KCF Technologies Blog

A Checklist to Consider: Calculating the Benefits of a Predictive Maintenance Program

Founded in 1989 in League City, Texas, STI Vibration Monitoring, Inc. "is the leader in providing vibration monitoring systems and hardware for industrial machinery predictive maintenance and condition monitoring custom products."

As a promotion for predictive maintenance in general--and its services in particular--STI has posted a "Field Application Note" on predictive maintenance.  This concludes with a detailed, helpful, and thought-provoking "PM Program Justification:" an itemized sheet to let you use the current costs of your business to calculate the potential benefits of condition-based maintenance.  If you've wondered about the amazing potential of PdM for your enterprise, be sure to look it over.

Part of STI's article is this bullet-point list of what a quality predictive maintenance program can do for a business that relies on rotating machinery.  It's quite a list, too:
  • Reduce Capital Investment
  • Reduce Machinery Breakdowns
  • Reduce Machinery Depreciaition
  • Increase Machinery Life
  • Increase Maintenance Staff Productivity
  • Reduce Dissatisfied/Lost Customers
  • Reduce Penalties
  • Reduce Unnecessary Machinery Repairs
  • Reduce Rework
  • Reduce Scrap
  • Reduce Warranty Claims
  • Increase Credibility and Reliability
  • Reduce Overtime
  • Increase Safety and Reduce Penalties
  • Reduce Injuries
  • Reduce Power Consumption
  • Reduce Spare Parts Inventory
  • Reduce Defects on New Machinery
  • Reduce "Wrong" Repairs
  • Reduce Insurance Costs
 If there's nothing on that list to interest you, you really must have quite the maintenance department.  Or, perhaps it's the fine workmanship of your perpetual-motion machines.

The Future Looks Brilliant for Asset-Monitoring Technologies

"The leading magazine for pump users worldwide," Pumps & Systems published a special November issue on power generation operations, emphasizing "Instrumentation, Controls & Monitoring."  Among the many worthwhile features in the issue was the first installment in a two-part look at "The Future of Asset Monitoring Technologies."  This forward look at the prospects for forward-looking technology was written by Roberto Piacentini, Preston Johnson, and Theresa Woodlei, three managers at National Instruments, "a producer of automated test equipment and virtual instrumentation software" headquartered in Austin, Texas.

The U.S. Energy Information Energy expects both power generation and demand to increase at 8% to 9% for at least the next seven years.  This growth-orient outlook contrasts starkly with their snapshot of of the current state of America's power generation industry: "...aging plants with equipment at the end of its life--increasing demands for reliability--and an aging workforce reaching retirement in the next few years.  All these factors exponentially increase the the need for effective and automatic knowledge transfer, training and new approaches to the maintenance of power generation assets."  The alternative is too grim to contemplate.

The authors accept asset monitoring enabling predictive maintenance as fundamental to the future growth and prosperity of U.S. power generation.  The Electric Power Research Institute (EPRI), a nonprofit funded by the electric utility industry, has come up with the numbers that back that claim:

"According to the research, a scheduled maintenance strategy is the most expensive to conduct at $24 per horsepower.  A reactive maintenance strategy is the second most costly at $17 per horsepower...[plus] the additional costs of safety being compromised.  A predictive maintenance strategy is the most cost-effective at only $9 per horsepower, and it nearly eliminates the risks of secondary damage from catastrophic failures."

To give that meaning, in 2009 the 1,436 coal-fired power plants in the U.S.A. produced an average of 235,885,794 watts each--roughly 316,200 electrical horsepower apiece.  Thus, predictive maintenance could save a plant as much as $4.75 million over the cost of scheduled maintenance, or $2.53 million more than reactive maintenance, not to mention the incalculable safety and production costs a catastrophic failure might incur.

The article lists other benefits of asset monitoring: lean and logical replacement parts inventory and positioning; greatly increased production reliability; more confident scheduling; and minimization of both production downtime and human health risk.  But, what makes it possible for predictive maintenance to produce these savings, beyond the innate efficiency of its paradigm, is a maturing technology that is arriving on the scene just in time to meet this urgent demand:

 Today, with the use of wireless vibration and power monitoring devices, reliability engineers can overcome historic cost barriers."

"Power generation providers are taking advantage of the cost effectiveness of wireless devices to add low-cost sensors to equipment.  Without the need to connect wires to transfer data, reliability engineers can expand instrumentation beyond critical assets and communicate condition monitoring data for many assets across systems."

The availability of all this data can itself be a problem, the authors admit, as "acquiring, analyzing, and managing this massive amount of data efficiently and timely, and communicating operations knowledge throughout the organization becomes a complex task."  Still, given the needed tools and technology, it's a task that may at least be within reach of America's power producers in the twenty-first century.

Matt's Musings: Wireless Vibration Monitoring at a Metal Rolling Plant

We're as far north as you can go in upstate New York--that's Canada over there across the river.  But what looks like any old, large factory here actually is much more.  They broke ground on this spot to build it in the late 1880s, and in 1902 it became the first plant built to smelt aluminum in the world.  When the Wright Brothers were looking for for a lighter engine to power the world's first flying machine, this plant supplied the aluminum for the engine that did the job.

Five or six generations of Americans have worked here, making what the nation needed, turning raw material to molten metal, then bending it to do their bidding.  One way they do that is with what are called rolling caster stands.  Endless glowing bars of aluminum go in, and are rolled to smaller, more versatile shapes sizes and shapes.  Each castor stand is enclosed by a big roof, but our SmartDiagnostics® wireless sensors work very well inside.  They're very cozy there in the dark, at about 150 degrees.  And, even in that hostile, demanding place, the little sensors send a a steady stream of information on machine performance.  You'll find them all along the production chain, vigilant sentries keeping an eye on things, letting the production team know through it's monitors just what's going on.

Read the rest of this post in Matt's Musings.

SmartDiagnostics® Wireless Vibration Sensors Get Class 1, Division 2 Certification

KCF's SmartDianostics® wireless vibration monitoring sensors have received Class 1, Division 2 certification.  This certification allows KCF to expand its market to customers that have potentially flammable gasses or vapors in the air in an amounts sufficient enough to be explosive or ignitable, such as petroleum refineries.  Visit KCF's website to learn more.

KCF Co-hosting APPA Drive-in Workshop in October

KCF Technologies will be co-hosting an Association of Physical Plant Administrators (APPA) Drive-In Workshop with Penn State University on October 30.

Topics that will be covered at the Drive-In Workshop include: "Improving Fiscal Spending by Integrating a Blend of Predictive Maintenance and Preventive Maintenance Programs," "Wireless Wibration Monitoring Deployment for Campus Wide Assets," and "The Whys and Hows of Knowledge and Skills Retention."

For more information about this workshop see the workshop's flyer, or to sign up for the workshop visit the APPA website.

When the Chips are Down, Critical Asset Management Matters

A petrochemical plant.

Founded in 1945, the International Society of Automation is a leading global non-profit that helps more than 30,000 worldwide members as well as other professionals solve difficult technical problems, while enhancing their leadership and personal career capabilities.  It also publishes the bimonthly InTech, a benefit of ISA membership available digitally, or by mail.  The journal takes a practical approach to explore and discuss industry challenges, new technologies, and fundamentals.

