One of the hot topics in electrical and mechanical training classes is the National Fire Protection Association (NFPA) 70E. Students question what 70E is and how it relates to the National Electrical Code (NEC), if 70E is a new regulation and if not why are they just now hearing about it, and if companies are required to comply with 70E.

This article will take some of the mystery out of 70E.

Troubleshooting live equipment, such as testing a contactor (left), requires hazard/risk level 2 PPE, suitable for protection from an arc flash of 8 cal/cm2, but racking of a circuit breaker (right) demands hazard/risk level 3 PPE, suitable for protection from an arc flash of 25 cal/cm2.

With the passing of the Williams-Steiger Occupational Safety and Health Act of 1970 came the need for occupational safety and health regulations. Congress directed OSHA to develop new regulations using existing “national consensus standards” and established federal standards.

For electrical safety regulations it originally adopted the most widely accepted electrical standard in the world—the NEC (National Fire Protection Association’s Standard NFPA 70). However, OSHA encountered several problems in attempting to use the latest editions of the NEC:

• With each new update of the NEC (which occurs every 3 years) OSHA had to go through the extensive legal process of adopting the new NEC edition and risk creating potential conflicts between the adopted version and the published version.
• OSHA needed a regulation that addressed installation, operation, maintenance, and repair in the workplace. The NEC is an electrical installation standard only.
• Because the purpose of the NEC is the practical safeguarding of persons and equipment and because it includes provisions for residential, it contains many provisions that are not relevant to OSHA and only confuse the reader.

To correct these problems and others, NFPA created a committee to develop electrical safety standards that would serve the needs of OSHA. This committee reports through the NEC technical committee and is called the Committee on Electrical Safety Requirements for Employee Workplaces—NFPA 70E. This standard has evolved over time:

• 1979: First edition published with only Part I (Installation Safety Requirements).
• 1981: Second edition added Part II (Safety-Related Work Practices).
• 1983: Third edition added Part III (Safety-Related Maintenance Requirements).
• 1988: Fourth edition had only minor revisions.
• 1995: Fifth edition updated Part I based on the most recent NEC and made some major additions to Part II.
• 2000: Sixth edition updated Part I based on the most recent NEC, made additions to Part II, and added Part IV (Safety Requirements for Special Equipment).
• 2004: The most recent edition made many significant changes including a total reorganization into the NEC format. In the reorganization Part II was moved to become Chapter 1, Part III became Chapter 2, Part IV became Chapter 3, and Part I became Chapter 4.

Is 70E a “national consensus standard”?
By definition NFPA 70E is a national consensus standard. In 29 CFR 1910.2(g), a national consensus standard is defined as a standard that is developed by the same persons it affects and then is adopted by a nationally recognized organization.

Organizations that publish national consensus standards include the NFPA, American Society for Testing and Materials (ASTM), and the American National Standards Institute (ANSI).

What does it cover?
In NFPA’s catalog it states: “70E covers the full range of electrical safety issues from safety-related work practices to maintenance, special equipment requirements, and installation. In fact, OSHA bases its electrical safety mandates—OSHA 1910 Subpart S and OSHA 1926 Subpart K—on the comprehensive information in this important Standard.”

The 2004 edition of 70E has an introduction, four chapters, and 13 annexes.

Chapter 1, “Safety-Related Work Practices,” is the meat of the 70E document. It discusses qualified vs unqualified persons and training. It requires an electrical safety program, electrical hazard analysis for shock and arc flash, energized electrical work permits, and lockout/tagout procedures. It establishes approach boundaries and discusses how to select appropriate personal protective equipment (PPE) and protective clothing. Arc flash protection also is addressed in this chapter.

Chapter 2, “Safety-Related Maintenance Requirements,” does not create much discussion. It basically requires that electrical components, wiring, and equipment be maintained in a safe condition.

Chapter 3, “Safety Requirements for Special Equipment,” covers batteries, lasers, and power electronic equipment. This chapter affects more installations than one might initially think because power electronic equipment includes electric arc welding equipment, and motor drives, UPS, and lighting controllers that contain rectifiers and inverters. There are no surprises in this chapter but those with the subject equipment should review it.

Chapter 4, “Installation Safety Requirements,” is a truncated version of the NEC. Here authors state that the requirements in Chapter 4 are based on the NEC and in the forward of the 70E document it states that this document is not intended to be used in lieu of the NEC.

Annexes A through M offer useful information including how to calculate flash protection boundaries.

What is the “general duty clause” and how does it relate to compliance?
This clause refers to a portion of the Occupational Safety and Health Act of 1970:

5. Duties
(a) Each employer
(1) shall furnish to each of his employees employment and a place of employment which are free from recognized hazards that are causing or are likely to cause death or serious physical harm to his employees,

Section 5(a)(1) has become known as the “general duty clause.” It is a catch-all for citations if OSHA identifies unsafe conditions to which a regulation does not exist.

The NEC requires field labeling (above) on equipment where arc flash is a hazard. A future edition of the code may require more extensive labeling (inset) that includes flash hazard boundary and PPE levels.

Is compliance mandatory?
In 2002, the NEC referenced NFPA 70E for the first time.

NFPA 70-NEC Section 110.16 Flash Protection requires field labeling of switchboards, panelboards, industrial control panels, and motor control centers that are likely to require examination, adjustment, servicing, or maintenance while energized to warn the qualified person of the potential of an arc flash. In Fine Print Note No. 1 that follows 110.16 it refers the reader to NFPA 70E for assistance in determining severity of potential exposure, planning safe work practices, and selecting personal protective equipment.

It is possible and in fact likely that the 2005 NEC may strengthen the language in 110.16 to require specific information on the field labels such as flash boundaries and PPE requirements, which are addressed in 70E. If this happens, facilities complying with the 2005 NEC will need flash hazard analyses completed for all new equipment or will need to default to generic tables provided in 70E to determine the boundaries and PPE requirements.

