the results validated workers’ impressions, while others provided some unexpected insights. For example, workers felt that mechanical problems were more critical than instrumentation and control, but the analysis revealed that both categories experienced almost equivalent failure frequencies and similar consequences of those failures.

Based on the failure analysis, the plant identified key maintenance tasks. Only 15% of the key tasks, mostly lubrication-related, were suitable for time-based maintenance, while the majority benefited from condition-based monitoring. Most maintainable components — motor, belt, pulley, crusher, and so on — had fewer than five key tasks.

Next, the tasks were grouped together into concise work orders that could be generated by the computer-based maintenance management system. To optimize technicians’ time, the work orders were organized around each major belt assembly and were designed to take between two and eight hours to complete.

When all was ready, new work orders for the coal-handling system were issued and performed, and the results were measured over a one-year period. The results include a 20% reduction in routine maintenance costs, more than a 90% reduction in emergency maintenance costs, and a dramatic reduction in total maintenance hours. In addition, system availability improved because several long-standing O&M issues were resolved. A chronic problem with rotary plows was fixed, for example, and numerous instrumentation deficiencies were corrected. Another important, but intangible enhancement, according to station managers, was that crew morale improved. "We’re getting support now," many of the mechanics and operators stated. The station’s success implementing the pilot RCM program is that much more impressive when you consider that it occurred while staffing levels were being reduced and a new dust-suppression system was being installed.

Some Specific Tips

Let’s assume you believe in the need for competitive maintenance strategies, you’ve read all the glowing case studies, and you understand the philosophy. Now you want some specifics that apply to your plant. That’s what we’ve compiled below: equipment-specific ideas that savvy power producers are applying to cut their O&M costs while achieving high reliabilities.

Extend Steam Turbine Overhaul Intervals. Major overhauls of steam turbines have an original equipment manufacturer (OEM)-prescribed interval of five years or so. But competitive power plants are extending the time between overhauls to as much as 10 years. Over a 20-year span, doubling the interval eliminates two overhauls and translates into, for a 250-MW combined-cycle plant, an approximately $4 million difference in reduced O&M cost and increased power sales.

Of course, the decision to extend maintenance intervals incurs some risk — not to mention the wrath of the OEM’s technical representatives and insurance carriers, who might bombard you with warnings of impending failure. To minimize the risk, decisions about maintenance intervals should be based on experience and methodical research. Intervals for steam turbine overhauls, for instance, didn’t suddenly leap from five to 10 years but were gradually extended as more and more overhauls showed turbines in good shape at the five-year point.

More important, the decision to extend maintenance intervals should be based, to the extent possible, on actual machinery condition using tried and true PdM technologies. EPRI’s Stress and Fracture Evaluation of Rotors (SAFER) computer program, for example, can help assess remaining life of steam turbine/generator rotors. SAFER reaches its conclusions by analyzing boresonic inspection results, calculating rotor temperature and stress distributions, estimating crack growth using actual flaw data, and computing the cycles required for cracks to reach critical size. EPRI estimates that the SAFER program has been used in the remaining life analysis of hundreds of turbine and generator rotors since its introduction in the 1980s.

Many steam-turbine operators are installing borescope ports to help make overhaul-extension decisions. Borescope ports, which can be installed on virtually any steam turbine regardless of age, allow visual and nondestructive testing of wear components without the need to totally disassemble the casings.

Apply New Repair Techniques to Electric Generators. A large number of 1970s-vintage water-cooled generators are now beginning to experience stator-winding problems. In response, users and manufacturers are striving to develop new techniques to detect the problems sooner and to repair them permanently. One of the problems is water leakage where the winding’s copper strands come together in a waterbox/clip arrangement. The brazed joints applied by the OEM occasionally fail, and the subsequent water leak damages insulation.

Over the course of several years, the leak can lead to complete failure of the winding. Conventional repairs such as TIG brazing may stop an existing leak, but they do nothing to ward off future leaks. The only permanent solution, traditionally, has been a complete replacement of the stator bars — an expensive and time-consuming task.

Several companies have introduced what many regard as permanent and cost-effective solutions to the problem. One company’s offering features a two-piece clip that eliminates the cast components and improves the brazed joints in both strand-to-clip and strand-to-strand areas. The supplier reports that the patent-pending two-piece clip can be installed in-situ, with the stator bars and field in place. It uses existing brazing methods and requires no new alloys or exotic machining practices. Other vendors have recently introduced new stator-leak repair techniques as well. For example, the OEM offers an epoxy-based repair, while a competing generator manufacturer offers a three-piece clip, similar in concept to the two-piece clip.

Another new maintenance tool for electric generators regards the stator wedge. A stator wedge is a small piece of electrical insulating material that fits in the stator slot of a generator to secure the copper coils in place. There are roughly 25 to 30 wedges in each slot, and 48 slots in the typical utility-scale generator. Loose or improperly installed wedges have been identified as a source of electrical failures in turbo generators. As hard, thermosetting resin/mica insulation replaced relatively soft asphalt/mica insulation in generator stator coils, this problem became more critical. A wedge-tightness detector is available that detects looseness, reportedly as accurately as manual hand-tap testing, while producing a permanent inspection record. And, its developers claim, it does it faster.

The instrument consists of a probe that is manually pushed along each slot. The probe is held magnetically to the iron core to avoid operator fatigue. An impact hammer strikes the wedge approximately 10 times/second, and the sensing head uses an accelerometer to detect the response. Accelerometer data are amplified and transmitted over a cable to frequency-filtering electronics and a notebook computer, where the resulting raw data are copied to the hard disk. Computer data processing and analysis software analyzes and displays the results in the form of a computer-generated map of the stator wedge. The map displays can be compared from one outage to the next, to track wedge tightness.

In one case, engineers using the instrument completed testing of a 900-MVA generator stator in two hours — far less than the eight hours required for manual hand-tap testing. Results were essentially the same, with 44% of wedges determined to be loose by manual hand tapping and 45% classified loose by the wedge tightness detector.

Monitor Gas-Turbine Blading. Infrared pyrometry is an increasingly popular tool for condition monitoring of gas turbines. For several years, high-performance military jet engines have used pyrometry for control purposes. In the industrial-turbine market, pyrometers are used to identify hot-section blading problems. Pyrometry detects and pinpoints individual blades that are running hotter than others in the same row.

Two approaches are commercially available for pyrometer installation. In the first, an insertion is made near a transition piece and the turbine blade is viewed through the stator nozzle. The second approach involves penetration of the high-pressure casing.

The condition of hot-section blading also can be determined by destructive metallurgical testing of blade samples. Nickel-based superalloys that are commonly used in gas-turbine hot sections contain a gamma prime phase that is precipitated during blade manufacture. With long-term exposure to over-temperature conditions, the gamma prime precipitate returns to solution, resulting in a loss of creep strength. Therefore, an assessment of blading condition and over-temperature conditions can be made by removing one or two blades from the turbine and destructively testing them for gamma prime precipitate solutioning.

Strain-gauge testing is a tool used by manufacturers to verify blading stresses and resolve compressor blading problems. Because the serviceability of strain gauges in the field is a significant problem, the technique is largely limited to test stand situations, although there are now field applications of the technology.

— Dr. Robert Peltier, PE, editor-in-chief.