Reliability Improvement For The New Millennium
By Chris Traianou, Senior Reliability Planner
Western Australia’s Water Corporation
Within many large scale plant based industries, maintenance costs can account for as much as 40% of an operational budget. The maintenance effort is therefore easily identified at a corporate level as a source of savings. Costs in maintenance can be cut in either a beneficial, or a detrimental manner. The best business outcome would be to both reduce costs and optimise current maintenance effort to increase reliability. It is in this environment that Western Australia’s Water Corporation (covering the vast State of Western Australia), has embarked on a project-based reliability improvement initiative.
More...Click here to read Reliability Improvement For The New Millennium (100k pdf)
Steve Shores
Vice President, Asset Management Solutions
The DEI Group
What will the reliability tools of the future be? How will maintenance managers and engineers use these tools to better support production requirements? Most companies currently work on a day to day basis on maintenance activities. Very few companies actually create a schedule in advance that they can stick to. But, what if there were ways to accurately predict when equipment will fail?
Condition monitoring test equipment has for years been touted as being able to predict failures. In fact, though, each condition monitoring tool when used as a predictive maintenance tool can only look at some of the failure causes for equipment. As companies expand their data collection capabilities, they collect more data than can be assimilated by a typical organization. They start asking the question “how do we get knowledge from this abundance of data?” Many software packages have been created to store and analyze subsets of this data. Proposed solutions for handling some of the data have ranged from a computerized maintenance management system to plant historians to enterprise asset management systems to data warehouses. However, most solutions that have been implemented to date have been only point solutions focused on one subset of the information such as the predictive maintenance analysis tools.
The challenge is to use reliability models integrated with data collection devices in such a manner as to be able to predict future performance of equipment. Next generation reliability is thus the prediction of equipment and system performance from data collection through the use of reliability models.
Steve Shores
Vice President, Asset Management Solutions
The DEI Group
The following excerpt from a paper by Marton Dundics and Ken Bevers, highlights an important area of Key Performance Indicators and using data from multiple sources ot predict future behavior.
Driving a corporation through only a rear-view mirror, receiving only weekly or monthly reports on historical (lagging) business performance indicators along with reactive analysis of production asset performance, and expecting to meet forward looking strategic objectives, leaves you exposed to uncertainty. This is even more difficult when everyone has a different rear-view mirror, turned to focus on their specific narrow activity, and not on their area’s tactical and strategic impact to the enterprise. Even the additional use of leading indicators, associated with internal and external process data, limit the range of accuracy and certainty of information required for time sensitive tactical and strategic decisions which impact future profitability.
To meet corporate financial objectives, minimize the risks of unexpected losses in fulfilling commitments, and take advantage of unplanned opportunities, managers at all levels need to have real-time access to analyzed information (knowledge) providing the “What-if” analysis relative to asset management decisions, that impact achieving future business objectives. This will allow not only budgeting, but dynamically predicting future cost, price, and asset availability, within an expected market demand environment, with a high degree of certainty
Traditionally, corporate and departmental performance has been measured based on metrics associated with a previous period’s actions (month, quarter, and year). To this end, the metric lagged behind the activity. Lost opportunity during an unexpected high demand period, cost/unit product produced, total output generated, employee accidents per month, are examples of lagging indicators. In order to provide an ability to estimate the economic resultant of currently measurable activities, relationships have been defined between certain historical information and expected (future) outcomes. The currently-used leading indicators provide some level of predictability relative to some of the defined key business performance indicators driven by process variables. For example, an increase in safety training should lead to reduced accidents; a measured improvement in the accuracy of the planning process should lead to reduced mean time to repair critical machinery.
Once validated, these leading indicators also define the causal relationships that allow management to adjust processes and make logistic decisions relative to meeting tactical and strategic objectives. Experience has shown that the set of correctly-defined lagging and leading metrics which provide information at all levels of corporate, operations, and maintenance, have both generic and unique components depending on the industry, organizational structure, and internal processes.
