When I Goggled “CMMS” I got 514,000 Web pages delivered. “Computerized Maintenance Management” returned 656,000 pages, and “Enterprise Asset Management” returned 3,890,000 pages. That is a lot of Web surfing. Besides, search engines like Google are now in the paid listing game—so the top results are not often the most relevant.

One good way to start your search is by learning what others are doing to be successful with CMMS and EAM. For details visit Maintenancebenchmarking.com. Look for the CMMS Best Practices link on the home page and download the pdf version. This study was conducted with more than 650 companies sharing what worked and, just as important, what did not work with their maintenance software implementations.

There are dozens of online CMMS software directories such as the MAINTENANCE TECHNOLOGY directory, the MaintenanceResources.com directory , and Cmmscity.

In addition to these Internet directories, the Web offers a new breed of helpful tools that include searchable databases of CMMS and EAM vendors and active comparison charting features that allow you to change the importance of several variables and more.

Software Evaluation is a new site that features a free spreadsheet that includes more CMMS/EAM software evaluation categories than you could want. It also suggests a scoring system. You can pare the list down to reflect the elements and issues that are important to you. The people behind this site have extensive experience with maintenance software and it shows. They also offer CMMS/EAM consulting services over the Web and e-mail at very reasonable prices.

Cmmscity has partnered with Technology Evaluation Center to create an intelligent CMMS knowledge base.

To use the service, you must register for a no-cost trial. Start by answering questions related to your specific software needs and desires. You can set the importance level of one element over another depending upon your requirements. Once you complete the questionnaire the knowledge base displays all the CMMS and EAM vendors who met your criteria.

You can select up to five vendors and dig even deeper into their capabilities and see how each supplier stacks up as you change the importance of the criteria you chose earlier. For example, if accounting integration was rated as less important during the initial software selection, you can change it to very important and see how that affects the recommendations about each of the five selected vendors.

There are also several good books on the subject that should be considered when evaluating new maintenance software. The connection to this Web column is that you can order them online. Visit Industrial Press and select the maintenance category. Check out Computerized Maintenance Management Systems by Terry Wireman (ISBN: 0-8311-3054-7) and Managing Factory Maintenance by Joel Levitt (ISBN: 0-8311-3189-6). Of course, Amazon.com also offers these books.

You also can attend conferences and learning events like the new SAP-centric EAM 2005 March 20-23, 2005 in Tucson, AZ; MARTS 2005—The Maintenance and Reliability Technology Summit May 23-26, 2005 in Chicago, IL; or CMMS 2005—The Computerized Maintenance Management Summit July 26-29, 2005 in Indianapolis, IN, to learn from peers and industry experts in a networked educational environment. MT

Installation of the permanently mounted fan balancing system (center) that continuously monitors fan vibration levels has enabled Boise to run the lime kiln with fewer production interruptions. Previously, massive build-up of calcium carbonate on the induction fan used for the kiln would throw the fan out of balance. The high level of vibration resulted in an average of 2.5 unscheduled shutdowns per year at a cost of $33,000-$35,000 each in lost production, makeup lime, and maintenance costs.

“When the fan started to vibrate, my coffee would actually shake in my cup as it sat on my desktop,” Blood said. “Shutting down is a serious process that typically takes 3 hours to allow the kiln to cool off, and another 3 hours to conduct the cleaning process. By the time we would get the kiln back in service, at least 10 hours of production were lost.”

In addition to lost time and production, frequent episodes of high vibration also were causing accelerated wear on the fan bearing, and the staff was often taken away from regular duties to troubleshoot vibration problems.

Searching for an answer
When faced with finding a solution to this situation, Blood remembered reading an article about a technology that might help solve his problem. After an Internet search, he located Lord Corp., Cary, NC, and called for more information. Andy Winzenz, sales manager for Lord, visited the plant and helped confirm the diagnosis.

“Bosie’s manufacturing process is dependent on the performance of the fans and their ability to maintain process air flow,” Winzenz said. “As such, when the fan was thrown out of balance because of build-up, the result was untimely and expensive shutdowns.”

After analyzing the problem, Winzenz recommended the company’s RealTime balancing technology—a permanently mounted fan balancing system that continuously monitors fan vibration levels. He recommended this balancing system because of its ability to make rapid balance corrections and to withstand the harsh environment surrounding the lime kiln ID fan.

How balancing ring works: The balancing ring of the system mounts permanently on the fan shaft. Vibration signals from the fan bearing are received and processed by a control system, which then determines the balance adjustments that are required. Insert heaters of the heavy chamber vaporize a high-density fluid, forcing it through transfer tubes to the opposite cooler chamber where it condenses into liquid, correcting the unbalance. This process continues until the controller senses that balance has been restored.

The balancing ring of the system mounts to the fan shaft. The ring houses liquid counterweight masses that can be repositioned to offset the unbalance detected in the fan rotor. Using vibration sensors, the system monitors the fan bearing vibration. Vibration signals are received and processed by a control system, which then determines the balance adjustments that are required. The controller relocates the counterweight masses to the desired position to minimize vibration levels. This process continues until the controller senses that balance has been restored. Typical balance cycle times range from 30-120 seconds, depending on operating speed.

Lord developed and patented the actuator coil assembly used in the balancing system. The actuator coil is traditionally mounted to support brackets located on the bearing pedestal. The noncontact power supply used in the actuator coil eliminates the need for maintenance, sending power across an air gap between the stationary actuator coil and the rotating balancer ring.

Implementing the solution
According to Blood, after the balancing technology was presented to the Boise team, a capital request was made and approved, allowing for an April 2003 installation during a regular scheduled shutdown. Installation involved moving the motor out of the way, pulling the coupling and bearing off the fan shaft, installing the balance ring, reassembling the bearing and coupling, and putting the motor back in and aligning it. Some minor trimming of the stiffening ribs on the fan housing had to be done to make clearance for the balance ring. Other work, such as installing power to the controller and then mounting it in a dustproof, waterproof box near the fan, was completed in advance of the shutdown.

Since installation, the team has endured only one unscheduled shutdown and that was caused by massive particulate build-up and throw-off. Blood said that after the installation, the fan ran so smoothly that the team forgot the fact that build-up was still happening.

“We received a wake-up call several months after installation when a large chunk of build-up flew off the fan,” Blood said. “The resulting vibration was more than the balancer could compensate for, so we had to shut down and sandblast.”

Because the balancing technology stores balance history and events, data can be analyzed—greatly aiding Boise’s process of calculating the build-up rate of the particulate in order to better plan for any necessary cleaning and sandblasting.

The proof is in the numbers
According to Blood, this process improvement has added up to big savings—an estimated $87,000 in 2003. Not only is Boise able to run the lime kiln with fewer production interruptions, it is also extending the life of the equipment and has minimized wear and tear on the fan bearings. Although Boise still shuts down the lime kiln three to four times per year for routine maintenance, such occurrences are typically planned and not in response to fan unbalance.

Blood can demonstrate the smoothness of the fan operation with the addition of the balancer by balancing a nickel on the edge of a fan bearing while in operation. Even more important, the production supervisors have little worries about the fan and are afforded the opportunity to perform their daily functions without the hassle of an unscheduled shutdown. Finally, the vibration figures speak for themselves. Before installation of the balancer, the fan registered 0.3-0.8 in./sec within 30-60 days of sandblasting. Today, however, Boise reports vibration levels of 0.04-0.06-in./sec within the same timeframe thanks to the new balancing technology.

“The new balancing technology paid for itself in six months, not taking into account the expense of wear and tear on the bearings,” Blood said. “Although the installation and learning curve had some glitches, the technology has more than met our expectations. A fan that was once a chronic problem and a constant worry is now one of the smoothest running pieces of rotating equipment in the mill.” MT

When properly used, thermography can reveal not only the liquid/gas interface, but also sludge buildup and floating materials such as waxes and foams. Similar techniques can be used to locate levels and bridging problems in silos containing fluidized solids.