OSHA regulation 29 CFR 1910 Subpart S Appendix A: Reference Documents also references NFPA 70E:

“The following references provide information which can be helpful in understanding and complying with the requirements contained in Subpart S:

NFPA 70-78 National Electrical Code

NFPA 70E Standard for the Electrical Safety Requirements for Employee Workplaces”

In a “Standards Interpretation” letter from OSHA in 2003 the following is from selected paragraphs:

“All your questions involve the NFPA 70E standard, which is one of many industry consensus standards developed by the National Fire Protection Association. NFPA 70E, which is titled ‘Electrical Safety Requirements for Employee Workplaces,’ is the NFPA’s consensus standard for workplace electrical safety. It covers employee protection from electrical hazards including shock, arc blasts, explosions initiated by electricity, outside conductors, etc.

“With respect to the General Duty Clause, industry consensus standards may be evidence that a hazard is ‘recognized’ and that there is a feasible means of correcting such a hazard.

“These provisions (1910.132(a) personal protective equipment) are written in general terms, requiring, for example, that personal protective equipment be provided ‘where necessary by reason of hazards…’ and requiring the employer to select equipment ‘that will protect the affected employee from the hazards…’.

“Industry consensus standards, such as NFPA 70E, can be used by employers as guides to making the assessments and equipment selections required by the standard. Similarly, in OSHA enforcement actions, they (70E) can be used as evidence of whether the employer acted reasonably.

“Under 1910.135, the employer must ensure that affected employees wear a protective helmet that meets either the applicable ANSI Z89.1 standard or a helmet that the employer demonstrates ‘to be equally effective’. If an employer demonstrated that NFPA 70E contains criteria for protective helmets regarding protection against falling objects and electrical shock that is equal to or more stringent than the applicable ANSI standard, and a helmet met the NFPA 70E criteria, the employer could use that to demonstrate that the helmet is ‘equally effective’.”

In September 1999 a major U. S. corporation experienced an electrical accident that resulted in serious burn injuries to an electrical apprentice employee. OSHA investigated the accident and issued a number of citations. The employer challenged the citations and the disagreement ended up before the Occupational Safety and Health Review Commission.

As part of the citation OSHA contended that the employer violated a federal regulation because it did not provide or require that its electricians wear appropriate flame-resistant or retardant personal protection, specifically, flame-resistant coveralls and insulated gloves. OSHA also contended that the employer violated a regulation when it did not provide or require that its electricians wear appropriate face protection.

In the settlement the employer agreed to develop hazard analyses in accordance with the personal protective equipment provisions contained in NFPA 70E. OSHA agreed that given the present state of its standards and regulations, the hazard analyses would achieve compliance with its requirements.

Points to remember
To summarize, you should understand:

• Several of the OSHA regulations are written in general terms leaving the details up to the employer on how to comply. (An example is requirements for personal protective equipment and clothing in 1910.132(a).) The employer is expected to use consensus standards to help in the selection of the best method to achieve compliance with OSHA regulations. NFPA 70E is a “how to comply” standard for specific OSHA regulations.

• Although NFPA developed 70E for OSHA, OSHA has not officially adopted or incorporated it by reference into its regulations. Instead in 1990, OSHA promulgated new safety-related work practices in 1910.331 based on the information in 70E at that time. However, NFPA has made major changes to 70E based on better information and research since OSHA developed its standard. The bottom-line is that 70E is not a federal regulation; it is just a national consensus standard like hundreds of other standards that are not laws or regulations. But compliance with 70E will assure compliance with specific OSHA electrical regulations.

• Some OSHA state plans are more restrictive than federal OSHA and as such may have adopted or incorporated 70E; however, this is on a state-by-state basis and should be evaluated by employer location. After researching several states on this issue, the responses were too varied to incorporate into this article.

• In the event of an injury or death due to an electrical accident, if OSHA determines that compliance with 70E would have prevented or lessened the injury, OSHA may cite the employer under the “general duty clause” for not using 70E to protect the employee(s). (Shock and arc flash are recognized hazards that employers should be aware of because 70E is now referenced in both the NEC and OSHA regulations.)

• It is important to get training on NFPA 70E and to implement it into your electrical safety program. MT

John C. Klingler, P.E., is vice president–site specific training and an instructor for Lewellyn Technology, Inc., P. O. Box 618, Linton, IN 47441; telephone (812) 847-3525

Much has been written about lean manufacturing and the lean enterprise—enough that nearly all readers are familiar with the concepts as well as the phrases themselves. But what about lean maintenance?

Is it merely a subset of lean manufacturing? Is it a natural fall-in-behind spinoff result of adopting lean manufacturing practices? Much to the chagrin of many manufacturing companies, whose attempts at implementing lean practices have failed ignominiously, lean maintenance is neither a subset nor a spinoff of lean manufacturing. It is instead a prerequisite for success as a lean manufacturer. This article will explain why.

The definition
The best starting point is to define lean maintenance:

Lean maintenance is a proactive maintenance operation employing planned and scheduled maintenance activities through total productive maintenance (TPM) practices using maintenance strategies developed through application of reliability centered maintenance (RCM) decision logic and practiced by empowered (self-directed) action teams using the 5S process, weekly Kaizen improvement events, and autonomous maintenance together with multi-skilled, maintenance technician-performed maintenance through the committed use of their work order system and their computer managed maintenance system (CMMS) or enterprise asset management (EAM) system. They are supported by a distributed, lean maintenance/MRO storeroom that provides parts and materials on a just-in-time (JIT) basis and backed by a maintenance and reliability engineering group that performs root cause failure analysis (RCFA), failed part analysis, maintenance procedure effectiveness analysis, predictive maintenance (PdM) analysis, and trending and analysis of condition monitoring results.