However, in order to manage current decisions that impact the future risks associated with operating a manufacturing or processing plant, the predictability of plant machinery availability, first pass yield, process throughput, and cost per unit output with an acceptable confidence level is paramount. The existing lagging and leading indicators by themselves do not provide this predictability with the accuracy and fidelity required for managing in such a narrow earnings environment. Therefore, a more robust, new set of asset-related predictive indicators, based on the analyzed interrelationship of machinery performance information, logistic support process information, and key business performance indicators must be deployed.
The good news is that corporations have invested heavily in plant process control upgrades, machinery data collection tools and monitoring systems, and data intensive logistic support systems such as Operational Process Data Historians and Enterprise Asset Management (EAM) Systems. A great deal of historical performance and activity-related data is being collected today by these systems, but seldom is crisp predictive knowledge being extracted and presented in an explanatory and advisory format. This is due to the fact that:
• The EAM systems, as the backbone of asset life cycle support process, themselves do not provide the functionality required for the plant manager to develop the optimal Operation & Maintenance strategy relative to planned activities, MRO materials, and data structures applicable for his business objectives
• The extent of the available operations and maintenance data has become large and it is scattered in disparate data bases and systems in incompatible formats
• Manpower reallocations have reduced the amount of time available for both data collection, sorting, and desk top analysis
• Lack of confidence in the data due to timeliness and accuracy has created a defensive view toward performance exposure
• The apparent complexity of developing knowledge from such distributed set of data has been culturally prohibitive
As a result, plant machinery life cycle models that provide the predictive indicators essential to forecasting of manufacturing and processing business performance, have not been deployed.
Manufacturing and Processing organizations must strategically invest in Business Performance Forecasting (BPF) systems which will provide predictive performance indicators related to strategic objectives, derived from relevant lagging and leading key performance indicators, analyzed through a risk based generation asset simulation. These systems will enable everyone to drive a Strategy-Focused Organization, concurrently using the same “window” and “rear view mirrors”, thereby achieving strategic objectives within a managed risk environment.
An overview of reliability centered maintenance (RCM) plus a close-up look at six variations and derivations will be covered in a day-long track at the Maintenance & Reliability Technology Summit (MARTS). Everything you need to know about RCM? Not quite everything, but enough to understand the features of six highlighted RCM or RCM-linked methodologies and possibly pick one that is right for you.
The maintenance and reliability educational event, which includes technical exhibits and preconference and postconference workshops, is scheduled May 24-27, 2004 at the Donald E. Stephens Convention Center in Rosemont (Chicago), IL.
The RCM methodology, which started when airlines began considering purchases of jumbo jets, has been used in various forms in the determination of maintenance requirements for civilian and military aircraft, nuclear ballistic missile submarines, electric power plants, and production systems and equipment in a variety of industries. MARTS will provide an opportunity to learn the history of RCM from Jack Nicholas, Jr. and explore the features of six different application approaches as explained by their developers, as follows:
Reliability Centered Maintenance, Its Variations and Derivations
Learn from Jack Nicholas, Jr., PE, CMRP, CEO, Maintenance Quality Systems, neutral but experienced observer who is no longer engaged in RCM analysis, how RCM got started and grew. Major variations and derivations will be outlined, setting the stage for detailed presentations by practitioners actively engaged in using RCM-linked methodologies to help organizations improve maintenance and reliability activity.
RCM Blitz
Learn from Douglas J. Plucknette, president, Reliability Solutions, Inc., how the RCM Blitz method differs from other traditional and streamlined methods and how it works to improve manufacturing reliability while reducing the risk of environmental health and safety incidents.
RCM Turbo
Learn from Kim Davidson, Strategic Corp., about structured and easily deployable approach to RCM that includes expert criticality assessment, optimized frequencies, workflow smoothing, grouped workflow optimization, and extensive failure mode libraries.