This article discusses the parameters and limitations that must be addressed and explains techniques that can be employed.

Fig. 1. Levels can be seen in two tanks (left and center), along with differing solar influences, while the right tank appears to be full. Fig. 2. Sludge completely covers the manway opening in this tank in a paper mill. Anticipating this condition will result in maintenance strategies that are safer and more cost effective.

Why inspect?
Instrumentation for locating levels in tanks and silos is often unreliable. The need for precise information about levels remains necessary, or even critical, in many instances.

For example, prior to the arrival of a tanker ship it may be necessary to verify a liquid level in a large storage tank. In continuous processes the operator must know how much capacity is available in each tank. Without that knowledge production may be impeded or, if an overflow occurs, a potentially dangerous situation could be created.

Sometimes existing instrumentation cannot determine levels (Fig. 1). Foams and waxes, for instance, are difficult to detect and measure accurately. A report from a paper mill identified a situation in which a tank was believed to be sized improperly when in fact it was full of foam rather than liquid. Defoaming the tank proved more cost effective than unnecessarily replacing it with a larger one.

A thermographer working in a petrochemical plant relayed a story about a contractor hired to clean out a large tank (Fig. 2). When the manway door was opened, sludge, which had settled to a depth high above the door, oozed out creating a dangerous and environmentally damaging situation. For industries needing to comply with the safety and process requirements of OSHA 1910, thermography may prove to be a particularly cost-effective tool to use.

How does thermography help determine levels?
Most of the time, the materials in a tank or silo behave differently when subjected to a thermal transition (Fig. 3).

Fig. 3. As it transitions from night to day and back, a simple liquid/gas interface influenced by the outdoors ends up all the same temperature twice. In large tanks liquids do not change temperature greatly, but gases do.

The materials often have differing rates of thermal capacitance. Gases typically change temperature much more easily than liquids. Water, for instance, has a thermal capacity that is 3500 times greater than air. One Btu of energy added to a cubic foot of water will raise its temperature 0.016 F while the same energy added to the same volume of air results in a 55 F increase.

While the thermal capacity of solids may be similar to liquids, the different way heat is transferred allows them to be seen. Solids, such as sludge, are influenced primarily by conductive heat transfer vs fluids (nonsolids), which are strongly influenced by convective heat transfer. The result is that the layer of solids in close contact with the tank wall, despite its often high thermal capacitance, heats and cools more rapidly than the liquid portion because it does not mix in the same way the liquid does.

What conditions are necessary?
Key to determining levels is to observe the tank or silo during a thermal transition. If viewed while at steady state with their surroundings, no differences will be seen. In fact, tanks and silos that are full or empty often appear identical, i.e., no indication of a level.

It is difficult to find tanks or silos that are not in transition, although it may not always yield a detectable image. Outdoors, the day/night cycle often provides enough driving force to create detectable differences. Even indoors, variations in air temperature are often more significant than they might seem.

Environmental conditions can influence detectability. Wind, precipitation, ambient air temperature, and solar loading can all, separately or together, create or negate differences on the surface.

Other factors to be considered include the temperatures of the products being stored in or moved through the tanks and silos, as well as the rates at which they are moving. Many tanks are insulated, although rarely to the extent that they obliterate the thermal patterns caused by levels. When insulation is covered with unpainted metal cladding, care must be taken to increase emissivity as discussed below.

What thermal patterns will be seen?

Fig. 4. Thermography is an important, cost-effective tool to verify or locate tank levels. A straightforward gas/liquid interface is shown here.

The most obvious pattern is the liquid/gas interface (Fig. 4). In a situation where the product is not heated, the gas typically responds quickly to the transient situation while the liquid responds slowly. During the day it is warmer than the liquid and at night it is cooler.

Liquid/sludge relationships may be more difficult to discern (Fig. 5). A larger transient may be required to create a detectable image. Thin layers of sludge also may be indistinguishable from the tank bottom. Sludge buildup in the center of a tank, i.e., not in contact with the wall, is simply not detectable, although product buildup on the side walls often is quite obvious.

Foams are often not difficult to distinguish from liquids but may appear similarly to gases (Fig. 6). Care should be taken to push the tank through a rapid thermal transition to reveal the differences.

Locating levels associated with floating materials such as waxes will typically require more persistence, skill, and a greater rate of transitional heat transfer.

Whether or not liquid/liquid interfaces, such as a mix of oil and water, can be seen depends entirely on their differing thermal capacities and, to a lesser extent, their viscosity. Simple experiments suggest it is fairly easy to locate the interface of oil and water, but further work needs to be done in the field to validate this technique.

Fig. 5. Sludge buildup in this tank is substantial, nearly 20 ft deep, a condition that had not been well understood. The outflow is pumped on the right side. Fig. 6. In this black liquor tank both the gas/soap and the soap/liquid interfaces are visible. The level indication equipment had been reading the gas/soap level rather than the liquid level. Fig. 7. Two silos are being filled, on alternate days, with lime from a kiln. The silo on the right had been filled the previous day while the one on the left was being filled with hot product at the time the image was taken.

Some solids, such as coal ash, plastic pellets, powered lime, and wood chips, behave as fluids and are called fluidized solids. While heat transfer in such materials is still primarily conductive, mass transfer of heat by the material’s movement can be significant. For instance, hot ash or lime blown into a silo carries its process heat to the silo (Fig. 7). Fluidized solids tend to behave similarly to liquids in the way they respond to gravity, except for the fact that they can bridge areas where liquids typically would not. In fact, locating bridging of fluidized materials is a valuable use for thermography.

Issues to be considered
Some tanks are covered in cladding, often unpainted aluminum or stainless steel. Detecting the kind of fine temperature differences necessary to reveal levels on surfaces such as these—ones having low emissivity and high reflectivity—is nearly impossible. The radiant difference is not detectable.

The problem, however, is most often rectified by applying a high emissivity target vertically. A painted stripe or a piece of tape on the tank, for instance, can work well. For outdoor work, use light colors and/or the shady side of the equipment to avoid solar loading.

Occasionally tanks are heated or cooled with a jacket. These often prove impossible to work with. In some instances it may be possible to see the structural stand offs between the tank wall and the jacket.

Tanks which are insulated also can prove challenging. Insulation levels are typically not great enough that they preclude seeing levels; rather the insulation changes the thermal dynamics to the point where a detectable level may not be obvious as often. Simple techniques, explained below, can help enhance thermal differences so they can be detected. In some instances it may be possible to cut small plugs out of the insulation at various levels that would more clearly reveal the tank temperatures.

Fig. 8. The impact of solar heating is evident on the left side of the tank, although it does not fully obliterate the level. Note the level is more clear on the shady side.

Although solar loading can enhance a pattern, more often it can cause subtle thermal patterns in a tank or silo to be obliterated (Fig. 8). It may be possible to view the device on the shady side, but sometimes it may be necessary to return when the sun’s affect is lessened.

Spheroid tanks offer another type of challenge in that, when viewed from one point, their reflectance varies so widely over their curved surface. It is not unusual to find the top of these tanks appearing cooler while the bottom appears warmer; all too often both patterns are related more to reflectance than emission.

Tanks located inside of buildings are not subjected to diurnal heating cycles. Some thermal cycling usually does take place, but it may not be enough to make the radiant differences detectable. Again, simple techniques, explained below, can be used effectively to enhance surface temperature differences.

Enhancing thermal patterns

Often thermal patterns can be enhanced by using simple techniques to increase transient heat transfer. It may be possible to heat or cool the tank/silo or the surface of the tank/silo. The gas head in the tank responds more quickly than the liquid. As discussed earlier, solids may respond in a more complex manner.