That is lean maintenance in a nutshell, albeit a rather large nut (except for a few details that were omitted here but will be covered later in the article). Let’s discuss the highpoints of this definition to be sure everyone understands the terms used:

• Proactive. This is the opposite of reactive where the maintenance operation reacts to equipment failures by performing repairs. In the proactive maintenance operation the prevention of equipment failures through performance of preventive and predictive maintenance actions is the objective. Repair is not equivalent to maintenance.

• Planned and scheduled. Planned maintenance involves the use of documented maintenance tasks that identify task action steps, labor resource requirements, parts and materials requirements, time to perform, and technical references. Scheduled maintenance is the prioritization of the work, issuance of a work order, assignment of available labor resources, designation of the time period to perform the task (coordinated with operations/production), and breakout and staging of parts and materials.

• Total productive maintenance. TPM is the foundation of lean maintenance. It is an initiative for optimizing the reliability and effectiveness of manufacturing equipment. TPM is team-based, proactive maintenance and involves every level and function in the organization, from top executives to the shop floor. TPM addresses the entire production system life cycle and builds a solid, shop floor-based system to prevent all losses. TPM objectives include the elimination of all accidents, defects, and breakdowns.

• Reliability centered maintenance. RCM is a process used to determine the maintenance requirements of physical assets in their present operating context. While TPM objectives focus on maintaining equipment reliability and effectiveness, RCM focuses on optimizing maintenance effectiveness.

• Empowered (self-directed) action teams. Action team activities are task-oriented and designed with a strong performance focus. The team is organized to perform whole and integrated tasks, hence requiring multi-department membership. The team should have defined autonomy (that is, control over many of its own administrative functions such as self-evaluation and self-regulation—all with limits defined). Furthermore, members should participate in the selection of new team members. Multiple skills are valued. This encourages people to adapt to planned changes or occurrence of unanticipated events.

• 5S process. There are five activities for improving the work place environment: sort (remove unnecessary items), straighten (organize), scrub (clean everything), standardize (standard routine to sort, straighten, and scrub), and spread (expand the process to other areas).

• Kaizen improvement events. Kaizen is the philosophy of continuous improvement, that every process can and should be continually evaluated and improved in terms of time required, resources used, resultant quality, and other aspects relevant to the process. These events are often referred to as a Kaizen blitz—a fast turnaround (1 week or less) application of Kaizen improvement tools to realize quick results.

• Autonomous maintenance. This refers to routine maintenance (e.g., equipment cleaning, lubrication, etc.) performed by the production line operator. The maintenance manager and production manager will need to agree on and establish policy for where in the production processes autonomous maintenance will be performed, what level and types of maintenance the operators will perform, and how the work process for autonomous maintenance will flow. Specific training in the performance of designated maintenance responsibilities must be provided to the operators prior to assigning them autonomous maintenance responsibilities.

• Multi-skilled, maintenance technician. Multi-skilled maintenance technicians are becoming more valuable in modern manufacturing plants employing PLCs, PC-based equipment and process control, automated testing, remote process monitoring and control, and similar modern production systems. Maintenance technicians who can test and operate these systems as well as make mechanical and electrical adjustments, calibrations, and parts replacement obviate the need for multiple crafts in many maintenance tasks. The plant processes should determine the need for and advantages of including multiple skills training in the overall training plan.

• Work order system. This system is used to plan, assign, and schedule all maintenance work and to acquire equipment performance and reliability data for development of equipment histories. The work order is the backbone of a proactive maintenance organization’s work execution, information input, and feedback from the CMMS. All work must be captured on a work order—8 hours on the job equals 8 hours on work orders. The types of work orders will include categories such as planned/scheduled, corrective, emergency, etc. The work order will be the primary tool for managing labor resources and measuring department effectiveness.

• Computer managed maintenance system. The information (maintenance) management software system performs, as a minimum, work order management, planning function, scheduling function, equipment history accumulation, budget/cost function, labor resource management, spares management, and a reports function that utilizes key performance indicators (KPI). To be effective, the CMMS must be fully implemented with complete and accurate equipment data, parts and materials data, and maintenance plans and procedures.

• Enterprise asset management. The EAM system performs the same functions that the CMMS does but on a more organization-wide, integrated basis, incorporating all sites and assets of a corporation. Even broader enterprise systems incorporate fully integrated modules for all the major processes in the entire organization and offer the promise to effectively integrate all the information flows in the organization.

• Distributed, lean maintenance/MRO storeroom. Several stores locations replace the centralized storeroom in order to place area-specific parts and materials closer to their point-of-use. Lean stores employ standardized materials for common application usage. The lean stores operation also employs planning and forecasting techniques to stabilize the purchasing and storeroom management process. This method requires that a long-term equipment plan is developed and equipment bills of material (BOM) are entered into the CMMS as soon as the purchase order for new equipment is issued.

• Parts and materials on a just-in-time basis. Stores inventories are drastically reduced (as are the costs of carrying large inventories) through a strong supply chain management team that uses JIT suppliers, and practices such as vendor-managed inventories in which the vendor is given the responsibility for maintaining good inventory practices in replenishment, in ordering, and in issuing the materials. The vendor is charged with the responsibility of controlling costs and inventory levels, the sharing of information with the facility, and making improvements in the process.

The supply chain management team advocates day-to-day supplier communication and cooperation, free exchange of business and technical information, responsive win-win decision-making, and supplier profit sharing.

• Maintenance and reliability engineering group. Because statistics indicate that up to 70 percent of equipment failures are self-induced, a major responsibility of maintenance engineering involves discovery of the causes of all failures. Reliability engineering is a major responsibility of a maintenance engineering group.