RCM Cost
Learn from Bill Keeter, ARMS Reliability Analysis, how to integrate information about failure mechanisms, maintenance costs, and operational costs of failures into the maintenance strategy decision process. The session will focus on the determination of strategies that optimize the maintenance task interval based on the Weibull characteristics of equipment failures.
Classical RCM
Learn from Glenn R. Hinchcliffe, PE, G&S Associates and co-author of RCM: Gateway to World Class Maintenance, why Classical RCM and Abbreviated Classical RCM are the right tools to establish your maintenance department as a corporate center for profit and, when combined with the RCM WorkSaver software, are fast, efficient, and cost effective.
RCM 2
Learn from Marius Basson, senior consultant, New Dimension Solutions, about RCM2, a process used to decide what must be done to ensure that any physical asset, system, or process continues to do whatever its users want it to do. Attendees will learn about how RCM2 extends the power of earlier RCM methodologies.
PM Optimization
Learn from Larry Johnson, Fractal Solutions, the basic principles and techniques of task-based analysis and when it should be applied. This includes a walk-through analysis to demonstrate how to perform all steps in the process. This method can be used to optimize an existing PM program before or after conducting RCM.
Multiple track program
In addition to the seven-part RCM session, the MARTS Reliability track will cover process reliability, risk analysis, and root cause analysis. Other tracks to focus on maintenance management, experience case studies, and technical and commercial innovations.
MARTS is being produced by MAINTENANCE TECHNOLOGY Magazine and Reliabilityweb.com. Further information about workshops, other conference tracks, and registration can be found at www.MARTS-2004.com
Successful programs require the combination of proper training, vibration and lubrication analysis programs, and good diagnostic and practical skills.
By John C. Robertson, AMSCO, Inc.
Originally published in the Jan/Feb 04 issue of Lubrication & Fluid Power Magazine
Click here for a printable 28k pdf version
Although maintenance departments have undergone numerous changes in the past 10 years, in many cases those changes have not led to significant improvements. Many plants have rushed into using the latest technologies at enormous expense, expecting immediate returns on the investments. This has not happened. Why have these technologies not produced what they were supposed to do?
Basically, the technology that is currently available is superb and, if used correctly, will contribute enormously to the reliable operation and maintenance of machines. Unfortunately, the transition from the analyses to the hands-on part of the job leaves much to be desired. This is due in part to the lack of skilled journeymen in the workforce.
Invest in training
As companies focus on equipment reliability, capacity, availability, and plant maintenance as means of improving performance and productivity, maintenance personnel training is often overlooked. Knowledge is expected to be handed down from one person to another, as was the case when a journeyman had an apprentice.
Often, a trainee (apprentice) is shared between two or more journeymen. Unfortunately, this exposes the trainee to bad habits as he or she works with different people who have different ways of working and different skills level and knowledge.
Training must be rigorously controlled to ensure that a trainee is assigned to the best journeyman for at least one year, during which time he or she should have minimal exposure to other journeymen. Bad habits are difficult to break and can lead to poor maintenance with costly consequences.
Get to know the equipment
To benefit from new technologies, there must be an understanding of how machines and their components work. Without this knowledge, the analyst/technician has no direction and the fault diagnosis may be inaccurate. The technician should be able to identify each component in a machine, explain its function, and mate the component to its signature in a vibration spectrum. If an analyst reports a shaft misalignment but cannot demonstrate how to physically correct it, his credibility is lost.
Ideally, an analyst should also be a technician who has hands-on experience in operating and maintaining machines. Unfortunately, industry now has too many certified analysts who can pass tests but are inexperienced in the art of machinery maintenance.
The Vibration Institute, Willowbrook, IL, has stated that 60 percent of those taking Level l Certification pass the examination. The failure rate is directly attributed to a lack of practical experience in the field. This situation can improve only if companies realize that practical skills training must be considered as the first step in preparing technicians and analysts to fully understand machinery basics.