Gas/liquid level Thermocline Sludge level Fig. 9. An unusual pattern can be seen in this tank where a thermocline has established itself in the liquid. A high-temperature flare in the background caused the “blooming” on the right side of the thermal image. (Courtesy Greg McIntosh, Snell Infrared Canada)

An industrial hot air gun can be used to heat the surface of small- to medium-sized tanks. Heating even a narrow area may dramatically reveal a level. Cooling can be provided simply by wetting the surface with water. As evaporation takes place, cooling drives transient heat flow and reveals or enhances the levels.

These techniques also are feasible for large tanks. Cooling in particular can easily be supplied with a spray of cold water hosed onto the tank surface. Add the element of time for the cooling to take effect and, in many cases, the image becomes readily apparent.

Get results
Many industries have a critical need to determine levels in tanks or silos or to validate existing level-indication instrumentation. In many instances, infrared thermography provides a cost-effective means of doing both. Conditions often allow for levels to be seen at almost any time of the night or day and throughout the year.

While levels are not always immediately obvious, persistence, careful imaging, and simple enhancement techniques can often produce remarkable results.

Acknowledgements
The authors would like to thank the following individuals for their assistance: Jeff Backer, Shane Brooker, Matt Clarke, Lee Colgrove, Jeff Cordova, Keith Dodderer, Patrick Lawrence, Greg McIntosh, Rob Spring, and Mark Soult. MT

How does your income match up with others in the maintenance and reliability community? It may be hard to find out where you stand because income figures vary so widely almost any way the data are tabulated.

INCOME DISTRIBUTION Half the respondents receive between $57,000 and $90,000. The average income is $74,995. INCOME BY AGE* Respondents 60 plus have the highest average income at $79,735. INCOME BY INVOLVEMENT* At $77,635, maintenance managers have higher average income compared to $72,208 last year. INCOME BY PLANT SIZE* Average income of readers in the largest plants (1000 or more employees) led this year’s sample. *The vertical brown line represents the income range within the group. The box shows the income range for the middle half (quartile two and three) of the respondents within the group. The green diamond within the box indicates the average income for the group.

That is what MAINTENANCE TECHNOLOGY Magazine has found out in all of its surveys of reader income. This year, respondents’ income ranged from less than $20,000 to more than $200,000, much broader than previous years. The lowest incomes were reported by practitioners working for nonindustrial facilities; highest incomes by executives with corporate or multiplant responsibilities in larger organizations.

Basic income profile
Average income of all readers responding to the survey was $74,995, somewhat more than the $71,153 registered last year and the $69,462 of the year before. The survey was conducted over a random sample of magazine readers, salaried and hourly, and we believe the data are representative of maintenance and reliability leadership.

Overall, 60 percent of survey respondents have worked in the maintenance trades or crafts. That level of craft experience was reported by respondents with a high level of responsibility: 44 percent for those who identified themselves with corporate/multiplant responsibility; 66 percent for plant managers; and 60 percent of maintenance managers.

Age and income profile
Half the respondents were 42 to 53 years old, with the average 47.3 years, older than the 46.6 years reported last year. Half the respondents received an income of between $57,000 and $90,000, more than last year. The income distribution chart illustrates the distribution in $10,000 groups. The midpoint was $74,000, lower than the average of $74,995.

The income by age chart that displays income distribution data vs age grouped by decades shows the wide variance of income within each of the groupings, with average income rising by age.

Education and registration
As expected, average income rose with the level of education. Average income rose from $67,260 for respondents with associate degrees, to $83,482 for respondents with bachelor degrees, and to $87,385 for respondents with advanced degrees.

Nearly 19 percent of respondents were registered professional engineers, certified maintenance and reliability

professionals, certified maintenance managers, or certified plant engineers. Average income of the professional engineer group was $86,509, roughly $4000 more than the average income of those with a maintenance manager or plant engineer certification, which were essentially the same.

Income by involvement
All respondents were involved in or responsible for plant equipment maintenance and reliability. That is the basic qualifying question on the application to receive MAINTENANCE TECHNOLOGY, and all respondents receive the magazine. However, they work at different levels and have varying responsibilities within the enterprise.

Respondents were asked to choose their level of involvement. Average income was $91,986 for corporate or multiplant involvement, $76,972 for plant management level, $77,635 for maintenance or reliability manager level, $68,021 for supervisor level, $76,971 for maintenance engineer, and $62,116 for technician level. The income by involvement chart shows a wide spread of income within each involvement sector.

Income by plant size
Respondents working in plants with 1000 or more employees received the highest average income at $80,220. The lowest average income occurred in plants with 50 to 99 and 100 to 249 employees, a pattern reflected in the data of earlier surveys.

Job responsibilities
Respondents were asked to indicate their job responsibilities by checking multiple items from a list. The portion of respondents checking various responsibilities in decreasing order are department performance, 65 percent; ordering or specifying plant equipment, 63 percent; ordering or specifying tools or supplies, 62 percent;

Hands-on troubleshooting of equipment, 61 percent; hands-on predictive maintenance analysis, 60 percent; time management and supervision of others, 59 percent;

Management of contract services, 58 percent; hands-on planning of maintenance work orders, 55 percent; department budgeting, 51 percent; engineering/design, 48 percent; hiring maintenance personnel, 45 percent; hands-on maintenance or repair of equipment, 44 percent.

The average span of responsibility included nearly seven out of the 12 responsibility items.

Respondents were contacted by e-mail and asked to visit a special Web site to fill out the survey questionnaire during a five-day period. The survey software did not collect respondent identification; however, respondents were given the opportunity to submit their email address separately to receive a copy of survey results.

For most questions, there were more than 990 usable responses. MT

There is a growing concern in the U.S. over a skill shortage in the technical trades crucial to run American industry. With many seasoned veterans planning retirement in the next 3-5 years, the question of how to transfer knowledge becomes more and more important. A well-planned and executed training program becomes the key to maintaining the level of expertise that is needed to keep American industry competitive.

Planning a training program
An effective training program starts with management support. Management should understand that there will be an up-front cost that has to be paid to put a program in place, but that the cost will usually be paid back quickly through numerous mechanisms, including decreased downtime because maintenance personnel have better skills to prevent equipment failures as well as less turnover in personnel who see the investment management is making in them.

A plan that describes the goals of the training program must be put together. The plan should draw upon information gathered in a training needs analysis (TNA), which is an exhaustive survey of plant operations, maintenance procedures, and equipment. It includes interviews with the technicians who maintain the equipment to determine the different skills they possess, and thus, the skills that should be maintained by anyone working on that equipment.

The TNA should include everything from union issues governing work practices to government regulations about plant operations to special ongoing problems in operations of all the different types of equipment in the plant.

From this needs analysis come the objectives of the training program. These objectives determine the exact nature of the needed training, which usually combines many methods, including classroom instruction, on-the-job training (OTJ), consulting for specialty subjects, and even outsourcing to companies that specialize in intense theory to practice-type training.

Classroom instruction
Classroom training offers a focused atmosphere in which to learn. The classroom curriculum should be based on the TNA. It should start with the basics: purpose and functions of equipment and basic operation, including theory of operation, how equipment interacts with the larger systems, etc.

The curriculum establishes a baseline of knowledge so that everyone is on the same level when more advanced training begins. Depending on the complexity of the equipment, this training could last anywhere from 1 hour to 1 week. Along with a good curriculum, the key to effective classroom training is a good instructor.

A good instructor has worked with the actual equipment and understands all aspects of it, from preventive maintenance required to how to troubleshoot unexpected problems. This person should understand the hazards involved in working with the equipment and the way the equipment interacts with other plant components. Safety should always be at the forefront of every discussion so as to incorporate it as part of the culture of plant technicians.

Before classroom instruction begins, the instructor should give a brief introduction of his or her experience to demonstrate expertise and understanding of what students will face in their day-to-day jobs. Next, students should give brief introductions of themselves to give the instructor an idea of the different skill levels and what they expect to get from the training. These introductions are crucial in setting up a comfortable place in which to learn.