The November/December 2011 issue featured a special section on asset management, including "Utilizing asset data for predictive asset management," written by three maintenance and engineering veterans, Alexandre Augusto da Silva, Geraldo Luiz Bellotti, and Heitor Hiroshi Chaya.  Bellotti and de Silva are senior staff of Braskem SA, the top producer of thermoplastic resins in the western hemisphere.  Three dozen Braskem plants in Brazil, the United States, and Germany produce over 16 million tons of resins and other petrochemicals per year.  Chaya, and electrical engineer with 15 years in industrial automation, runs his own consulting firm, Chaya Automação, in San Paulo, Brazil, where Braskem maintains its corporate headquarters.

"Asset management is enabling Braskem...to move toward predictive maintenance and away from costly preventive and reactive maintenance.  The key is learning what assets need special attention as their performance begins to degrade....Making repairs or replacements at the optimum time limits costs and downtime.  There is also less risk of an unexpected failure shutting down...before a repair can be completed."

"One of the plants [in Brazil] has experienced a rapid return on its investment in asset management while maintaining high safety standards for people and equipment.  In fact, by identifying just one previously unrecognized control valve leak, Braskem saved nearly $300,000 per year..."

The story is simple--given the complexity of Braskem's asset-management system architecture, as is capably summarized in the six-page article--but that six-figure bottom line is a guaranteed attention getter.  Furthermore, the savings came not from heading off a maintenance problem at the end of the valve's service, but from detecting an entirely unrecognized problem that had been introduced at the start of its service life:

"While performing valve signature tests [intended to verify that it can function as required, from fully opened to fully closed], Braskem discovered a problem with one pressure control valve on a propylene storage tank that had to remain 100 percent closed under normal conditions.  The test showed the valve to be partially open.  In fact, it remained three percent open even though the control panel indicated it was closed.

"This small opening was allowing about 20 kilograms of pressurized propylene to escape every hour--a loss valued at $297,500 per year.  It was later determined that the problem was caused by a calibration error at commissioning.  The solution was to perform a valve-travel calibration...The air supply pressure was also increased from 32 psig to 40 psig [pound-force per square inch gauge, a unit of pressure relative to the surrounding atmosphere].  This critical valve now closes fully, and the loss has been eliminated."

Still, it's called asset management for a reason.  Sometimes the optimal choice is not at all obvious, and relies on a professional placing of a calculated wager as to which of multiple possible outcomes will provide the most advantageous for the company:

"Braskem technicians...identified excessive vibration in an axial pump on a loop reactor.  In fact, they were even able to identify a specific bearing as the cause an recommended shutting down the process as soon as possible.  However, the ploant was producing a specific product grade and needed to continue operating for another 15 days to fulfill its commitment to a customer."

"The pump's operation was carefully scrutinized over the next 14 days using continuous on-line monitoring along with frequent visual checks at the pump....Although the bearing ceased functioning before the14-day period was up, the pump kept working."

"Staying in production over that 14-day period met customer commitment and produced sales revenue of about $29 million that might have been lost to a competitor it the process had been interrupted."

When the pump was opened up, only the bearing had to be replaced, which was completed in two days.  If an emergency shutdown had been necessary due to an unexpected failure, five days would have been needed to identify the problem and make emergency repairs, custing the company more than $6 million in lost revenue."

Clearly, the old gambler's mantra about the importance of knowing when to fold and when to hold 'em has a powerful place in today's predictive maintenance toolkit.

Photo by Stopgo (Own work) [Public domain], via Wikimedia Commons.

Wireless Vibration Monitoring Breezing into Wind Industry

The following article, authored by KCF Technologies employee Christopher Shannon, was published on RenewableEnergyWorld.com on August 8, 2013.

A SmartDiagnostics®
wireless vibration sensor
mounted on a wind turbine
blade root.
Wind turbines can, and do, fail. And, when they do fail, they do so in spectacular fashion. A quick search on YouTube for “wind turbine failures” yields numerous videos of exploding wind turbines, wind turbines on fire, or involved in other high-stress calamities. 

However, predictive maintenance with wireless vibration monitoring can offer a solution.  By using wireless vibration monitoring to continuously monitor the health of wind turbines, wind turbine operators can access information that can be used to reduce operational downtime.  By tracking vibration patterns that indicate an urgent need for maintenance the operator can take action to repair the turbine before a catastrophic failure occurs.

Furthermore, by using the vibration monitoring information to indicate when maintenance activities need to be performed the operator can reduce unnecessary maintenance and shutdowns, enabling additional significant savings to be realized.  First, let us look at five specific phenomenon that vibration monitoring may detect to avoid excessive downtime or failure.  Then, we will take a look at the potential value of wireless vibration monitoring on a hypothetical fleet of 100 turbines.  Finally, a real-world use case will be discussed.  Read the rest of the article on RenewableEnergyWorld.com.

Photo courtesy of Trans-Tag.  All rights reserved.

Vibration Monitoring Works on Low-Speed Equipment, Too

State College, Pa., home to both The Pennsylvania State University and KCF Technologies, shares the same time zone with Chile's Universidad de Concepción,even though the two are about 5,400 miles apart from north to south.  Still, it says a lot about 21st-century technologies in the Americas that both these esteemed institutions of higher learning and their neighboring industries share a passionate interest in predictive maintenance technologies.

Founded in 1915 as the Technical Association of the Pulp and Paper Industry, TAPPI is today a registered not-for-profit international organization of about 14,000 members--pulp and paper engineers, scientists, managers, and academics.  Two such academics, Pedro Saavedra and Edgar Estupiñan of the Mechanical Engineering Department at UdeC, coauthored a paper in the May 2002 TAPPI Journal titled "Vibration analysis applied to low-speed machines in the pulp and paper industry."  This peer-reviewed predictive maintenance column "used some real-world historical cases from the pulp and paper industries to illustrate that with an integrated analysis of the vibration spectrum and waveform, and the use of averages and a fine frequency resolution, it is possible to detect defects in bearings of low-speed machines."  These are defined by them as machines operating at speeds from six to 300 cycles per minute.

"Most mills have been using the technique for some time to identify deterioration of vital equipment components...predicting and preventing catastrophic failures.   However, monitoring of low-speed machines is more complicated than general machinery monitoring.  In low-speed machines, the magnitude of the dynamic forces generating the machine vibrations decreases as the rotational speed of the machine decreases.  In addition, low-speed machines are typically massive in size.  Therefore, the resultant vibration on the bearing housing...is often very low and can be hidden by background noise."

The six-page article uses numerous images of vibration spectra and waveforms to show how the analysis was done, and how the detected low-frequency aberrations appeared in practical applications during the study.  "Frequency (or spectral) analysis is the most commonly used method for detecting machines faults such as rotor unbalance, shaft misalignment, mechanical looseness, and bearing damage.  The fundamental idea of frequency analysis is to find the relation between the spectral component frequencies and the frequencies of the dynamic forces producing the faults."

The authors conclude that "...research has shown that it is possible for field engineers and technicians to monitor the condition of low-speed machines by using integrated vibration analysis techniques and by paying strict attention to the selection and use of vibration measurement equipment.  Concerted efforts to improve the signal-to-noise ratio of the measurement are required."