Their responsibilities in this area also include evaluating preventive maintenance action effectiveness, developing PdM techniques/procedures, performing condition monitoring/equipment testing, and employing engineering techniques to extend equipment life, including specifications for new/rebuilt equipment, precision rebuild and installation, failed-part analysis, root cause failure analysis, reliability engineering, rebuild certification/verification, age exploration, and recurrence control.

Other terms
Here are descriptions of some of the terms related to the maintenance and reliability engineering group:

• Root cause failure analysis. One of the most important functions of the maintenance engineering group is RCFA. Failures are seldom planned for and usually surprise both maintenance and production personnel and they nearly always result in lost production. Finding the underlying, or root, cause of a failure provides an organization with a solvable problem, removing the mystery of why equipment failed. Once the root cause is identified, a fix can be developed and implemented.

There are many methods available for performing RCFA, such as the Ishikawa, or Fishbone, diagramming technique; the events and causal factor analysis; change analysis; barrier analysis; management oversight and risk tree (MORT) approach; human performance evaluation; and the Kepner-Tregoe problem-solving and decision-making process.

• Failed part analysis. Examination, testing, and/or analysis by maintenance engineering on failed parts and components, removed from equipment, determines whether the parts were defective or an external influence, such as operating conditions, faulty installation technique or other influence, caused the failure. Physical examination is often required in order to determine where to begin RCFA. For example, when a bearing fails the mode of failure must be determined by examining the bearing,. If electrical erosion/pitting is found, then stray ground currents (the cause of electrical pitting in bearings) must be found and eliminated.

• Procedure effectiveness analysis. Among the responsibilities of maintenance engineering for the establishment and execution of maintenance optimization is the use of CMMS-generated unscheduled and emergency reports and planned/preventive maintenance reports to determine high-cost areas, and establish methodologies for CMMS trending and analysis of all maintenance data to make recommendations for changes to preventive maintenance frequencies, corrective maintenance criteria, and overhaul criteria/frequency. It also must identify the need for the addition or deletion of PMs, establish assessment processes to fine-tune the program, and establish performance standards for each piece of equipment. The maintenance engineering group also establishes adjustment, test, and inspection frequencies based on equipment operating (history) experience.

Additional responsibilities include the optimization of test and inspection methods and the introduction of effective advanced test and inspection methods. Maintenance engineering performs periodic reviews of equipment on the corrective maintenance (CM)/PdM program to delete that equipment no longer requiring CM/PdM, or to add to the CM/PdM program any equipment or other items as appropriate. The maintenance engineering group also communicates problems and possible solutions to involved personnel and controls the direction and cost of the CM/PdM program.

• PdM analysis. A major role of maintenance engineering is optimizing maintenance. One of the most widely used tools in this regard is PdM to forecast necessary maintenance actions. Depending on the quantity and kinds of production equipment in a plant, the array of PdM techniques can range from as few as two or three to as many as 10 or more. Whether a PdM technique is outsourced or performed in-house, the results and recommendations must be analyzed by maintenance engineering and maintenance actions scheduled prior to predicted failure or out-of-specification condition.

• Trending and analysis of condition monitoring. Condition monitoring, actually a subset of predictive maintenance, usually involves the use of installed metrology (gauges, meters, etc.) to derive the equipment’s operating condition. Examples can be as simple as a differential pressure gauge across a filter or the head-flow characteristics of a pump.

Maintenance engineering must establish operating limits for the condition(s) being monitored and trend the observed data, obtained from a log sheet or planned maintenance procedure, to determine when the operating limits will be exceeded so that required maintenance can be performed. This is referred to as condition-based maintenance and can be both more effective and less costly than periodic or fixed frequency maintenance.

Leadership changes
The foregoing provides a good, basic definition of lean maintenance by describing the activities and job responsibilities of those involved in the lean maintenance operation. Lean maintenance is also about fundamental changes in attitudes and leadership roles. In the lean environment the shop floor-level employee is recognized as the company’s most valuable asset. Management and supervisory roles change from that of directing and controlling, to a role of supporting.

The lean maintenance organization is a flat organization with fewer layers of middle management and supervision because, with the establishment of empowered action teams, much of their direction comes from within. The remaining supervisors spend the majority of their time on the shop floor providing technical advice and guidance and identifying first-hand the problems and needs of the action teams.

The foundation elements, in particular TPM, must be in place before an organization can effectively build on the maintenance management pyramid with elements such as autonomous maintenance and before it can sustain continuous improvement.

A company transitioning to lean manufacturing will not have a sound basis of maintenance support without first implementing many of these necessary and fundamental changes in the maintenance operation. As the foundation of lean maintenance, TPM must be operating and effective, as shown by the key performance indicators, prior to launching a plant’s lean manufacturing initiative. MT

Ricky Smith is the executive director of maintenance solutions at Life Cycle Engineering, Inc., 4360 Corporate Rd., Ste. 100, North Charleston, SC 29405-7445; (843) 744-7110

Terry Wireman, C.P.M.M., Editorial Director

KPI–itis (kA–pE–I–Tis): a management disease where everyone in the company is obsessed with trying to find key performance indicators (KPIs) that are right for the their company.

Symptom—asking other companies what they measure, reading books and articles on KPIs, using industry benchmarks as KPIs, and using excessive amounts of KPIs.

Treatment—A personal vaccination of education to personally learn what the company’s business is and how shareholder/owner value is created. Then, the next step is to communicate to each employee how each of the company’s core business processes contribute to shareholder value/profits and how to measure whether or not this is being accomplished. Once this is understood, then the proper KPIs can be designed and utilized to manage the business.