The information should be complemented with visual aids, such as a computer-projected slide presentation, a white board or flip chart, or show-and-tell with parts from the machinery.

Also, there should be question-and-answer sessions throughout the presentation, and instructors should be able to answer questions effectively. It is common that if one student asks a question, over half of the class has the same question. Although a question from a student can disrupt the flow of the material, good instructors use the disruption to engage the class even more fully, and in doing so the instructor shows that questions are desired and will be answered in a way that will add to the comfort level of the learning situation. (See accompanying text “Steps to Answering Student Questions.”)

Other attributes that make a good instructor are the ability to communicate the curriculum clearly and concisely, the ability to maintain control of a classroom and keep the students engaged and on topic, and the desire to improve. A critique of the instructor should be done by all students at the end of the class so the instructor can learn what was done well and what can be done to improve the presentation. A good instructor and curriculum are crucial for the beginning stages of any effective training program.

Hands-on instruction
When a trainee understands the theory behind how a piece of equipment operates, the next step must be taken—on-the-job training. OJT reinforces what students learn in the classroom with a hands-on, learn-by-doing approach. Again, it is very important to have developed a curriculum by which to carry this out.

OJT can consist of students assisting qualified technicians in carrying out daily routines or assisting in preventive as well as repair maintenance. OJT must be incorporated into the work scheduling process and controls must be in place to ensure that it is effective and safe. The Occupational Safety & Health Administration, Environmental Protection Agency, and other regulatory agencies have guidelines and regulations involving training that discuss liability issues and the safety of workers.

Tracking progress
OJT should include a system in which to track a trainee’s progress. For example, a series of tasks could be listed in which the trainee has to participate under supervision of a qualified technician, who then would sign and date the task when it is accomplished to the supervisor’s satisfaction.

The curriculum for the OJT also should have a pre-existing set of questions that the qualified technician uses to quiz the trainee during the task being performed to ensure that the trainee fully understands all that is involved for each task. With an OJT program in place and running effectively, the knowledge transfer begins to take shape.

Follow up
To know whether the training program is effective and to continually improve upon it, follow up should be done with trainees. Surveys should be done periodically, possibly 30, 60, and 90 days after completing each segment of training.

These surveys should include questionnaires and quizzes that can gage the retention of knowledge by trainees and the applicability of what they have learned. They also should include a place for comments, so the trainees can make suggestions for improving the program.

Outsourcing and consulting
As stated earlier, outsourcing and consulting can be effective tools in setting up a training program. Experts in a particular field can assist in a variety of ways, from building the program from scratch to a more limited role, such as reviewing the safety controls in place to ensure regulatory compliance.

Some companies have already developed generic training curriculums with a wide range of industrial topics to choose from. These courses can be tailored to a specific arrangement or used as they are, the next step being that the organization will then build the OJT portion in-house.

Training pays off
The skills shortage is not an easy thing to fix. It takes management support, exhaustive TNAs, well-thought-out objectives and goals, participation from veteran technicians, and a program that employs the most effective means available. It includes an understanding of the complex workings of industry from regulatory compliance and safety to basic theory of machinery.

However, it is a continuing part of successful companies who understand that training does not cost—it pays. It pays through decreased downtime as technicians have the tools to troubleshoot costly problems, as well as decreased turnover as employees understand the investment the company is making in them. This endeavor, if done well, will keep the focus on smooth operations that, in the end, contribute to the success of the company. MT

Alan Lovett is the director of new product development for National Technology Transfer, Inc., P.O. Box 4558, Englewood, CO 80155; (800) 922-2820

SMRPCO (or certification) experience. As a member of the SMRP Certifying Organization board, I have been able to watch the impact that earning the CMRP status has had on several individuals and their companies. Some companies have implemented a policy of an automatic salary increase for those passing the exam. Others have provided public recognition and praise.

I believe it is fair to say that almost all companies have at least recognized the importance of the achievement for those individuals who have studied and worked to improve their skills and knowledge by becoming CMRPs. Has it made a difference in payback on their individual investment? I think so, if not immediately, certainly in the longer term. Whether in personal satisfaction, peer recognition, salary, or some combination of the three, I think almost every CMRP would tell you that it was well worth the effort.

Graduate studies experience. As a coordinator for our distance education (or off campus) Graduate Program in Maintenance Management and Reliability Engineering, I have seen a number of industry professionals earning degrees, or certificates, or simply participating in courses to further their education.

Most of these individuals report increased respect and recognition—and often, salary increases and new job opportunities due to their development endeavors. They continue with the program until they reach whichever goal they have set for themselves. Are they receiving the return on their investment? I think so, or they wouldn’t stay with the program. Here is what some of them say:

“I am sure that my participation in this course was a deciding factor in winning the job, and the knowledge I am continuing to gain will be more and more valuable.” (promoted to asset manager)

“The Graduate Certificate was instrumental in helping me win the job of production supervisor at a new gas plant. A great opportunity for further advancement, apart from about $40K increase in salary.” (promoted to supervisor)

College student experience. I have recently been reviewing our UT undergraduate intern reports from the recent summer as well as some feedback from two of our recent graduates from our maintenance and reliability engineering program. I think these excerpts tell the story better than I can:

“The extra specialization really helped when it came time to find a job during my senior year. When recruiters saw that I chose to deepen my knowledge in one particular area, they became more interested, and it oftentimes led to a second interview. Both at work and during the interviewing process, most of the engineers were surprised to see that I was familiar with various maintenance technologies.” maintenance engineer, May 2004 graduate

“Now that I have moved on to graduate school, my Maintenance and Reliability Certification impresses professors who thought I was just another face on campus. My research project focuses on diagnostic and prognostic analysis for condition-based maintenance for space shuttles. I believe that had I not gone through the MRC certification program, the opportunity to work with the manufacturing department at Georgia Tech would have been greatly reduced.” graduate student, May 2004 graduate

A Good Investment? So, does professional development pay? The answer is generally a resounding “yes” at all levels from seasoned veteran to university student. Certainly increased salary is a positive return. Promotion and/or new jobs also are normally considered great return on investment.

Air Liquide America (ALA) is part of the Air Liquide Group, the global leader in industrial and medical gases, headquartered in Paris, France. Its products include air gases such as nitrogen, oxygen, and argon as well as hydrogen, CO2, electricity, and steam. In the United States, the company maintains more than 125 production facilities and 700 customer installations.

Late in 1999, ALA’s management team realized that a higher reliability performance was key to existing and prospective clients. Concerned that current performance levels needed to be raised, a standardized benchmark assessment of maintenance and reliability capabilities was commissioned.

This article will summarize the original assessment in 2000 and the ensuing improvement efforts from 2000 through 2003. The focus is on the improvements made from 2000 through 2003 and the benchmark update in late 2003. The progress achieved by ALA over 3 years is highlighted.

Situation in 2000
Since the company’s beginnings in the United States, the maintenance function was decentralized and primarily the responsibility of the plant managers. The plants were supported by technical resources at headquarters, but were largely autonomous. Few reports or key performance indicators (KPIs) measured maintenance or reliability performance. Performance was mainly measured by two indicators—costs and headcount. This arrangement served ALA well for many years, but by the late 1990s, reliability issues began to affect customer satisfaction and maintenance costs were rising and unpredictable.

In 1999, ALA commissioned a regional maintenance concept designed to support up to two dozen sites from regional reliability centers (RCs). Initially the RCs were staffed primarily with a manager and technicians from the plant sites. Maintenance engineering support was provided as was planning and scheduling. However, the RCs were mostly reactive in nature, trying to provide resources for plant shutdowns and emergency responses.