More candid still are the "Insights" Saavedra and Estupiñan added following the article itself, including this: "Mill maintenance staff can use vibration analysis to predict faults, but the most important thing is the correct analysis of the information.  Significant cost savings are possible through enhanced maintenance planning and fault prediction."

SmartDiagnostics® Feature Highlight: Customizing Monitoring Bands

In order to make it easier for you to set the same monitoring bands for multiple machines, SmartDiagnostics®provides the ability to create a template of monitoring bands and give that template a name.  Then later on, for other machines, you can apply that template rather than create more individual bands.

To do this, create your monitoring bands on a monitoring location once by hand as described in the manual in Section 4.5.1 Configuring an Individual Monitoring Band.  When you are ready to save the whole group of monitoring bands as the template, do the following:
  1. Click the Save As Template button on the Monitoring Bands Summary view to show the Save Monitoring Band Template pop-up as seen near by.
  2. Type in a unique name to use for your template in the Name entry field to create a new template or select a name from the Existing Templates List to repalce the monitoring Bands in that template.
  3. Click the Save button to save the template for use on another machine.
 To use an existing template of monitoring bands on your monitoring location, do the following:
  1. Click the Apply Template button on the Monitoring Bands Summary view to show the Monitoring Band Templates pop-up as seen near by.
  2. Select the template you want to use on the monitoring location and press the Apply Template button.
If you want to remove an existing template from the system simply click the Remove Template button on the Monitoring Band Templates pop-up.

The value of the monitoring bands template feature is that it allows you to easily configure the same type of monitoring locations on different machines of the same type.  You simply create the monitoring bands once, save them as a template, and you can apply the same bands to each machine at that monitoring location to allow for consistency and ease of configuration.

In reality, similar machines may still operate somewhat differently.  Therefore, once you have applied a template, you can override it to allow for the uniqueness of each machine.  Remember, too, that operating frequency levels are often listed in the User's Manual for each particular piece of machinery, can be obtained by contacting a dealer's representative or main office, or may be available from professional maintenance organizations.

You can learn much more about KCF Technologies' SmartDiagnostics® technology here.

There's a Major Difference Between Vibration Monitoring and Vibration Analysis

A SmartDiagnostics® wireless vibration sensor on a pump at a wastewater
treatment facility.
Dating back to 1905, Gardner Denver Nash --"your source for industrial vacuum and compressed gas solutions" -- has its offices in Pennsylvania, Brazil, Germany, and China.  The firm provides improved global service and technical support for Nash liquid ring vacuum pumps, compressors, and engineering systems serving the chemical, petroleum, power, paper, mining, environmental, food, and wastewater treatment industries.  It also produces a thought-provoking quarterly newsletter.

One of its newsletters last year featured a brief, unsourced piece on "Vibration Analysis vs. Vibration Monitoring."  It's conclusion was unexceptional: "Implementing the proper maintenance plan will have a positive impact on the longevity of you equipment."  But, it found that the difference between monitoring and analysis can be the difference between merely piling up a lot of numbers and insightful comprehension of actual machine performance.

"Companies heavily depend on maintenance to keep equipment running, but simply monitoring bearing vibrations at set intervals may not be the best way to evaluate equipment.  When monitoring and analyzing bearings the trend data, the equipment may seem to be operating correctly when in fact it is not.  An example of this comes from a 2007[-09] case study at a wastewater facility in Washington, D.C."

"Briefly, the wastewater treatment facility began tracking the reliability of it's pumps in 2007 using vibration analysis.  Working with a simple analyzer, a rotating machinery technician examined any pump with high vibrations.  Due to the quantity of pumps and the multiple plant locations, it was difficult to service the pumps before they had issues.  By the time attention was given to problematic pumps, extensive repair was needed--which increased the cost as well as the downtime."

In 2009, the company shifted...to a monthly monitoring schedule.  This new approach helped to reduce the cost associated with catastrophic failures.  The problem with monthly trending is that you can occasionally have higher than normal vibration peaks and they could go unseen in the overall trend data.  Even though they were having fewer failures, they were still not able to determine the problem pumps before there was an issue.  After discussion with an outside consultant, it was determined that a continuous monitoring system would solve their problem."

Triaxial vibration/temperature sensors on the pumps were connected to data modules that sent measurements by radio to a communications module, which relayed the signals to a tower, which transmitted them to a company network.  The result?  "Plant technicians, maintenance personnel, and other operators could view the data directly on their computers, via the internet, from any location.  The results showed that vibrations spiked from a low level of approximately 0.05 inches per second...to a high level that was close to the 0.5 inches per second alarm level.'

Predictive, preventive, and proactive maintenance all have their places in any water and wastewater treatment system.  But, vibration monitoring--just observing that pumps vibrate from time to time--is by no means the same as vibration analysis--recording patterns of vibration when and as they happen, tracking non-standard patterns and peaks, and intervening in time to schedule needed maintenance and avoid downtime.

Photo by Christopher Shannon/KCF Technologies.  All rights reserved.

Maximize Cooling Tower Energy Efficiency Through Scrupulous Maintenance

A cooling tower fan at an university HVAC plant.
"Bringing Knowledge to People...Promoting Efficient Use of Energy" is the mantra of Energy Manager Training, a website mandated by India's Energy Conservation Act of 2002, which is financed jointly by the government of India and by the Federal Ministry of Economic Cooperation and Development of the government of Germany.  Among its pragmatic offerings is an unattributed three-page technical brief with some helpful "Cooling Tower Tips," presented in .pdf format.

"Reducing energy expenditures for your cooling tower may be as simple as regular maintenance.  This Technical Brief explains how proper maintenance will optimize heat transfer and help your equipment operate more efficiently.  It also identifies strategies for upgrading cooling tower performance."

Whether it's relatively small rooftop unit air conditioning a university, office building, or hotel, or one of the massive hyperboloid structures we associate with nuclear power plants, and large chemical complexes, cooling towers work by venting unwanted, or waste, heat that is carried by the cooling water into the atmosphere.

But maintenance matters, and costs of neglecting it can be steep:

"An improperly maintained cooling tower will produce warmer cooling water, resulting in a condenser temperature 5° to 10° F higher than a properly maintained cooling tower.  This reduces the efficiency of the chiller, wastes energy, and increases cost.  The chiller will consume 2.5 percent to 3.5 percent more energy for each degree increase in condenser temperature.  For example, if your chiller uses $20,000 of electricity each year, it will cost you an additional $500 to $700 per year for every degree increase increase in condenser temperature.  Thus, for a 5° to 10° F increase, you can expect to pay $2,500 to $7,000 a year in additional electricity costs.  In addition, a poorly maintained cooling tower will have a shorter operating life, is more likely to need costly repairs, and is less reliable."

Optimizing the performance of any cooling tower means keeping in check the process and problems that can plague them.  "The performance of a cooling tower degrades when the efficiency of the heat transfer process declines.  Some of the common causes of this degradation include:
  • Scale Deposites
  • Clogged Spray Nozzles
  • Poor Air Flow
  • Poor Pump Performance"
Poor air flow and poor pump performance often are readily detectable by excessive or uncharacteristic vibration.