The evolution of maintenance
Maintenance has evolved over the years from a pure reactive mode to preventive maintenance to a proactive mode. Reactive maintenance was costly because there was no control over breakdowns, large inventories of spares had to be maintained, and the cost of call-outs was expensive.
Preventive maintenance was better than reactive maintenance because it expected failures were based on time intervals. It was still costly for some of the same reasons. A large inventory of spare parts was still required, and materials and components were being changed out when there was still useful life left in them.
Preventive maintenance is still complementary to predictive or proactive maintenance because oil and air filters still have to be changed and oil and water levels still have to be maintained in order to keep machines running well.
Blend old skills with new technologies
There are a number of predictive maintenance technologies including vibration analysis, infrared thermography, ultrasonics, and lube oil analysis.
Because most machines are part of an overall mechanical system, vibration analysis is the most common diagnostic tool and is a major component of precision maintenance programs.
But vibration analysis does not supply all the information that is necessary for implementing a precision maintenance program. As a minimum, a lube oil analysis program should support the vibration analysis program. A certified laboratory can analyze samples of lube oil and deliver a report that accurately shows which materials in bearings or in the oil are degrading and the rate of deterioration.
If lube oil is sampled on a monthly basis, the lab report will provide a database for the machine over a six-month period showing the last five or six analyses taken. A good analyst/technician should be able to take that information and compare it against a vibration spectrum obtained from the same bearing or machine. When the results are plotted together, the plots will run alongside each other like railroad tracks, and as the condition deteriorates, both plots will signal an alert or alarm at almost the same time.
However, if the analyst/technician does not know the composition of Babbitt metal that is used on sleeve bearings, for example, then the high lead, tin, and copper levels will not mean much and consequently will not prevent the failure of the bearing. Likewise, if the viscosity and flash point of the oil relationship is not fully understood, the bearing will fail.
In the case of a diesel engine, this ignorance can lead to a crankcase explosion that could injure or kill someone. Clearly, the analyst/technician must have some theoretical skills beyond basic practical skills.
Building a good lubrication program
The three lubrication requirements that a good program must have in place are:
• The lubricant must not be exposed to contaminants. Most component failures can be directly attributed to this avoidable problem.
• Lubricants must perform despite the presence of some contaminants.
• Lubricant consumption should be minimized to prevent spoilage, promote safe working conditions, ease the burden of spent lubricant disposal, and promote longer antifriction bearing life.
Lubricants should reduce energy consumption whenever possible. The analyst/technician has the resources available to determine the correct amount of lubricant to correctly grease antifriction bearings.
Bearings can be correctly greased using a vibration analyzer, an ultrasonic leak detector, or by simply ensuring that the bearing’s cavity drain plug is removed during the greasing and is replaced approximately 30 min after the grease insertion has stopped. Overgreasing motor bearings is responsible for approximately 80 percent of motor burnouts.
The analyst as doctor
When diagnosing problems, it is also helpful to know how equipment works. Pumps provide a clear example. One common pumping problem that can be confusing is cavitation—or is it recirculation? These are two very distinct problems, but both share the same characteristic noise and destructive internal erosion that can tear impellers and casings apart.
Cavitation occurs in the low-pressure suction side of the pump and does not show vane pass frequencies on the vibration analysis spectrum. Cavitation happens when the fluid is vaporized in the suction line and low-pressure side of the impeller.
This phenomenon occurs when there is insufficient net positive suction pressure, or the saturation temperature of the fluid is higher than its corresponding pressure. As a result, vapor pockets form in the fluid. As the vapor pockets reach the surface of the impeller or casing, the higher fluid pressure causes the pockets to collapse, which creates noise, vibration, and possible structural damage to the pump (specifically on the suction side of the impeller and casing).