After a year, the RCs were having moderate impact on reliability and maintenance costs remained unpredictable. As a result, the plants saw minimal value in the new centralized approach. Rather than abandon the effort, ALA executives decided to commission a Maintenance Benchmarking Study to define the issues that could accelerate progress.

Benchmarking process
Benchmark assessments usually involve the collection of pertinent data and a mechanism to validate the data. For most studies, a base of comparison data already exists. The challenge is to collect “apples-to-apples” client data. Using an unvalidated database can introduce a wide variation in data submissions for key information such as maintenance costs, replacement asset value, and personnel counts. These variations, in turn, can significantly affect comparisons and interpretations.

In the ALA studies in 2000 and 2003, all data was validated through on-site review of definitions, data reconciliation, and interviews. Similarly, comparison data also was validated project-by-project to ensure that comparisons were as consistent as possible.

On-site validation not only provides an opportunity to validate submitted data, but also allows observation of maintenance practices. In reality, the validation visit provides the opportunity to:
• Validate data
• Interview personnel
• Tour and observe the plant and its conditions
• Develop data comparisons and key issues in a team-based environment
• Draw conclusions that the plant team understands and supports
• Develop consensus lists of plant strengths and improvement opportunities

Interviews conducted during the site visits allow the process to move beyond collection and comparison of data. The interviews typically highlight problem areas, obstacles to improvement, and, very often, support conclusions implied by data comparisons.

When the on-site work is completed and the benchmark team has discussed the issues and key improvement needs, the assessment report documentation begins. At this stage, the team understands the hard number comparisons and the key areas for improvement. The final report and the subsequent presentation to management are designed to highlight issues to be addressed and resources required. The report also quantifies the potential financial gain from improvements.

Benchmark 2000 conclusions
In applying the benchmark assessment process at ALA in 2000, standard techniques were used with accommodations for the typically smaller size of ALA’s United States plant locations. The same validation processes, interviews, and team approaches were applied, as described above.

The areas identified for improvement from the initial 2000 assessment were:
• Improve cost control through improved reliability
• Coordinate maintenance and reliability with capital projects
• Restore key support resources
• Redesign the reliability center concept
• Strengthen or replace the computerized maintenance management system (CMMS)
• Develop a contractor management strategy
• Institutionalize root cause failure analysis (RCFA)
• Perform reliability centered maintenance (RCM) analyses
• Institute work planning and work scheduling
• Strengthen spare parts management

Benchmark comparisons were provided in the assessment to allow the ALA team to gauge appropriate staffing levels for direct maintenance as well as for support functions including planning, reliability improvement, and parts management. The comparisons with external data and practices also provided a frame of reference for total maintenance costs, maintenance organization structures, and maintenance philosophies.

Given the number of issues highlighted by the study, it was clear that ALA would have to prioritize its targets. A potential cost reduction of up to 25 percent was identified, but it would come slowly, given the economic downturn in 2000.

Using the benchmarking report as a basis, a maintenance improvement team began to develop strategies and implementation plans. The elements of that strategy are described below.

Gaining control of the work
Figure 1 shows that at the start of the change process, the plant audits were generating 50 percent breakdown work. After applying the new tools and processes, ALA would control the work and equipment to the point where 90 percent of the work would be plannable.

The level of existing emergency work required a significant amount of resources and overtime. In addition, plant turnaround performance was inconsistent, hampering cost control and causing additional overtime. Before any of the improvements could be implemented, it was imperative that the plants and the RCs gain control of the work.

The RCs decided to concentrate on improving shutdown performance. Experienced professional planners were hired to supplement the existing planners, who were provided training. Planning tools were developed and planners began the planning process months in advance. To better manage the shutdowns in the field, additional field supervisors were hired and developed.

However, the most important aspect was the consolidation of quality contractors. ALA turnarounds were small compared to those of our large customers such as refineries, chemical complexes, and steel mills. ALA plants usually shut down when their customers do. However, because the customers dominated the labor market during shutdowns, ALA was often left with few quality contractors.

ALA immediately identified three high-quality contractors to do general mechanical, high voltage electrical, and major compressor repairs, and signed national contracts with them. These efforts gave ALA improved planning, good field supervision, and quality contractors. In less than a year, a measurable improvement in turnaround performance was observed.

There was not much that could be done about the frequency of emergency breakdowns at this point, but they could be better managed and investigated. Maintenance and reliability engineers were hired at each RC. These engineers were assigned to provide engineering analysis to determine the optimum scope for the repair.

The new planning resources provided some planning, and the increased field supervision was deployed to better manage emergency repairs. The new national contractors improved the quality of repairs and RC performance in emergency situations started to improve. The new maintenance and reliability engineers were all sent to RCFA training and began to perform RCFAs on incidents in an effort to understand and prevent them in the future.

Implementing a process
One of the major barriers to improved coordination and communication between the plants and the RCs was a lack of process by which they would work together. There were no agreed upon roles and responsibilities. Each group did what it thought it should be doing.

A simple maintenance management process was developed that included work identification, approval, planning, scheduling, execution, and documentation. Each of these functions was broken down into tasks. For each task, roles and responsibilities were spelled out in RACIs. RACI is a simple tool that details for each task who is responsible (R) for doing the task, who is accountable (A) to see that the task is completed, who is consulted (C), and who is informed (I) . Training manuals were developed and all RC and plant employees were trained.

Soon thereafter much of the wasted energy and confusion, trying to determine who was going to do what, began to disappear. Employees began to better understand their roles and how their roles supported the process.

As new and unexpected issues arose, collaborative troubleshooting took place. Improvements in the process were documented and after a year the process was reviewed, the manual was revised, and employees were retrained.

CMMS implementation
The study revealed that the legacy CMMS was not being fully utilized. The strategy development team evaluated reconfiguring the system or replacing it with a new one. After a thorough analysis, the team recommended replacing it with a new version of Maximo from MRO Software, Bedford, MA.

The new software was configured to support the new processes. Asset data was gathered from the sites and placed in a multilevel hierarchy. Engineers entered legacy PMs where they existed and created new ones where they did not. The workflow feature of the software was mapped to route work orders for approval according to the approval limits set forth in the company’s delegation of authority. The software was interfaced with the company’s financial software to enable cost tracking and enable the purchasing and inventory modules. Virtually all software functionality was exploited.

This CMMS project is discussed in the article “Enhancing an Enterprise Asset Management Project”.

Vendor consolidation
Because the maintenance function was decentralized for many years, the company employed hundreds of mostly small local vendors. Most of these vendors were responsive to the local site but many lacked the level of quality needed to improve long-term reliability of the company’s assets. Furthermore, the fragmented service providers did not leverage ALA’s expense in the basic maintenance services.

The maintenance department teamed with corporate supply management and began to identify common services used at all of the sites. Collaborative cross-functional teams of experts and stakeholders were assembled to evaluate potential service providers.

A matrix of critical success factors was weighted and agreed upon by the team, and potential service providers were identified based on the quality of their work with ALA and the potential of providing their services on a national basis. Each potential service provider submitted a proposal and made a detailed presentation to the team.

Based on the data and final matrix score, a finalist was selected and formal negotiations took place. After negotiations were completed, the contacts were signed for a 3- to 6-year period. Both ALA and vendor KPIs were established to assure that both parties benefited from the new agreements. For ALA, KPIs were established to measure quality and savings. For the vendor, leakage and timeliness of invoice processing were measured.

Prior to rollout, each agreement was communicated to both maintenance and operations employees. Service providers handed out catalogs, normal and emergency contacts and phone numbers, and answered questions. Problems that arose were quickly addressed by the team to ensure that small issues did not escalate into bigger ones.

After 2 years, 17 contracts were signed including gas turbine maintenance, general mechanical services, compressor maintenance, electric motor shop repairs, safety relief valve management, electrical switchgear PMs, inspection services, transformer oil analysis, vibration monitoring, MRO supplies, and expander turbine shop repairs.