All of these sources of degradation stem from unwanted elements in the water itself.  Thus the keys to maximizing energy efficiency in a cooling tower rely on:
  • Prevention of Corrosion (due to dissolved oxygen and/or acidic pH)
  • Prevention of Scale (due to dissolved minerals that precipitate out in evaporation)
  • Prevention of Fouling (due to dust, dirt, algae, fungi, and bacteria)
The brief recommends controlling and preventing all three through the use of Treatment Dosing Equipment: "In order for the treatment products to to work effectively, they must be properly fed into the cooling system.  Corrosion and scale inhibitors should be maintained at a constant level at all times, whereas biocides are most effective when applied in slug doses on a product alternated basis."

The good news is that even many neglected cooling towers can be returned to efficient operation with the right treatment: "It only takes three to six months to dissolve two tons of solid impurities from the coiler in a cooling tower after the installation of the chemical-free Aqua Correct physical water treatment.  The approximate two tons [of] deposits was the amount...removed from the cooling towers at the worldwide known dairy company MD Foods/ARLA as well as at the Danpo Chicken Slaughterhouse and in hundreds of other cooling tower plants."

Photo by Dr. Jeremy Frank/KCF Technologies.  All rights reserved.

Wireless Sensing + Vibration Energy Harvesting = "A Great Combination"

A SmartDiagnostics® vibration sensor and
harvester.  In his article "Wireless Condition
Monitoring Arrives (and Just in Time),"
ARC Senior Analyst, Harry Forbes says,
"...wireless sensing and vibration harvesting
make a great combination."
Established in 1896, Power Engineering magazine is "the comprehensive voice of the power generation industry that provides readers with the critical information needed to remain efficient and competitive in today's market."  That's why, for three years in a row, it has been named the most read and useful magazine in the power industry.  It is ably supplemented by Power Engineering Online, which provides up-to-the-minute energy news, stock quotes, five years of searchable editorial archives, power generation conference schedules and details, and an industry product and services guide.

Among the worthwhile articles in the online archive is "Wireless Condition Monitoring Arrives (and Just in Time)" by Harry Forbes, a specialist in power generation, transmission, and distribution and a Senior Analyst with the ARC Advisory Group.  His overview begins with a survey of then-new wireless predictive maintenance technology, followed by a short but insightful perspective on what it all means, and the future of equipment condition monitoring (ECM) and prognostic aqpproaches in the power gen industry.

So what does Forbes foresee?  To begin with, still greater challenges ahead:

"Today's utility engineer may be responsible for performance of a 500MW or 750 MW unit, where problems with any of roughly 50 rotating shafts located somewhere on the unit can affect unit availability.  Tomorrow's utility engineer is going to confront an even greater challenge.  For one thing, the low cost baseload generation of future power markets will include far greater amounts of wind power.  In the future, engineers will have to monitor the condition of a far larger number of machines, and the machines will not all be located within a few minutes walking distance from their desk."

Forbes believes "that the cost, flexibility, and other advantages of new wireless sensing technologies will expand the coverage of equipment condition monitoring to a far higher fraction of critical equipment in the power generation industry....This will make condition monitoring a greater challenge for utilities, who will rely on a much larger number of machines to deliver power (and profits) during peak hours."

The crucial question ahead: "What recommendations should a utility follow?"

"First, utilities include automated analytics and diagnosis (not just automated data capture) as factors in their evaluation of new EMC offerings."

"Second, they should use wireless sensing systems to expand both internal and external collaboration in the ECM area."

"Third, greater levels of collaboration also will be essential for condition monitoring in the future, so start building this capability now."

"And, finally, wireless sensing and vibration harvesting make a great combination.  Utilities should expect their suppliers to deliver products that leverage this match."

Photo by Christopher Shannon/KCF Technologies.  All rights reserved. 

SmartDiagnostics® Feature Highlight: New Cloud Capabilities

With the release of version 1.2 in November 2012, KCF Technologies has added capabilities to the SmartDiagnostics® suite.

The objective of SmartDiagnostics® in the Cloud is to make this inexpensive yet powerful monitoring system even easier to install and operate.  To do so, SmartDiagnostics in the Cloud uses a pre-configured collection server to deliver data to the customer from a a centrally located system in the Cloud, thus allowing the customer to implement the system with minimal training and expertise.

SmartDiagnostics® in the Cloud is a customizable platform for monitoring the health and usage of a variety of assets.  The system collects, processes, displays, and reports dynamic sensor information from a variety of wireless sensors to provide near real-time diagnostics.  The system enables management by exception through the use of configurable monitoring bands that interpret sensor readings to determine whether there is an alarm condition that needs to be acted upon.

"Predictive maintenance offers huge cost savings and productivity improvements by enabling actionable data from rotating machinery," according to KCF President Dr. Jeremy Frank.  "However, the sensors must be low-cost and easy to deploy to deliver on the value proposition.  Eliminating the wire cuts out a major deployment cost, and the next step is to make the flow of data automatic and simple."

"The Cloud offers this capability, because data can now flow from the sensor to the collection server, to the Cloud, and down to the user software with minimal local setup of data infrastructure.  The benefit to the user is the easiest possible installation to achieve low-cost predictive maintenance."

Also new to SmartDiagnostics® in version 1.2 is the ability for users to setup e-mail or text alerts to warn them of potential failures when they are away from the computer.

Of the new e-mail alert feature in SmartDiagnostics®, Dr. Frank added, "Modern day plant managers are responsible for successful operations of the plant, and the reality is that predictive maintenance practices are rarely the primary focus.  Often, that means it is difficult to pay attention to the health of rotating machinery, which creates a challenge for taking full advantage of predictive maintenance."

"E-mail alerts help address this challenge by automatically bringing warning or alarm condition to the attention of the right person, increasing the likelihood of catching a developing problem proactively."

You can learn much more about KCF Technologies' SmartDiagnostics® technology here.

All Vibration Monitors Are Not Alike: Troubleshooting vs. Predictive Maintenance

Ken Piety is the Vice President of Technology at Massachusetts-based Azima DLI, which "delivers machine health reliability solutions with global reach that reduce risk, improve safety, increase production, and optimize efficiency" at pulp and paper plants and other facilities that use rotating machinery.  Piety also narrates a brief but brilliant little video titled "Why Vibration Troubleshooting Instruments are Inadequate for Predictive Maintenance" (PdM) on the Reliable Plant website and on YouTube.

"Sometimes the distinctions can be confusing," Piety says, noting that tools involved in both troubleshooting and PdM, "often do the same tasks.  Both instruments often measure vibration, frequently [taking] detailed measurements like vibration frequency spectra...and both instruments have software associated with them, and may do detailed fault analysis.  However, the fact that there are many similarities does not mean the instruments are capable of accomplishing the same purposes."

"With predictive maintenance...you are able to have a current and up-to-date indication of the health of your machines, and to do that it's important that those machines be screened or scanned on a periodic basis, that their state of health be evaluated....It's extremely important in a predictive maintenance program to capture machines that are beginning to fail at a very early stage.  Often we talk about this as incipient failure detection, and we may be able to capture the fact that the machine is going to fail weeks--perhaps even months--ahead of time."  Accurately anticipating such failures before they happen means parts can be ordered and needed maintenance scheduled when it is most convenient, thus maximizing uptime.