While this fictitious dictionary entry may not seem realistic, most companies have caught this disease or at least have its symptoms. You only have to visit companies, listen to their discussions at conferences and tradeshows, or interview company executives. They all talk about their “symptoms”; how they cannot find the “right” KPIs. If they only read the fictitious medical handbook, they would realize that the cure (finding and then properly utilizing the right performance indicators) is not that complicated.

The process of developing the right KPIs begins with the clear understanding of the business—having a shared vision from the boardroom to the shop floor. Employees must understand the business at a level where they can connect their department to the company’s profitability. If they cannot achieve this understanding, the organization loses focus and the business efforts are inefficient and will consume more resource (and ultimately profits) than they should.

Once the business connections are understood, then process strategies should be developed with the company’s financial goals in view. There is always a corporate financial measure (total cost to produce, total cost of occupancy, etc.) that executives measure. Each of the business processes in the company need to connect to this measure. For example, consider the maintenance business process. Many companies measure maintenance cost as a percentage of the product cost (maintenance cost per unit produced) or of occupancy (maintenance cost per square foot maintained). These indicators are common in many plants. However, here is where a disconnect occurs. They fail to connect the next level of indicators to the company financials.

Financial indicators are impacted by the efficiency and effectiveness of the maintenance organization. If there is waste in maintenance execution, then the costs are higher than necessary. This highlights the need for a level of indicators to monitor the efficiency and effectiveness of the maintenance organization. For example, an indicator such as level of maintenance technicians’ productive time could be utilized. If the productive time is low (less than 50 percent), this indicates an area of waste that needs to be investigated.

Low productivity may be caused by excessive reactive maintenance. In this case another indicator showing the comparison of planned, reactive, preventive, and predictive work activities would be useful. Another possible cause could be lack of spare parts from the stores and purchasing function. An indicator such as stores service level or stock out percentage would be useful.

There are other indicators that may need to be investigated that also impact on the productivity of maintenance technicians. These are tactical indicators that focus on the proper organizational deployment and support. Ratios of technicians to supervisors or technicians to planners should also be utilized. An indicator monitoring response time may show that the maintenance organizational structure is contributing to low productivity—for example, utilizing a centralized deployment model when an area deployment model would be more efficient and effective, thereby raising the maintenance technician’s productivity.

Robert C. Baldwin, CMRP, Editor

The meeting convened in Baton Rouge, LA the evening of September 13, as Hurricane Ivan was moving into the Gulf of Mexico.

The first morning session began with a hurricane update. By the coffee break time there was considerable concern by the members about whether they were going to be able to get home before the storm hit. We decided to dispense with the plant tour and cram as many presentations as possible into a one-day marathon so we could arrange to start for home early the next day.

Even though we would have to reconvene after dinner, we kept a panel discussion on equipment reliability metrics on the agenda. (Why did we keep this discussion and not wrap up earlier? Perhaps comments by Terry Wireman in his Viewpoint editorial on pg 50 will provide some insight.)

Dofasco’s Ed Ray moderated this informative session. He and roundtable members Jay Blosser of DuPont, Richard Shirer of Baxter Healthcare, Don Latiolais of Syngenta, and Jennifer Vicknair and Rich Hall of Honeywell provided talking point slides for discussion.

Ray’s talking points, based on Dofasco’s decade-long enterprise-wide journey of maintenance and reliability improvement, set the scene by noting that business units in his company are responsible for results and they use metrics to manage and communicate. The following slide did a neat job of summing up the issues.

Why Measure?

• To effectively and efficiently manage the equipment reliability process
• To permit on-going demonstration of the business value of equipment reliability
• To ensure broadly communicated and thoroughly understood awareness of current performance
• To drive performance improvement through the creation of appropriate tension for change
• To enable a proactive management environment

Presentation coaches often suggest that you begin with a summary of your conclusion. Ray did that with a quote from management guru Peter Drucker:

“It is not possible to manage what you cannot control and you cannot control what you cannot measure.”

That invites a follow-up question: What are you measuring, and are they the right things? MT

As time went by, others adapted RCM further, some adding elements and other taking them away. At last count there were over a dozen major RCM derivations and many other less popular methods. The only thing some of the processes have in common are three little letters: R, C, and M.

Like many technical subject areas, you do not know what you do not know about RCM and its derivations until it is too late. Luckily the Web is a great place to learn more about the world of RCM.

Let’s start with the archives at MAINTENANCE TECHNOLOGY magazine. Visit www. mt-online.com and click on Articles. There are dozens of archived RCM articles including many by John Moubray.

Reliabilityweb.com also features an RCM knowledge base linked from the home page that includes articles, and streaming tutorials.

Reliability Radio includes an audio interview with Anthony “Mac” Smith, co-author of RCM–Gateway to World Class Maintenance.

You can also visit Aladon, the company founded by John Moubray and the current promoter of RCM-2 methodology . There are several excellent articles and papers that anyone exploring RCM should read.

You can also read some articles by Mac Smith at the JMS Software site including one that covers implementing an RCM program.

The Society of Automotive Engineers has written a rigorous RCM standard that some companies have adopted to ensure a consistent RCM service level. There are advantages and disadvantages of using a standard like SAE-JA1011; however, understanding it should be included in your RCM learning goals. Visit www.sae.org and search for Reliability Centered Maintenance to purchase the standard.

Jack Nicholas Jr. is building on the work of Nowlan and Heap, John Moubray, and Mac Smith with the help of Doug Plucknette, the developer of RCM Blitz, to create an RCM scorecard method for applying consistent metrics for short and long term RCM results.

Jack will deliver a full day workshop at the Reliability Centered Maintenance Managers’ Forum, March 9-11, 2005, in Clearwater, FL and welcomes e-mail input from people who have experience implementing reliability centered maintenance. Specifically Jack is inviting comments that explain:

• How the results of RCM projects are measured during and after completion
• What pitfalls and problems have been encountered when identifying adequate metrics
• What solutions may be beneficial to practitioners, vendors, customers, managers, sponsors, or anyone else interested in the progress and eventual outcome of an RCM project. MT

Fig. 1. Sample of 20th century building trends (sq ft by year).