Organization changes
As it existed in 2000, the maintenance organization was comprised mostly of technicians and specialists. To support the new processes and tools, the RCs needed to be reorganized. The new organization increased support in the areas of engineering support, and work planning and scheduling.

Additional maintenance engineering resources were hired in each RC to apply engineering principles to determine the optimum repair scope for equipment failures or equipment with known component degradation. They were required to document the repair scope and write detailed job plans in the CMMS. They followed the repairs, examined the parts, and made necessary adjustments to the scope and plan based on what they found. They documented the repairs and conducted RCFAs.

Reliability engineering resources also were hired to monitor and improve equipment health. They were assigned to monitor vibration, oil condition, and infrared scans for the plants in their zone. For each recommendation, they were to write work orders in the CMMS with target dates based on their review of the data.

In the first year, more than 500 interventions were made prior to failure, saving large amounts of downtime and significantly reducing costs. The reliability engineers also were assigned to do an in-house version of streamlined RCM called a vulnerability study. For each piece of equipment, failure modes were identified and the appropriate predictive or preventive activity was identified. Spare part stocking levels were determined based on criticality and lead time.

Work planning was also strengthened. Additional planners and schedulers were assigned to each RC. All RC employees attended planner training to instill the importance of the activity throughout the organization. Additional schedulers worked with production to plan work on a weekly and monthly basis, according to business needs and equipment availability.

At the technician level, both in-house and outsourced competencies were determined. It was decided that ALA needed to keep instrument, electrical, and control system technicians in-house because the intimate knowledge of plant processes could not be readily obtained off the street. On the other hand, it was decided that almost all mechanical maintenance services could be safely outsourced because of the high quality of service providers available.

Importing and developing talent
Perhaps the most critical success factor was the hiring of new talent to fill many of the key positions and the development of the substantial technical talent that existed.

The new department inherited some outstanding technical talent, but there were not enough resources to meet the needs and there were some gaps in certain areas. The gaps were identified and an extensive talent search began. Over 2 years, some of the best technical talent in the nation was recruited to join the team. Together these technical specialists made an immediate impact better managing major overhauls and repairs.

However, the most critical positions needing an immediate influx of talent were the five RC managers. In the early stages of the implementation, several of the technical specialists were used for these positions because it was thought that these senior-level engineers would be the perfect mentors for the influx of younger talent being hired as maintenance and reliability engineers.

However, this was not the best use of their talents. What was needed was managers experienced with change management and implementing maintenance and reliability best practices.

In 2002 and 2003, several new RC managers were hired with this type of experience. Most of them had successfully passed the Society for Maintenance & Reliability Professionals’ certification examination and were Certified Maintenance and Reliability Professionals (CMRP).

Each of these new managers was able to articulate to their employees and counterpart in production how and why we were doing what we were doing and could paint a vision of what the end state should look like. They focused on roles and responsibilities and did extensive troubleshooting of issues and work processes that hampered performance. Once on board, the new managers applied the knowledge and experience of the new and existing technical specialists in a more effective manner, focusing on long-term reliability improvement.

The new department also recruited some world-class planners to lead the new planning efforts and to serve as mentors for the fleet of planners in training. After 2 years, the depth of planning talent has increased significantly. These new planners were critical in troubleshooting issues that arose in the new planning and scheduling processes.

The new department also filled entry-level engineering positions with extremely bright young engineers. With their enthusiasm for their assignments and the new managers’ ability to use them effectively, their impact exceeded expectations and made the future of the department bright.

Condition monitoring
By 2002, the focus was on improving condition monitoring. The benchmark study indicated that the existing vibration program was not optimal, primarily because data was not collected frequently enough to avoid most failures.

A team was assembled to design a state-of-the-art system that could increase the frequency of data collection and implement a process that assured action prior to failure. The idea was to partner with a nationwide company that could provide data collection, analysis, issue reports and provide ad hoc troubleshooting.

After evaluating several proposals, the team selected Rockwell Automation as that partner. The program provides three full time and dozens of part time Level 2 (Vibration Institute Certification) vibration technicians to collect data at all plants on a monthly basis and issue monthly reports to the reliability engineers. The program also provides a full time program manager who is located at ALA headquarters. This program manager is considered a valuable member of the ALA maintenance leadership team, just as if he was an RC manager.

KPIs were established to assure timeliness of data collection, the issue of reports, the review of the reports, completion of recommendations, and the number of saves attributed to the program. As mentioned earlier, more than 500 work orders were issued that intervened in component degradation prior to failure.

The program has since expanded to include infrared scanning and oil condition monitoring. The predictive maintenance data and reports for all three technologies are kept on a single database accessed from the ALA intranet. ALA technicians, engineers, and managers anywhere in the world can access this data for analysis or troubleshooting of issues with any piece of equipment at any plant in the United States.

Given the challenging geography of the assets, this is an invaluable resource of information that is constantly being accessed to address and preempt equipment component degradation and avoid breakdowns.

KPIs and monthly reports
Once the new people, tools, and processes were in place, KPIs and monthly management reports were created to drive improvement. As ALA defined them, KPIs are quite different from monthly reports.

KPIs were initially established to measure the effectiveness of the initial implementation and subsequent adjustments to the processes. For example, during initial implementation of the maintenance management process (MMP), it was discovered that simply writing work orders for maintenance work was inconsistent. So a maintenance work order compliance measure was developed for each plant and zone. This identified plants with high and low compliance. Plants with high compliance were recognized publicly and additional training and coaching were provided for plants with lower compliance.

As work order compliance increased, the next problem arose—the lack of resources and skills to adequately plan the increased amount of work orders. There were work order planning KPIs for volume and quality but when work order compliance was low, the planning KPIs looked pretty good. As the planning KPIs fell, additional planners were hired and others were trained until the volume and quality improved.

As planning volume increased, it was discovered that the scheduling process was under-performing. Managers, planners, and engineers worked hard evaluating and troubleshooting the issues and bottlenecks in the scheduling process until it became streamlined and effective.

This process was called “getting around the bases.” In other words, we had to get to first base before we could reach second base. And we had to get to second and third base before we could get to home plate and score a “run.” A run was defined as getting a well-written work order; having it properly approved, well planned, scheduled days to weeks in advance; executing it on time and with high quality; and documenting the problem, cause, and remedy. In this environment, KPIs were use for tactical improvement and modified weekly to measure new issues as they arose.

As problems were solved and improvements became sustainable, old KPIs were dropped to make room for new ones. To assure that the improvements were institutionalized, a higher level of the most critical KPIs was incorporated into monthly reports for management and corporate executives.

The relatively few monthly report measures were much different than the numerous KPIs. Instead of being malleable measures used at the field level to measure the implementation issues of the day, the monthly reports measured performance at a higher level and were held consistent for at least a full year. They were used to assure management that the new processes were working and improving.

The monthly reports consist of two sets of measures—one set for the MMP and one set for equipment reliability. For the MMP, four items are measured:
• Age of the work order backlog. This measures the age of each work order so it can be determined if the volume of work orders exceeds the capacity of the RCs to complete work orders. The measure easily shows how many work orders come into the system each week, are completed, and grow old past the target of 9 weeks (excluding turnaround work orders).
• Preventive maintenance work order scheduling compliance. This measure shows compliance to the schedule date of PM work orders.
• Predictive maintenance work order target compliance. This measure shows compliance to the target date of work orders that were written as the result of a discovery by a condition monitoring report.
• Maintenance activity type. This measures the percentage of man-hours worked for each of the maintenance management plan work types: normal (planned at least a week in advance), preventive maintenance, urgent (planned less than a week in advance), and emergency (no planning). (See Fig. 2.)