"For a predictive maintenance program, it's common that, perhaps on a monthly basis, a technician will go out and screen these machines with a vibration analyzer, collecting large quantities of data, bring those back, scanning the information, and determining out of the hundreds of machines that you collected data on, which of the few machines may have developed a problem.  [The technician's] task then is to analyze that information, and determine what are the specific faults , and recommend the actions...that need to be taken to correct those faults, and what are the priorities of those actions."

"This is where the overlap begins with a troubleshooting instrument," which is, Piety notes, "capable of collecting and analyzing vibration data, and trying to make a determination of what's wrong with the machine.  However...if you haven't been screening on a periodic basis, how do you know when to do that?"

"Typically, the fault [that the troubleshooting instrument detects] is going to have to be in an advanced enough state that it's been detected by people just from their ears, or their eyes, or the smell of something burning, so it's in a very advanced state.  Even though you may be able to analyze what is wrong with the machine, there's less time to be able to take action and correct that in a way that will not cause the plant to be shut down or to have to reduce production."

"Summarizing," he concludes, "I would say that...vibration analyzers that are used for predictive maintenance programs and those that are used for troubleshooting have many similarities, but the goals and what they can accomplish for you are vastly different.  A troubleshooting instrument is used in an isolated situation to solve a single problem, versus monitoring large numbers of machines in order to optimize the maintenance that is performed in your plant."

Predictive Maintenance is Vital in Fixing America's Aging Water Infrastructure

Gregory M. Baird, Managing Director and Chief Financial Officer of AWI Consulting LLC, alsohas served as the CFO of Colorado's third-largest utility, and finance officer of California's 17th-largest city.  He has thus had a sharp eye on the emerging crisis of America's crumbling water and wastewater infrastructure--both from the perspective of revenue-strapped public utilities in a water-challenged West, and as an advocate of the best ways to meet the challenge.

Baird summarized much of his thinking in "The Aging Water Infrastructure," from which his consultancy's acronym was derived. It appeared in the November/December 2010 newsletter of Water Utility Infrastructure Management, available free in the US and Canada.  Those who prefer a graphic-rich presentation online can read Baird's 26-page pdf "The Aging Water Infrastructure (AWI): Needs and Challenges," prepared in October 2010 for the Rocky Mountain Chapter of the North American Society for Trenchless Technology (RMNASTT).

It is hard to overstate the extent of the challenges faced by the systems that carry and cope wit US water and wastewater, vital and related utilities we rely on everyday.  "Community water systems include over 1.8 million miles of network pipes," Baird notes, along with the breathtaking estimate that there are 21 feet of sewer pipe for each of 314 million Americans.

More disturbingly, much of the vast network of buried pipes and pumps that serve our largest cities were installed when most transportation was by horse and electrification was an emerging technology.  Any changes or improvement will be massive undertakings.  Furthermore, the gap between what it will take to repair and update this critical infrastructure and the money actually set aside for that work grows each year.

What Baird champions is an asset-management strategy that is both affordable and sustainable, in which work is allocated to extend the life of water and wastewater assets in the most cost-effective manner possible.  "There is no one-time fix," as he pointedly reminds us.  "This momentous task of addressing the aging infrastructure dilemma requires overcoming many challenges, especially during this extended economic crisis."

One of the keys to doing so, Baird says, must be predictive maintenance.

"As city councils are educated on asset-centric business practices, they begin to comprehend that the water and wastewater utilities are the most capital intensive industries," he observes.  "...in order to attain cost savings, operational efficiencies and lower future risks a return to properly maintaining our assets and extending an asset's useful life in a cost-effective manner is required."

"About 90 percent of US water and wastewater utilities use a geographic information system (GIS).  Every utility is actually on an asset-centric path using GIS for mapping, ...next expanding with additional GIS applications and finally achieving an enterprise-wide operation.  When the investment in GIS is the focus and the whole enterprise is the vision, the full power of GIS tools and functionality can be employed for long-term cost savings."

"An asset registry (geo-database) combined with a [computer maintenance management system] (CMMS) creates a foundation for an enterprise asset management system (EAMS) as promoted by the EPA.  This simple and powerful combination captures asset data, work history, and condition assessments necessary to produce cost-effective, condition-based and predictive maintenance programs."

"This era of sustainability, deliberation, and economic downturn is not for the weak of heart," Baird concludes."  [Water and wastewater] rates will need to increase, and if affordability is truly a core concern then there must be a change from the crisis management approach of waiting for the next sink hole and fixing it to a predictive methodology to avoid even higher rate increases."

Non-Intrusive Electric Load Monitoring Still Promises Tantalizing Diagnostic Potential

Dr. Michael R. Brambley has more than 30 years of academic and research experience related to energy.  He has spent the last 22 years at the U.S. Department of Energy's Pacific Northwest National Laboratory (PNNL) focused on improving energy efficiency in buildings.  At PNNL, Dr. Brambley has served in a wide variety of roles including principal investigator, project and program manager, technical group leader, department chief scientist, and research contributor.  Most of his work over the past 15 years has focused on improving the operating efficiency of buildings and other energy systems, including air conditioning.

In September 2009, PNNL published a short study by Dr. Brambley with a long title: "A Novel, Low-Cost, Reduced-Sensor Approach for Providing Smart Remote Monitoring and Diagnostics for Packaged Air Conditioners and Heat Pumps."  It begins by noting a basic conundrum: "Operation Faults are common in packaged heating, ventilation, and air conditioning (HVAC) equipment...commonly used for space conditioning space conditioning of commercial buildings with less than about 50,000 square feet and many larger buildings with three floors or less....Remote diagnostic monitoring systems have been developed, bu they are expensive and, as a result, have not achieved significant penetration into the market.  Both hardware and installation costs are too high."

Noting that, "Smart monitoring and diagnostic systems (SDMSs) built for field testing...ina a follow-on to the current project had an estimated cost of approximately &1,000 per SDMS unit," plus, "an installation cost of another $200 to $1,000," Brambley conludes, "a much more lower-cost monitoring and diagnostic system is required to serve this market effectively."

His proposed solution?  "Basic non-intrusive electric load monitoring (NIELM) techniques can be used to extract information about the electricity use and efficiency of individual components of the heating, ventilation, and air conditioning unit from measurement of power supplied to an individual HVAC unit.  By using very few sensors, the capital cost and time/cost required for installation will be minimized, creating a monitoring and diagnostic system with a cost an order of magnitude lower than previous systems developed by the research team..."

Brambley details the original idea as developed by George Hart at MIT in the 1980s, and its evolution into various diagnostics systems.  "We hypothesize that much smaller sampling periods of tens of seconds to a couple minutes might be used to distinguish the on-off events of packaged unit compressors and fans to quantify...electric energy consumption."  He contends, "This together with measurements of outdoor-air temperature (and possibly return-air or supply-air temperature) should be sufficient to detect," six different faults on larger packaged air conditioning or heat pump systems.

What could make a NIELM system powerful and affordable is that it takes these limited power samples from the air conditioner or heat pump only when starting.  Key steps needed to convert the concept into workable technology are, "Adaptation of algorithms from previous work and development of some new algorithms for using NIELM to extract on-off times, power draw, energy use, cycling frequency of packaged unit compressors, and fans from the power connection to the unit ans implementation of them in software," followed by, "Development of a very low-cost hardware package with the necessary processing, data storage and communication capabilities for implementing the NIELM and fault detection algorithms."