Maintenance managers are constantly looking for new ways to communicate value propositions to senior managers. Part of the communications gap lies in the fact that senior managers generally look 5-10 years into the future, whereas maintenance managers are primarily focused on the next 1-2 years.

Most organizations go through large surges of capital purchasing, followed by a settling-in period. Looking at a sample of public infrastructure building over the past century, the patterns are clear (Fig. 1).

Renewal spending surge
The post-war building surge has required a certain level of maintenance spending—but all that is about to change. The average age of public assets in North America is 50 years, and the typical renewal lifespan for components in a building is 50-60 years. Thus, over the next 10 years, renewal spending is naturally going to go through the same surge that initial asset spending went through 50 years ago. Just as school requirements follow demographic trends—high birth rates require a large number of schools 5 years later—renewal spending increases a fixed amount of time after asset creation.

The same principles apply to younger organizations, just in different categories. If a company replaced 10 of its boilers 20 years ago, and those boilers have a 25-yr lifespan, the company will face a large renewal spend in 5 years. These clusters of renewal spending also may occur in completely unrelated categories. In 6 years, the same company may be facing the failure of three HVAC systems, four electrical subsystems, nine roofing systems, and a group of exterior windows.

This kind of future spending cluster is beyond the operating radar of most maintenance managers. But knowing about it is crucial for financial managers, who often work in the 5-10 year range, and demand more choices about how to deal with upcoming spikes in renewal spending. With longer expenditure planning profiles, they can choose to decommission facilities, drop or sell product lines, seek interest advantages on future borrowing, and so on. The more warning financial managers have of upcoming spending spikes, the more options they have.

If maintenance managers are able to effectively communicate their company’s long-term capital renewal requirements, they are already speaking the financial manager’s language. If they can warn the organization of approaching spikes in renewal spending, they make the leap from reactive to proactive communicators. The further ahead they can look, the more they will be able to help their organizations avoid dangerous and reactive capital spending by exercising proactive capital renewal options. It almost goes without saying that their value within an organization also will increase and be perceived as more strategically important.

Gathering the data
The process of transforming multi-year renewal planning from an art to a science can be called total capital planning (TCP). Assuming that this kind of predictive ability is worthwhile, how can it be achieved?

Fig. 2. Ten year renewal spending variations.

First, a comprehensive asset inventory must be acquired. Some organizations already do this by sending out auditors on a bi-annual or annual basis to review the existence and condition of their major capital assets. To really be able to bring TCP into play, however, there are a few other pieces of information that need to be gathered, specifically installation date, theoretical life, and replacement cost. (This does not need to be done for every nut and bolt on-site. There are different levels of cutoff for capital spending—one rule of thumb is to exclude any asset which has a replacement cost of less than $10,000.)

Once this information is obtained, managers are in a position to begin exploring future spending requirements. For instance, looking out into the next 10 years, they might see potentially catastrophic spending variations (Fig. 2).

This kind of information is crucial for financial managers. If they cannot see these trends, they will be unable to plan for the staggering differences between the 2006 requirements ($200,000) and the 2009 ($2.4 million). They will be forced to make hasty decisions as renewal requirements spiral upward, without necessarily knowing that those requirements drop by almost three quarters in the next year.

Funding TCP?
There is no doubt that if any financial manager could snap his fingers and have this data, he would do so. However, in the real world, all funding must be justified, and so it is important that managers have a strong understanding of the financial drivers for instituting TCP.

Bulk purchasing. With airlines, the further ahead a traveler can book, the better the price. Similarly, being able to examine renewal requirements 5, 10, or more years out gives managers strong leverage when it comes to negotiating bulk purchase orders with vendors, or interest rates with lenders. This can help cover the additional costs of implementing TCP. To get the most value out of this strategy, however, it is important that organizations categorize their assets by type (Fig. 3). Within limits, the finer the definitions, the more that can be saved.

Fig. 3. Sample asset categories.

Facility assessment deferral. By prepopulating a database with known life cycle data, it is possible to be able to reduce or eliminate on-site assessment time and minimize bottom line costs. A good rule of thumb is to avoid performing inspections on any asset which is more than 5 years away from the end of its modeled, theoretical life. (Some organizations may have entire sites which fall into this category.) When an auditor does go on-site, he or she can ignore assets which have a long way to go before they need replacing. This reduces costly on-site time by bypassing assets until they age further.

Capital savings. If financial managers know years in advance that they are going to need significant additional funds for renewal, they can negotiate lower interest rates, or avoid getting into costly capital projects which will draw money away from upcoming renewal needs. Subject to analysis, it also may be possible to mitigate some renewal needs by decommissioning facilities or converting them to other uses.

Singularly, or in combination, the above strategies should more than cover the costs associated with gathering the additional data for TCP.

Software. Gathering, interpreting, and reporting on this volume of data usually requires a software solution. This raises the costs of implementing TCP, but it also increases the opportunities for savings.

Where applicable, capital renewal events should recycle themselves. For example, pushing events (projected capital requirements) out of a TCP system and into a computerized maintenance management system (CMMS) will ensure that these events can be tracked to completion in a familiar environment.

Once the work orders are completed in the CMMS, the TCP system should recycle them back into the future, ideally using the actual cost. Thus, a boiler replacement might be modeled as costing $200,000, but come in at $225,000. When the event is recycled into the future, the real-world cost should be used to represent the event in a specific year based on its theoretical life. In this way, the more a TCP system is used, the more accurate it becomes. Similarly, events can be pushed to a project management system, and then recycled as the projects are completed.