Recirculation occurs when the output of a pump is drastically changed by throttling the discharge valve, increasing the horsepower of the motor without redesigning the size of the discharge pipeline, or by some other discharge restrictions. As the fluid exits the impeller, the velocity is reduced and, as a result, the fluid no longer passes smoothly into the volute and discharge piping. The fluid now tends to impinge on the cutwater and induces a vibration at a vibration frequency equal to the vane pass (number of vanes x rpm). It also shows large amplitude readings at vane pass frequency.
The analyst as diplomat
When shaft misalignment is diagnosed, can the analyst/technician offer a corrective solution? In many cases, there is open warfare between the millwright and the analyst with each claiming the other person is wrong.
In reality, both could argue they are right, especially when the millwright shows his cold alignment calculations. But if the millwright did not take into consideration allowances for soft foot, shaft runout, bracket sag, pump-to-piping nozzle distortion, and thermal rise, then the vibration analysis diagnosis is correct.
In addition to being skilled in the art of proactive maintenance, the analyst/technician also must be well versed in diplomacy, especially when that person does not know how to physically correct the problems diagnosed. Under such circumstances, an analyst should take the time to go out on jobs to observe the repair work. Often, when it is shown that there is an interest in finding out how repairs are carried out, a better relationship will develop.
The analyst as detective
Once a problem has been identified, the analyst/technician may feel his contribution to success is over. If the identified problem was related to a failed antifriction bearing, the analyst would be wise to observe the bearing removal and to keep the part for inspection. When bearings are discarded, valuable information is thrown away. This information could reveal whether the bearing suffered damage during installation.
Was it the correct bearing for the job? Did the bearing show signs of lack of lubrication or too much lubrication? Was the bearing’s designed physical characteristic correct? These are some of the questions that must be asked when a part fails, and it is the responsibility of the proactive maintenance group to conduct such questioning.
The installation of the replacement bearing also should be witnessed and the bearing checked with vibration analysis after it is back in service. These are some of the basic steps that put precision maintenance ahead of proactive maintenance. For the education of the analyst/technician, the failed bearing should serve to verify the findings of amplitudes and frequencies in the vibration spectrum that characterize antifriction bearing failures.
Practical skills
Taking analysis a step further brings the analyst/technician into other areas where practical skills are needed to help resolve problems. Simple V-belt installations are an example. Belts may be mutilated when they are being installed because of laziness, ignorance, or both.
Instead of detensioning the system by adjusting the motor’s position to slacken the belts, V-belts are pried from the sheaves with a pry bar, and the new belts are installed using the same method. The inner fibers of the new belts are torn apart and rendered useless from the start. The sidewalls of the sheaves can be easily chipped when belts are forced with pry bars, causing unbalance of the sheave.
Sheave misalignment is common, and some people believe that the flexibility of the V-belts will compensate for this misalignment. This is a big mistake. To determine these faults, the analyst/technician must be aware that three spectra have to be identified in order to do a proper analysis: driver frequency, belt frequency, and driven unit frequency.
Each of these components has its own characteristic harmonics, which point to potential problems in a specific system. When these three fundamental frequencies are identified and broken down, belt problems can be easily resolved, depending on the experience level of the analyst/technician. But if they are not recognized, serious problems will arise and cause considerable damage to the machine.
As soon as belts have been installed and properly tensioned, a straight, thin white line should be painted across the width of the belts. This line will serve as an indicator if belt slippage is taking place when the belts are turning and will be highly visible under a correctly tuned strobe light.
If a strobe light does not come with the vibration data logger or analyzer, the analyst/technician would be wise to recommend buying one. This tool is indispensable for troubleshooting rotating equipment, especially fans and drive belts.
Knowledge and experience are keys
An effective analyst must take the time to master the intricacies of machine design and operation, and couple this with a solid foundation in hands-on maintenance at the shop floor level. He must be prepared to continually ask questions from experienced people and build on those experiences until he is competent.