In conclusion, Brambley opines, "The changes possible from successfully developing and implementing the NIELM-based technology...will help transform how packaged HVAC equipment is operated and maintained, increasing its operating efficiency and decreasing the energy used for space conditioning the 90 percent of commercial buildings and the 55 percent of commercial floor area that these units serve."  The savings could be still greater, he notes, if the technology were built right into the packaged units.

The promise that Dr. Brambley notes in this intriguing study is real enough, but so too are the barriers that have yet to be overcome to make NIELM-based technology a diagnostic game-changer.  Perhaps someday they will fulfill that great promise.

'The Advantages Are Clear:' Wireless Vibration Monitoring Answers a $65,000 Question

You could scarcely ask for a more upbeat title than "Wireless Vibration Monitoring - Improves Reliability and Enhances Safety," written by Travis Culham.  Culham is the Rotating Machinery Engineer at the Barking Power Station in the large, eastern suburb of Dagenham in London, England.  His article was published three months ago in maintenance.co.uk, "a monthly E-zine, featuring a mix of news and editorial articles" produced in the United Kingdom by Conference Communications.

"Taking advantage of the ease [with] which new measurement devices can be introduced to an existing Smart Wireless network," writes Culham, "our Barking Power Station is using a wireless vibration transmitter to monitor rotating equipment remotely and in real time.  The introduction of this device is helping to improve maintenance schedules and prevent unexpected downtime..."

"Monitoring techniques for rotating machinery have improved greatly in recent years," Culham notes, "with advanced vibration monitoring and analysis tools now able to identify even the slightest changes in the condition of an asset--as they are taking place.  Online vibration monitoring can help to predict when a failure will occur and alert maintenance as to the health status of the equipment.  Early warning of impending failures can prevent process shutdowns that lead to lost production."

"Continuous monitoring is making an important contribution....At Barking Power many of the largest and most critical pieces of rotating equipment have vibration monitoring permanently installed: ideally, all rotating equipment should be monitored..."

But just how important can such a high level of proactive predictive maintenance be?  Culham gives the seemingly trivial example of a gas turbine starter motor, housed in a hard-to-access compartment, from which, Manual readings were taken using a handheld collector and then downloaded for analysis."  This enhanced level of maintenance scrutiny was, important as these motors have a history of problems that can lead to total failure requiring replacement of the entire motor."

"Despite the potential problem, we wanted to continue to run the the motor; otherwise this affected our ability to run the related turbine, reducing our maximum output capacity by 200MW.  Shutting the motor down...and completing a total overhaul would make the turbine unavailable for approximately 36 hours.  Potentially this could cost our company as much as £50,000 in lost revenue (about $65,000), depending on the price and demand for power that day."

A wireless vibration monitor proved to be just the ticket.  "This success gave us great confidence," Culham recalls.  "If smart wireless technology could be applied here, then it could be applied pretty much anywhere throughout the plant."

In addition to its accuracy, remote vibration monitoring had other advantages: "Without the wireless vibration transmitter, we would have been unable to monitor the starter motor safely and would have had to take it out of service--with all the negative production impact that would have entailed.  Additionally, removing the need for maintenance personnel to visit the plant floor reduced risk."

"At Barking Power we want to continue to use technology to avoid forced outages...," he concludes.  "The advantages are clear.  We no longer need to have plant personnel make as many trips to the field, so safety improves.  We receive vibration data transmitted from the motor....This enables us to estimate when a motor is going to fail.  The real-time information from the wireless vibration transmitter provides valuable insight that can prevent unplanned shutdowns and improve maintenance scheduling and reliability."

SmartDiagnostics® Feature Highlight: Dynamic Alert Icons

Last month, our SmartDiagnostics® Feature Highlight briefly explained how to set a base line after a sensor node is placed on a properly working machine.  Page 35 of the SmartDiagnostics® Vibration Monitoring System User's Guide introduces the important subject of alarms and warnings (Section 5.4).  One of the principal values of the SmartDiagnostics® Vibration Monitoring System (VMS) is that it facilitates exception-based monitoring of machine condition.

If you are implementing a facility-wide condition-based maintenance program, you may have a large number of machines to monitor on a regular basis.  When you configure and apply monitoring bands with warning and alarm thresholds, you can let the system do much of the daily basic monitoring for you.  When the VMS detects that a machine's vibration level is exceeding one of the thresholds, it automatically tells you of the event so that you can take action.

VMS indicates alert conditions in several ways to allow you to quickly and easily navigate to the machine that is having a problem and to see visually when and why the alert condition was raised.  In addition to the system wide-alert log, VMS presents four different alert indications.  The first of these are Dynamic Alert Icons on the Navigation Tree, discussed in Section

Any time a vibration sample exceeds an alarm or warning threshold in one or more monitoring bands, an alert is generated and presented as an alarm icon, or warning icon on the monitoring band, monitoring location, machine and facility levels in the Navigation Tree.

If there are multiple alerts triggered at the same level in the Navigation Tree, the higher level will propagate up the tree to the next level.  For example, if the warning threshold is breached in a band on one machine and the Alarm threshold is breached in a band on another machine, then the alert level for the facility as a whole will be set to the alarm level and will be indicated with the alarm icon.

The nearby Navigation Tree shows an example where one machine--Chiller One--has triggered and alarm and a second machine--Chiller Two--has triggered a warning, so the alert level of the West Campus HVAC Plant as a whole shows as an alarm.

The alert icons are dynamic and only show the alert level of the most recent vibration sample.  As such, if the machine temporarily exceeds a threshold vibration level and then returns to normal operation, the alert icon will show up briefly and then go away.  In this way, the alert icons in the Navigation Tree always give you a snapshot of the current state of your machines.

You can learn much more about KCF Technologies' SmartDiagnostics® technology here.

Wireless Sensors Work

In the world of vibration monitoring, wireless sensors can provide a cost-effective alternative to traditional machine monitoring methods.

The following article, authored by KCF Technologies employees Christopher Shannon and Matt Cowen, was published in the June/July 2013 issue of Uptime Magazine.

SmartDiagnostics® sensor
on a compressor.
Traditionally, machine vibration monitoring is performed in two ways: machines can be periodically monitored by utilizing a temporarily mounted sensor and a portable analyzer machine, or machines can be continuously monitored by permanently mounting sensors and wiring them into a high end diagnostic system in the plant.

The advantage of a portable system is that it can cost less to procure and install since there is no permanent wiring required. However, if a facility decides to hire an outside firm, even this option can be costly, running between $600 and $1,200 per day while still providing some level of predictive monitoring. The disadvantage of a portable system is that machine problems do not follow a schedule and there is a very real possibility that a machine can develop problems or even fail between the periodic assessments.

Permanently mounted sensor systems attempt to address the issues presented by portable systems, but they do so at a very high cost.  Acquiring and installing a permanent system can run into the hundreds of thousands of dollars when you factor in the costs of the sensors, diagnostic machine and software, and the installation and maintenance of long wire runs that are necessary to power the sensors and collect the vibration data. These costs can dramatically affect the return on investment (ROI) of continuous machine vibration monitoring for predictive maintenance and put such systems beyond the financial reach of most companies.