What to look for in a TCP

Fig. 4. Sample TCP data flows

A number of functions should be present in TCP software:

• Event recycling. When a renewal event is completed, is it re-created? Does the recycled event use real-world costs/theoretical lives?
• Modeling and costing expertise. Does the software create the draft database without requiring on-site visits?
• On-site assessment tool. Will it reduce on-site time and errors?
• The ability to explore spending scenarios. Will it calculate “If I increase renewal spending by 5 percent” for example?
• Project management. Will it allow bundling events into capital spending projects?
• Easy integration into existing systems
• End-user/DBA customization
• New information tracking. Can the software be easily configured to do so?
• Strong security features. Can data integrity be maintained as capital requirements and projections are very sensitive to the organization?

The internal sell
To determine if there is a benefit from TCP, the most important step is to review how much an organization is spending on facility assessments.

This is typically where the greatest and most immediate cost savings can be found. If some or all of the investment in a TCP system can be recouped quickly by reducing these assessment costs or improving capital purchasing practices, the value proposition should be presented to financial executives. Include spending clusters that might affect the organization, as well as a preliminary return on investment on anticipated savings.

As North America’s facilities and infrastructure age, renewal requirements will dramatically increase from historical levels. In 10 years, organizations may well be largely defined by how well they anticipated and responded to this growing need for renewal capital. Putting a TCP process in place before assets end up on life support will greatly increase the chances of being counted among the survivors. MT

Mature maintenance and reliability programs, ones that have the basics covered, look to reliability analysis as a primary path to improved maintenance practice. The analysis efforts are directed toward physical assets and systems whose failure will produce significant negative impact on safety, environmental and regulatory compliance, productivity, or costs.

Reliability centered maintenance (RCM) is accepted as the most effective and comprehensive approach to reliability analysis. Also effective is failure modes and effects analysis (FMEA) which is an element of RCM that can be practiced independently. In both approaches, the equipment or system is analyzed by a cross-functional team of plant personnel according to a prescribed procedure. The team is led by a trained facilitator.

RCM and FMEA produce a lot of data which can be difficult to organize and maintain. There are a number of data and information management software packages that have been developed to streamline the RCM and FMEA data handling chore. The following information includes a listing of 15 such packages, along with suggestions from the software producers on how to get started with an RCM or FMEA program and how to sell it to management.

Getting started with RCM
Maintenance and reliability practitioners and consultants stress the need for appropriate training prior to launching a RCM or FMEA project.

“First time RCM analysts,” notes Anthony “Mac” Smith, RCM consultant, “should use an experienced facilitator to guide them through the initial process. In-house personnel should be involved in making the RCM decisions, not outside contractors. This helps provide buy-in and increases the success and benefits of RCM.”

According to Chris Kelly of Strategic Corporate Assessment Systems, “Practitioners need to have a full understanding of the principles of reliability centered maintenance in order to obtain sustainable outcomes. A maintenance strategy review needs to be undertaken like any other project, i.e., with milestones and expected deliverables. Thereafter, the reliability approach needs to be embedded into daily life for maintenance people.”

How to sell the program
When it comes to selling upper management on the benefits of RCM, JMS Software’s Nick Jize suggests that you should “let management know that the process will shift the paradigms of why maintenance is done from a reactive, time-based approach to a proven reliability and condition-based, less intrusive approach. Knowing that, in-house personnel will perform the analysis and thereby increase their expertise on how plant equipment fails and how to prevent it. Most important, show them the cost and benefits of the process. Too often, management sees only the initial costs and does not see the instant payback. They need to understand the process and see the differences in task selection after each system is completed.”

“Nontechnical management must be convinced that maintenance is a business function and integrally linked to production,” councils Kelly. “We need to impart to management that the maintenance function has far-reaching business impacts which include the preservation and extension of the life of expensive assets. We also need to impart a quantification of the business value associated with an increase in reliability, availability, and production associated with an optimized RCM- based maintenance regime.”

“There is a clear business case for the improved reliability that results from a technically sound work identification strategy like RCM or FMEA,” says Ivara’s Ann Christie. “Upper management will be interested in understanding the business case in terms of speed of payback and multiples of return. In addition, senior management will look at nonfinancial business impacts particularly those that reduce risk. For example, in regulated industries improved reliability will enable a company to more easily meet the regulators’ demands. Other impacts such as improved safety and environmental compliance can easily be attributed to improved reliability.”

Reliability software
The information on the software listed here was provided by the suppliers of the software. Each description begins with a notation of whether the software developer claims the package is designed to facilitate RCM, FMEA, or both.

If you have used or know of software helpful to managing the information associated with RCM or FMEA other than that listed here, we urge you to notify the editors so that it can be considered for the next edition of this directory. MT

Anyone listening carefully to operating machines recognizes a number of characteristics. There is the steady hum of a rotating machine at a tone usually determined by shaft rotational frequency. The tone may be pure, at a single frequency, or contain overtones. It may be modulated, varying in intensity due to an interaction with another component or machine. Cavitating pumps generate an irregular sharp bubbling sound because that is what is happening inside the pump.

Fig. 1. The impact transmitter can be mounted on compressor crosshead.

Although these transient clicks and impacts may be quite clear to a human ear, they typically do not constitute a significant portion of the total energy of the overall signal sensed by a vibration transducer. Thus, while the characteristics can be heard and seen by a human, they will not appear in a conventionally detected level displayed on a vibration monitor. That is the primary limitation of conventional vibration monitoring as a means to identify the condition of a reciprocating machine.

Components in a reciprocating machine can be compared to many people talking in a noisy room. While the crankshaft is going around, rods are going back and forth reversing on every stroke. Tones may be present but they are typically not primary condition characteristics.