Precision maintenance is a common-sense approach to avoiding and resolving problems before they become disasters. The analyst/technician must have in-depth field experience in all aspects of machinery operation and maintenance. If a disastrous breakdown occurs, the analyst/technician will be called upon to provide information in a root cause failure analysis investigation.
Based on the quality of data and actions taken up to the point of failure, the proactive maintenance program will be justified or condemned. Without experienced people running the program, such a scenario can be disastrous.
John C. Robertson is a maintenance reliability specialist and president of AMSCO Inc., 105 Goldenrain Way, Simpsonville, SC 29680
Ever since the dot com bomb, some companies have been content to simply allow their corporate Web sites to serve as a collection of electronic brochures.
Unfortunately for the companies that have not converted to value-based information delivery, the meltdown of hyped up, over-inflated Internet stock prices did nothing to slow the real benefits of connecting computers and, more importantly, people from around the world.
The Internet forecasts made in 1999 have now been borne out as e-commerce has expanded to well over $100 billion in 2003 (a year behind the prediction). The number of computers with Internet access in 2003 actually exceeded 1999 predictions with 633 million people now online.
Nearly half of U. S. Internet users have built Web pages, posted photos, written comments, or otherwise added to the enormous variety of material available online, according to a report released last month by the Pew Internet and American Life Project.
As some companies squander the opportunity to create online relationships based on value, the Internet continues to fulfill its promise of changing the way information is distributed. Information flows like water around “roadblocks” and those companies that are not in the “information flow” simply get navigated around as Web surfers find their way to more helpful resources.
Sites deliver value
That said, creating and maintaining a useful Web site is a big job. Reliabilityweb.com announces the Top 100 list of maintenance and reliability Web sites each year as a way of delivering value to readers and as a way of acknowledging the extra work that these companies put into creating Web sites that contribute to the overall maintenance and reliability community.
We hope to encourage other companies that publish Web sites to follow these fine examples as well as ask current Top 100 site publishers to continue adding even more value to their Web sites on a consistent basis. When it comes to delivering maintenance and reliability information, there is still a long way to go.
The following criteria are evaluated before a site is selected for the Reliabilityweb.com Top 100 list:
• Web site must be nominated.
• Web site must offer valuable maintenance and/or reliability information.
• Accessing this information must be fee free.
The Reliabilityweb.com Top 100 Web sites are ranked by link popularity, as value exists only in the mind of the visitor. The list is generally assembled in the order we receive the nominations and they are included on our list after being reviewed. The order changes as more people click the online resources they find most useful.
Increase in number of sites
Things are definitely progressing from 2000 when we had a difficult time finding 50 Web sites that met the criteria for the list. It is heartening that we received more than 200 new nominations and many sites had some positive information benefit for visitors beyond product or service information.
The Top 100 listings happen without any communication with the Web site owners or operators and there is no obligation on anyone’s part to take any further action. Some companies choose to display a Top 100 logo and others simply continue with their good work.
A complete list of the Top 100 maintenance and reliability Web sites is available online at www.reliabilityweb.com/forms/rw100_list.htm, including a link to nominate your favorite Web sites for next year’s list. You can even nominate your own site if you think it meets the criteria.
Yahoo, Google, and other search engines are fine, but lists like the Reliabilityweb.com Top 100 can provide a more focused starting point that will speed your way through the search engine clutter to more applicable solutions.
Please help us recognize your favorite maintenance and reliability Web sites by nominating them for our next Top 100 list.
This is a headline right out of USA Today on March 2, 2004.
"poor maintenance is now the second most likely cause of crashes - about 30% from 1995 to 2001 - supplanting other causes such as weather and mechanical failure." according to the National Transportation Safety Board.
"Improvements in cockpit technology and pilot training have caused a dramatic decline in crashes caused by pilots. But far less has been done to modernize maintenance." That about says it all
Terrence O'Hanlon
Reliabilityweb.com