While permanent machine monitoring has traditionally been performed using wired sensors, costs for wiring vibration sensors are high, ranging from $50 to $100 per foot. Wire installation costs are a driving factor that limits the affordability of vibration monitoring. Wireless sensors address this cost issue.  Additionally, wireless sensors offer to simplify sensor installation, reduce maintenance associated with wiring faults, permit new sensor locations that would not have otherwise been accessible with wired sensors, and offer greater flexibility with easy installation or removal, as required.

In summary, wireless sensors have the promise to make vibration monitoring practical for most companies.  Read the rest of the article on the Uptime Magazine website.

Photo by Matt Cowen/KCF Technologies.  All rights reserved.

Pointing the Way for Practical Vibration Monitoring at Paper Mills Today and Tomorrow

President of J.M. Robichaud Professional Services in New Brunswick, Canada, Mike Robichaud is a Strategic Adviser to Acuren, where he has been General Manager of Reliability Engineering since 2010.  Previously he was founder, president, and principal engineer of Bretech Engineering Ltd. for more than 21 years, certified as both a vibration analysis and as a maintenance and reliability professional.

It was during this long period of professional development that he wrote "Practical On-Line Vibration Monitoring for Papermachines"--a 12-page paper accompanied by an 82-slide PowerPoint presentation.  It was first presented at PaperCon 2009, organized by TAPPI, a non-profit, international organization of 14,000 member engineers, scientists, managers, academics, and others involved in pulp and paper, founded back in 1915 as the Technical Association of the Pulp and Paper Industry.

"Vibration analysis is one of the most powerful condition-based maintenance technologies," Robichaud begins, "the cornerstone of many predictive maintenance programs.  It is also widely utilized for troubleshooting and fault diagnosis for machinery and structures.  In recent years, much emphasis has been given to on-line or permanently installed vibration monitoring for machinery that is inaccessible, critical to process, and/or very expensive.  This article will provide a practical overview of system components, installation considerations, and benefits of on-line monitoring."

After briefly reviewing the increasing recent use of both hardwired monitors for condition-based maintenance in the pulp and paper industry, Robichaud opines on what lies ahead: "The next logical progression is a paradigm shift from focus on maintenance practices to focus on asset management....Clearly, augmenting the process information with relevant equipment condition data, in an easily understood format, can lead to substantial improvements in productivity and profitability."

He then turns to "several obvious system features, which determine the overall success of the system."  He gives examples of user interfaces in which "data...[are] presented in an easily understood format," using a simple "equipment schematic...with 'traffic light' alarm indicators."  These should be supplemented with displays of "discrete frequency alarm bands" and "trends of vibration amplitude" to give maintenance specialists "an indication of fault condition and progression."

A catastrophic paper machine failure.
Other "advanced diagnostic features" that are championed by Robichaud include Fast Fourier Transform (FFT) and time waveform plots, enabling informed, prognostic maintenance decisions in a timely manner.  The worst-case alternativeto informed decision-making can be the sort of "catastrophic failure" seen nearby, which took place in the gear drives of a dryer section at an old newsprint facility in central New Foundland that has since been shuttered--one of several pulp and paper facilities with machinery that was examined in depth in this report.

"The shift towards asset management strategies creates a requirement for a 'central repository'...of all relevant data," says Robichaud.  "Open systems are widely seen as the most significant challenge/opportunity facing industry--and are critical to the wide acceptance and success of on-line vibration monitoring systems."

The author then reviews the technical requirements of "highly configurable on-line vibration monitoring systems" which have "the best opportunity for plant-wide acceptance....Another important property is configurable data acquisition based on time, process condition, and alarm condition.  Signal processing, including band alarms and advanced diagnostics, may be programmed using configurable analysis 'blocks.'"

Robichaud sums up before proceeding to detailed examples of how vibration monitoring worked at two Canadian mills: "The success of any on-line vibration monitoring system depends entirely on the engineering.  Engineering includes all aspects of selection, installation, and configuration of hardware and software....As monitoring systems become more flexible, open, and configurable, the importance of engineering increases."

WWTP Return on Investment: 'Is What We Are Doing Giving the Best Return?'

A blower at a wastewater treatment plant.
For the past eight years, Saul Cisek has been Lead Maintenance Planner for Virginia's Upper Occoquan Service Authority (UOSA).  As wikipedia notes and presents in a table, "UOSA operates under the Virginia Pollutant Discharge Elimination System (VPDES) Permit...issued by the department of Environmental Quality (DEQ)," the limitations of which, "are among the most stringent in the State of Virginia and possibly the United States."

With successful adherence to these high standards as his professional bona fides, Cisek presented "Predictive Maintenance ROI for Waste Water Treatment Facilities" in July 2011 on ReliabilityWeb.com, an outstandingly useful, "specialty publishing company focused on information delivery of articles, videos, audio podcasts, case studies, iPresentation tutorials, web workshops, benchmark data, tips, and how-to information for maintenance and reliability professionals."

"Predictive Maintenance (PdM)," Cisek states, "adds great value and helps to navigate in an inconsistent, illogical, and disorganized world.  PdM after all is simply using scientific tools to help determine asset condition.  The top tool for most machines is vibration analysis, [but] adding other technologies (ultra-sound, oil analysis, thermal, and electrical analysis) can enhance the results."

Cisek notes that consistent record-keeping is crucial in enabling managers to accurately measure return on investment (ROI).  "having data for costs for acquisition, power consumption and upkeep will aid finance departments in valuation and improve making decisions of when to replace versus when to repair."

The key to maintaining infrastructure and successful PdM is not technology so much as the people using it.  "Develop stakeholders, not just technicians." Cisek advises.  "Train them and allow them to create their own work plans.  If they identify imminent equipment failure, plan, schedule, and execute the fix.  This will help them feel empowered and want to dig deeper for more savings."  And, their knowledge and experience make PdM such a powerful maintenance strategy.

"The PdM technician/engineer is essentially an asset actuary.  For them failure modes are second nature.  They know the common failure modes of their equipment....They know the crash points.  They understand when to launch corrective actions."

Using a blower failure as a practical example , Cisek reviews procedures for assessing gains and losses, noting, "The math for this process is basic, but the substantial knowledge of the enterprise is complex."

He concludes, "The foregoing principles are largely self-evident.  The goal of this piece is to put them into a memorable form for action."  And, he reminds maintenance managers that not everything has a value measured in dollars: "The quest of ROI is not simply saving money.  Additional benefits include improved safety and reliability."

Photo by Christopher Shannon/KCF Technologies.  All rights reserved.

What is Your Facility's "IAQ"?

Inside an AHU.
Laura Rygielski Preston is vice president of the Global Healthcare Practice for Ingersoll Rand, including Trane, now celebrating its first century of helping improve lives around the world through innovative heating and air conditioning systems, services, and solutions.

In the May 2006 issue of Health and Facilities Management, she authored, "Clean Air Acts," her take on, "Designing, installing, and maintaining HVAC systems with air quality in mind."  As Preston notes, "a hospital's indoor air quality (IAQ) is a significant consideration when trying to ensure a healthy, safe, and comfortable environment.  Good IAQ is particularly important in sensitive areas of hospitals, such as intensive care units and surgical suits, because of the growing challenge of protecting patients against hospital-acquired infections."

Humidity and temperature are key concerns when assessing IAQ.  Meticulous planning and communication, "must occur during the design and installation stages of hospital HVAC systems as well as in the preparation of a maintenance regimen that will be used after the HVAC installation is finally completed."