Squeaks, ticks, knocks, and other periodic impact type sounds are generated as components move. Clearances click and should not bang as forces reverse direction. There will be whooshing noises from valves opening and closing as well as flow noise through open valves. Examples of anomalies manifested as impact or spike type transients are knocks from loose parts and liquids in gas, clicks as cracks come together and open (often observed visually as bubbles or oil weeps), and hissing flow noise through closed valves and squealing gasket leaks.

This simplified description shows that in terms of vibration and sound, a reciprocating machine is quite different and far more complex than a rotating machine. For this reason, conventional vibration monitoring and analysis techniques that work for rotating machines do not work well on reciprocating machines. There is simply too much din and transient activity in a sound or vibration signal to accurately identify what a single component is attempting to say.

If the vibration system is set to a high sensitivity in order to recognize small changes, false trips are likely. If the sensitivity is set too low, the monitored machine may suffer extreme damage without a trip from the monitoring system.

A different approach
Instead of looking at steady state tones, it was suggested, why not build a monitoring strategy that is sensitive to the impacts generated by reversals unique to reciprocating machines? With loosening bolting and clearances, leaking valves and other common problems in a reciprocating machine identified by impact type events, why not use this information? When ignored, conditions identified by impacts only get worse.

As the thought process evolved, it seemed logical that a reciprocating machine in normal condition should have some level and pattern of impacts. Any deviation, such as bolts or fits loosening and internal leakage, should produce a measurable change in the pattern of amplitude and the number of impacts in a given time period. The theory was tested and it worked well in practice.

Fig. 2 . To meet Div 1 area classification, the impact transmitter is installed inside this explosion-proof housing. The transmitter has a Class I, Div 2 rating.

Recognizing that the situation with an unreliable vibration monitor could not continue, the company initiated a search to find something better that would be capable of preventing another expensive failure. As a result of the search, the company installed impact transmitters and embarked on an extensive test program. The test program had two objectives: identify optimum transducer configuration and compare performance to the vibration monitoring system.

With the impact transmitters in place and connected to analog input channels on their PLC, the system was tested with an impact wrench on crankcase through bolts. Plug wires were removed to induce misfiring. The engine was overloaded to about 115 percent. The impact transmitters did not respond to either condition. With deliberately increased valve lash, the impact transmitter tripped—on intake valves only.

Following the test, the company concluded that the impact system offered significantly improved protection compared to conventional vibration monitoring. The system is still under evaluation; however, the phantom alarms and trips have ended, and in one case a bent pushrod was found following warning from an impact transmitter.

This plant is particularly concerned about broken valve keepers. In the past, keepers have broken, allowing the valve to contact the piston with serious damage occurring in less than 5 sec. The impact transmitters are expected to significantly increase protection against this type of failure. There also is concern about link rod failures, when wear increases bolt stress until they break. The company hopes the impact system will recognize this type of wear, and the increasing bolt stress, prior to failure.

Another gas plant in California experienced a similar result. After a diet of valve and power piston parts killed the turbocharger on a large integral gas compressor, the conclusion was clear. There had to be a better way than conventional vibration monitors to protect these expensive machines. Following an extensive test program, impact transmitters were installed as primary protection on approximately 15 gas compressors ranging up to 6000 hp.

Installing better protection appears to have frightened away the valve and piston failure the system was designed to avoid. However, one major save and a number of minor saves have demonstrated the effectiveness of the impact method of protection. The major save occurred when an impact-initiated warning led to the discovery of a cracked crankshaft. In addition to this major save, the plant has experienced instances where impact transmitters have warned of loose bolting, including a rod nut, in time to take corrective action before any damage occurred.

Due to the high vibration environment in which the impact transmitters are installed, the plant is considering safety wiring connectors to prevent loosening and false signals. It also wants to devise a similar method to prevent loosening of the impact transmitter from its mounting.

Fig. 3. Typical mounting locations for an impact transmitter.

In one case, an impact transmitter indicated a problem that defied identification. Clearances were checked, and there were no external signs of looseness such as bubbles or oil weeping. Everyone jumped to the usual conclusion—it must be the instrumentation. Another impact transmitter produced exactly the same results. Finally, after much searching, two 11/8 studs were found improperly torqued. When the studs were torqued to their proper value, the impact transmitter returned to normal—an indication of the sensitivity of the technology.

At a compressor site, an operator related an incident where a mechanic was tightening bolts with an impact wrench about 5 ft from the impact transmitter. Although the wrench impacts were not very large compared to those a machine in trouble might produce, the impact system activated and initiated a shutdown.

How it works
The impact transmitter is designed for sensitivity to high amplitude, short duration transient spikes that characterize most potential problems found on reciprocating machinery. These spikes typically do not represent a significant portion of the total energy within a steady state vibration signal. They are lost in the traditional signal processing used for monitoring rotating machines.

Within the impact transmitter, special peak detection circuitry captures and counts impact events above a threshold value during a specified length of time. This counting method has proven reliable in practice. Transient conditions where impact events appear and go are differentiated from mechanical flaws where impact events appear and stay.

Since the impact transmitter is focused on identifying conditions such as looseness, cracks, and leakage, it is typically mounted on the crosshead or distance piece perpendicular to rod motion. The transmitter produces a 4-20 mA output equivalent to the number of impact events above an adjustable threshold level within a preset time window.

In this configuration, the transmitter is a cost-effective addition to the control and monitoring systems present on most reciprocating machines. It effectively uses the inherent dynamic characteristics of a reciprocating machine to warn and trip in the event of defects such as loose and cracked parts, valve and gasket leaks, and other problems unique to this type machinery. MT

Information supplied by Metrix Instrument Co., 1711 Townhurst Dr., Houston, TX 77043; telephone (713) 461-2131.