The equipment reliability measures are a series of mean time between repair (MTBR) calculations for our most critical equipment. These classes and subclasses are:
• Compressors—large (>2000 hp), medium (200-2000 hp), and small (<200 hp)
• Motors—large (>2000 hp), medium (200-2000 hp), and small (<200 hp)
• Expander turbines—All

These calculations take the number of significant work orders generated for a class of equipment and divide it by the number of pieces of equipment in that class. Because spared equipment is less prevalent in our industry, we do not attempt to account for on-line spares or actual run time. What the metric lacks in pure theoretical accuracy it more than makes up for in consistency. Therefore, its repeatability makes it ideal for internal continuous improvement. The thought is that if the number of significant work orders is reduced, equipment reliability must improve.

Each of the MMP and equipment reliability measures is published in charts on the ALA intranet for all employees to see. Each has the ability to show any of the measures by plant, production zone, maintenance zone, or business unit. This allows any manager at any level in the company to view a customized report with his staff anywhere in the country (see Fig. 4). Managers can now set improvement targets for any plant or group of plants in a zone or business unit.

ALA also has an extensive set of production availability and production disruption reports generated by other departments. Coupled with these measures from the maintenance department’s databases, ALA now has a balanced and comprehensive view of key performance indicators.

Designing the rebenchmark study
In late 2003, ALA decided it was time for an external assessment of improvement progress. The benchmark approach was used again to get a consistent, 3-year progress analysis. The team approach was applied as it was in 2000. Data collection was conducted on the same basis as in 2000. The focus of this assessment was to gauge the extent of improvements over the 3-year period and to bring attention to lingering issues still to be addressed.

Certainly, we all wanted to identify progress and recognize the hard work invested by the entire organization. Once again, quantifying the continuing financial stake was an objective. Another key objective in the rebenchmark study was to provide input for a more formal strategic plan.

With the list of issues from the 2000 assessment, ALA opted to tackle the low hanging fruit without the delay imposed by developing a formal strategy. The expectation in the 2003 assessment was that the remaining issues would likely be more difficult to address and might require a longer-term investment in shifting the culture. The concept of a 3-5 year formal strategic plan approach forces people to keep all issues on their radar, but follow a disciplined schedule of tasks and resources.

2003 results
Substantial improvements were made in the 3½ years between the initial benchmark study and the follow-up study in 2003. Key improvements included:
• Personnel changes to facilitate planning and reliability improvement
• More disciplined processes for planning, reliability analysis, and work scheduling
• Better coordination of contractors
• Improved material management
• A comprehensive set of performance tracking measures
• A substantially improved morale and teamwork environment
• An emerging reliability culture in place with support at the highest levels of management.
• A number of technology tools supporting improved practices including a new CMMS, vibration technology with routes and analysis, RCFA, vulnerability studies, and improved cost definitions, reporting, and analysis
• National agreements with key service providers and suppliers

Of course, there are still issues to improve going forward.
• Develop the new warehousing system into a comprehensive set of integrated national and regional warehouses
• Continue to support the growth of the new reliability culture by finding ways for maintenance, operations, and engineering to collaborate on solving reliability issues and integrating the operating technicians into the new reliability systems.
• Develop a mechanical integrity program to improve reliability for fixed equipment after focusing on improving rotating equipment over the past few years.
• Take the new systems such as the CMMS, the RCFA process, and condition monitoring to the next level.

Measurable progress
Between the recent benchmarking report and the new monthly reports, progress in several areas can now be validated and, more importantly, can drive continuous improvement.

Work order execution. Putting the MMP in place with its defined roles and responsibilities and implementing the CMMS to support it has dramatically increased the department’s ability to do more work more efficiently. While the work order volume was not known prior to the changes, the documented number of work orders completed each month has more than doubled in the past 18 months while reducing costs.

The primary reason is that the two most efficient MMP streams of work, PMs and normal (full planning), have increased from below 50 percent to greater than 75 percent, and are fast approaching the target of 80 percent. Consequently, the most inefficient streams, urgent (minimal planning) and emergency (no planning), are now less than 25 percent with the least efficient emergency work now at less than the target 10 percent.

The next opportunity is to reduce the urgent work to below 10 percent so PM compliance can be increased.

Work order backlog. When the MMP and CMMS went live, the number of work orders being written skyrocketed as previously undocumented maintenance work became documented.

For about a year, the size of the backlog increased steadily each month. Then in 2003, as the influence of the MMP and the CMMS began to grow, the size of the increase began to lessen. By the beginning of 2004, the increase stopped and the total backlog remains remarkably constant.

This is another indicator that the ability to complete work orders has equaled the volume of work orders created. The next opportunity is to continue to increase efficiencies so that the number completed exceeds the number created and the work order backlog can be reduced.

PM work order schedule compliance. Again, when the MMP and CMMS went live, the number of PM work orders increased. As expected, initial PM schedule compliance was low. However, the volume of completed PM work orders also has increased and is now four times what it was a year ago. This has increased PM compliance to about 80 percent in some months, but it is not consistently at that level.

This is related to the number of urgent work orders being above target. If the amount of urgent work can be reduced, the 90 percent target for this measure should be attainable.

PdM work order schedule compliance. When the condition monitoring programs were put in place, they initially generated large volumes of work orders because far more equipment was being looked at far more frequently than ever before. As expected, initial schedule compliance for these work orders was lower than desired. But as the work order capacity of the department grew, the backlog of these work orders was minimized and compliance is consistently up from 50 percent to more than 75 percent, and is fast approaching the 90 percent target.

Maintenance costs. Maintenance costs decreased between 2000 and 2003 despite doing more maintenance work. All the efforts from implementing the MMP and CMMS, to consolidating vendors, to increasing the engineering, managerial, and planning talent have played a role.

Downtime for planned maintenance. The amount of planned downtime has more than doubled in the past 4 years, confirming the fact that far more work is being controlled than ever before. The ability to do this work on our own terms instead of the equipment’s terms not only improves the efficiency of maintenance, but also all of the other business unit costs associated with taking a plant out of service. Now that major maintenance is more defined and better planned, it can be better coordinated with customers and the internal supply chain, reducing costs and increasing profits for both.

Supply availability. Like most everyone, during the recent downturn in the United States economy, business volumes were lower than they are today. Therefore, supply availability remained high due to the availability of excess capacity. Now that the economy has picked up markedly, our volumes have steadily increased and in some business units are at record levels. This has put a premium on the reliability it takes to operate at these rates with less backup capacity. During the past 4 years the traditional high levels of supply availability have held steady at these higher rates, again confirming that the investment the company made in the area of maintenance and reliability during a down economy has positioned it well for an improved one.

Reliability. ALA now has two new measures for reliability. The one at the plant level is the number of production incidents, no matter how small and no matter if they had any affect on supply availability. At the equipment level, the MTBR is measured for all motors, compressors, and turbine expanders. Both measures are too new to give a trend, so the 2004 data is being validated and improvement targets for 2005 are being set.

The future
ALA still has several areas to complete and optimize along this never-ending journey toward reliability excellence. In the short term, Air Liquide Group has completed the acquisition of Messer Griesheim’s operations in Germany, the United Kingdom, and the United States (MG Industries).

In the United States, more than two dozen production facilities will be gained. These plants will be incorporated into the MMP, CMMS, condition monitoring program, PM program, and the national service and supply agreements. In addition, at least four new plants under construction must also be incorporated into these systems.

The commissioning of a new national warehouse in Houston has been completed, and now regional warehouses need to be created to hold spare parts currently located at the plants. This is a huge project that will tell what is on hand. In parallel, vulnerability studies will be completed at all our sites over the next 2 years. Among other things, this will tell what we need. Then the spare parts stocks can be optimized from what we have to what we need.

A cross-functional team of operations, maintenance, and engineering has been commissioned that will determine and define collaborative reliability activities to instill a reliability culture throughout the company. The team will explore several elements of total productive maintenance including operations-driven reliability, equipment improvement teams, early equipment management, and the increasing use of RCFA.

Earlier this year a mechanical integrity project was approved and staffed. This team will implement a more formal and comprehensive mechanical integrity program. They also will institutionalize all of the new tools and processes with new maintenance and reliability policies and procedures and establish a formal audit protocol that will ensure compliance and improvement.

And, benchmarking approximately every 3 years will always be a part of the long-term plan to validate progress, see where other companies are moving, and adjust the strategic direction to meet business needs and incorporate industry best practices.

At ALA, it is our firm belief that our investments in improving reliability will separate us from our competition and, in the end, make both ALA and its customers stronger and more profitable partners. MT

Mark E. Lawrence, PE, CMRP, is director of maintenance and reliability, Air Liquide America, LP, 2700 Post Oak Blvd., Suite 1800, Houston, TX 77056; telephone (713) 624-8181. Edwin K. Jones, PE, is a consultant who can be reached at telephone (863) 699-9196

Early maintenance process

Fig. 1. This is an early ALA strategy to gain control of maintenance work by applying planning and scheduling to half of the work that could be planned and scheduled and using national contractors and improved supervision on both the planned and emergency work. For gaining control of the equipment, the planned work would flow into CMMS work history and the unplanned work would flow into RCFAs. RCM would proactively generate the PM and PdM programs and plant assessments would detail needed restoration. All of this would reside in the CMMS and the outcome would maximize planned work and minimize breakdowns.

back to article

Maintenance work categories

Fig. 2. The maintenance management process designated four types of maintenance work: emergency (no planning) where work starts immediately, urgent (minimal planning) where work starts in 24-72 hr, normal (full planning) where work is scheduled weeks in advance, and preventive maintenance (preplanned and approved) that can go straight to scheduling.

back to article

Task Roles and Responsibilities

Fig. 3. RACI chart is used to explain the roles and responsibilities of all employees involved in the maintenance management process. R = responsible, A = accountable, C = consulted, and I = informed

back to article

Completed work report

Fig. 4. Sample monthly report shows the percent of man-hours spent on the four maintenance management process flow paths. These reports have drill-down capability to the business unit and plant level and can be viewed by any ALA employee anywhere in the country.

back to article

2003 benchmark improvements

Fig. 5. These two benchmarking charts showing estimated replacement value per planner and per engineering support person confirm that both planning and engineering resources were brought more into line with world class companies in the benchmark database.

Robert C. Baldwin, CMRP, Editor

MAINTENANCE TECHNOLOGY surveyed its readers to find general information about current practice in certain areas of plant equipment maintenance, reliability, and asset management. It was a part of our annual survey of maintenance salaries.

We found that most respondents are using multiple predictive maintenance or condition monitoring technologies, which was expected. Infrared thermography led the list with 74 percent of respondents reporting they are using or have used it in their facility. Oil and fluid analysis was a close second with 73 percent using, followed by vibration monitoring and analysis at 66 percent.

Less than half the respondents were using or had used each of the other listed technologies: Ultrasound inspection, motor circuit analysis, electric power monitoring, and process parameter monitoring. However, when figures for respondents who were considering using the technologies were included, each technology was being considered or had been used by at least two-thirds of the respondents.

We also asked readers about their use of several maintenance and reliability tools and techniques: Reliability centered maintenance (RCM), total productive maintenance (TPM), root cause analysis (RCA), PM optimization, benchmarking and key performance indicators, precision maintenance, and Six Sigma. For each tool, more than half the respondents said they were using or dabbling in it. The success rate was best for precision maintenance (precision alignment and balancing), where 82 percent of those using it reported moderate or extreme success. PM optimization was second at 76 percent and root cause analysis third at 68 percent.

The overall failure rate in applying predictive maintenance technologies and maintenance reliability tools was about 10 percent, which suggests that all techniques are providing significant value to the approximately 1000 practitioners who responded to the survey.

We all know there are many plants in the fail-and-fix repair mode that have yet to discover the predict-and-prevent world of modern maintenance. And, thanks to the readers who shared their personal information, we are confident that you are making progress with predict-and-prevent.

We wish you continued success in 2005. MT

The clutch is being used on drive units that power a vertical lift and transfer system in the assembly plant where the Saturn VUE small sport utility vehicle and the Saturn ION sedan and quad coupe are built. This assembly is part of a condition monitoring program that is being used to optimize maintenance at the facility.

Avoiding downtime
Significant downtime has been attributed to the vertical lifts and transfers which move the vehicles through the assembly process. These enormous drive units carry large amounts of weight which causes massive torque loads and sizable vibrations. These drives can fail for a variety of reasons: gearbox failures, motor failures, bearings, sprockets, etc.

If one of these drives fails, the line stops. Although most units are equipped with a backup drive, it can take maintenance personnel anywhere from 45 min to 1 hr to manually switch to the backup. A projected loss of up to $3500/min makes this downtime extremely expensive.

Prototype installed
A prototype clutch, designed jointly by Autogard Corp., Rockford, IL, and GM Spring Hill, was installed on selected lifts. The clutch quickly switches between drives and can be coupled with a torque ring and telemetry system for continuous monitoring of the torque and velocity of the drive shaft.

Using the torque ring outputs, the primary drive can be switched automatically to the secondary drive in 1 min rather than the 45 min at minimum that it takes for a manual changeover. Because this switch occurs automatically, the operations continue to run smoothly, allowing the maintenance personnel to redirect their work to other priority areas.

Fig. 1. Right side angle shot with lifting carriage in full down position

Reduced interruptions in the flow of parts to the line and the stream of vehicles through the plant results in improved uptime and lower operating costs. The end result is that the maintenance department and operations management see improved efficiencies that favorably impact the bottom line.

Monitoring identifies problems
The data ring monitoring system provides key running load information on the condition of the vertical transfer. In one instance, information from the new unit identified a problem caused by a gearbox failure during the first 5 min of operation. The limit switches and spring preloads on the vertical transfer were adjusted incorrectly resulting in overtorque of the gearboxes by a factor of four times the maximum rating of the gearbox.

The monitoring system was used to aid in the proper adjustment of the limit switches and preload springs to bring the elevator back to its designed parameters. The system facilitates the detection of developing problems and allows the scheduling of maintenance based on conditions requiring attention before an impact to production occurs.

The result has been improved preventive maintenance scheduling, more efficient use of maintenance personnel, and early detection of mechanical problems through continual monitoring of the equipment.

Once the torque ring receiver is attached to a programmable logic controller, a computer program is then able to monitor the torque and velocity readings. If the threshold is reached, the system warns the maintenance team that something is about to fail. The monitor can be used to determine how much counterweight is needed.

Lessons learned
While considering the implementation of this project throughout the plant, several key lessons were learned:

• If designed correctly, the clutch can be an extremely cost effective, efficient, and useful way to switch to the backup drives. In the current situation, it paid for itself the first time that it was used.
  
• The torque monitor is not always required for brake solutions.
  
• The torque monitor is not required for clutch implementations but without the data ring, the clutch cannot switch automatically between the primary and backup drive. With the torque monitor, the clutch switches to the backup during a failure, and provides an indication that the primary has failed. The data ring also can warn of imminent failure, allowing repair or replacement of the primary drive during scheduled downtime.
  
• If a condition based monitoring system is in place, it is not imperative to have a backup. However, it should be evaluated on a case-by-case basis, depending on the criticality of the equipment and the efficiency of the maintenance organization. If planned downtime and occasional unplanned downtime can be tolerated, then the data ring should be sufficient.

Understanding each unique environment will help to ascertain which components and strategies need to be implemented. MT

Thomas A. Rogers, Ph.D, is vehicle systems engineer at General Motors’ General Motors Spring Hill Manufacturing, 100 Saturn Parkway , P.O. Box 1500, Spring Hill, TN 37174-1500. He can be reached at (931) 486-6782