After rigorous design and construction practices create a health care facility--or any other large public buildings, such as offices, schools, or libraries--"a predictive maintenance program can help facilities managers mitigate moisture problems, enhance IAQ, and avoid facility disruptions....Proper maintenance can also reduce operating costs by incorporating energy-efficient technologies.  In fact, the Department of Energy's Building Technologies Program suggests that a maintenance program can reduce energy costs by five to 20 percent in existing health care facilities without significant capital investment."

"With predictive maintenance programs, facilities managers and maintenance personnel can optimize HVAC system performance and maximize the facility life cycle."Crucial to making this work for all parties, as Preston enumerates, are design, budgeting, education, communication, and smart staffing.

For best success, architects, engineers, contractors, HVAC solutions providers, facility administrators, and infection control or risk management professionals must commit to not only the equipment and design that create a healthy hospital, but also to the maintenance regimen that will assure its continuation.  "When such collaboration is successfully employed," Preston concludes, "the end result will be an integrated, reliable and efficient HVAC solution that provides optimized facility investments, increased comfort levels for staff and patients, and improved patient outcomes.

Photo by Christopher Shannon/KCF Technologies.  All rights reserved.

Predictive Maintenance Tip: For Predictive Maintenance to Succeed Practitioners Must Know what Fails

Alan Friedman is Senior Technical Advisor for Azima DLI, a worldwide company self-described as "the leader and premier provider of predictive maintenance analytical services and products that align with customers' high standards for reliability, availability, and uptime."  Friedman recently began what he envisions as a series of articles presented by reliabilityweb.com on the subject "Why do Predictive Maintenance Programs Fail?"

"In the coming months," Friedman begins, "I will be writing a number of articles addressing the subject of why PdM programs succeed or fail from managerial, technical, and financial perspectives.  The purpose of the series is to identify failed strategies and show how they can be avoided."

Acknowledging the constraints we all face in a tough US economy, the author challenges us "to think about how we conduct maintenance and determine how to do it more efficiently and intelligently in the future...."  It's well worthwhile to read Friedman's essay in full, but an overview of his answers offer some insights as well.  Reasons for failure are many, and include:

Lack of Vision: In Friedman's view, PdM "should change the culture, philosophy, and work flow of the maintenance department.  It is not just the addition of a new technology or tool, but a different approach of strategy towards maintaining one's assets...."  Using new tools without buying into the underlying strategy and remembering to "benchmark the gains"--a key task in PdM--all too often leads to failure.

Using a Tool Without Understanding Why: Understanding how to use technology, such as a vibration data collector, without recognizing its predictive power and purpose is a serious impediment to developing PdM as a practical maintenance strategy.  As the author puts it, "the use of the technology as an end in itself without an overall vision of why the technology is being employed" does little or nothing to unlock the potential of PdM.

Failure to Justify the Program: could also be called "failure to do the paperwork."  Even if maintenance successfully adopts the PdM culture and workflow, regularly documenting its success (in savings, uptime, and longer minimum time between failures) is absolutely crucial to obtaining ongoing support from management.  "In other cases," Friedman recalls, "the person managing the PdM program left and no one picked up the ball."

Lack of Consistency: "There are many causes for this,ranging from a failure to commit adequate personnel, lack of proper training, loss of skilled personnel, change in program direction/technology, failure to adequately define the program at the start, and, finally, the lack of a consistent model to monitor the efficacy of the program over time."  What they all have in common is that they undermine sustained commitment to the PdM idea, making success of this highly rewarding maintenance strategy unlikely.

Failures in Training and Partnering: Indifferent training--not Friedman's comitted "combination of complimentary technology and managerial expertise," but merely the usual rote tool-and-manual employee hand-outs--also can fail to instill the strategic focus and entrepreneurial attitude without which PdM cannot succeed.  Even the finest technical instruction is not adequate: "It is important to take these courses, pass the exams, and become certified, but this training alone will not necessarily translate to running a successful PdM program."  Knowing how to gather, sort, and assess the data is all for naught unless trained maintenance professionals also know what to do with it.

Lack of Procedures & Methodology: "...A successful monitoring program...depends on consistency and repeatable performance."  As Friedman observes, "...one needs to test the assets in a repeatable fashion, month after month and year after year for many years.  When this is understood, one will see that a successful program depends much more on consistency and program management...then it does on technical prowess."

Lack of Experience & Commitment: When one doesn't have experience, it's hard to have commitment, and both of these can be hard to come by in a world where everybody has heard of PdM, but few have extensively used it.  "Even if one has the best intentions and the highest level of commitment," cautions Friedman, "it may take a long time to train an employee or group of employees to the point where they can implement a good maintenance program....Like many things in today's world, PdM is becoming a highly specialized area of expertise where...it takes a great deal of dedication and time, which unfortunately, may not be compatible with the other 100 duties you are expected to care of as part of your other work."

Make now mistake; there are authentic hurdles to overcome if predictive maintenance is to succeed.  However, as Friedman himself concludes, "...understanding why things fail is the key to understanding how to get them to work!"

Photo illustration by Pablo X [CC-BY-SA-3.0 (http://creativecommons.org/licenses/by-sa/3.0)], via Wikimedia Commons, and Christopher Shannon/KCF Technologies.

SmartDiagnostics® Feature Highlight: Setting a Baseline

Page 34 of the Smart Diagnostics® Vibration Monitoring system User's Guide includes a brief tip about setting a baseline (5.3.1: Setting Baseline Vibration Spectrum for a Monitoring Point).  This quick click of the button saves a waveform as a baseline, enabling a plant manager to compare data taken from when a machine was known to be "working perfectly" to data showing how it is currently working.  This makes for a quick and easy comparison and will help both with spot checks to confirm consistent operation and with early diagnosis of any potential issues that may develop.

When a sensor node is placed on a properly working machine, it is necessary to take a baseline value.  This baseline will give you a starting point and will help you identify any mechanical issues in the future.
  • After successfully installing a sensor node at a monitoring location, allow data collection to begin.
  • Select an observation that you believe is representative of stable operation.
  • Click the "Save as Baseline" button on the right side of the Vibration Data Chart.
  • Click "Yes" in the confirmation pop-up dialogue box to save the baseline.
 The baseline is now set and is shown by the light grey line in the Vibration Data Chart.  You can prevent the displaying of the baseline spectrum data by clicking the "Hide Baseline" buttom on the right side of the Vibration Data Chart.

Vibration Monitoring is a Key Tool in Condition-Based Maintenance for Hydroelectric Generators

Generators at a hydroelectric plant.
Published in December 2011 by the Hydro-Power Advancement Project (HAP), Best Practice Catalog - Machine Condition Monitoring is an 18-page study prepared by the Chattanooga-based engineering and consulting company Mesa Associates, Inc., and Tennessee's Oak Ridge National Laboratory, representing the U.S. Department of Energy.

The work opens with a brief synopsis of the needs of U.S. hydroelectric installations and potential advantages of vibration monitoring for predictive maintenance:
"Condition monitoring of hydroelectric power generating units is essential to protect against sudden failure.  Fault development can occur very quickly.  Many hydro units are located in remote areas making regular inspection difficult.  It is required to have a monitoring system that continuously checks machine condition, remotely indicates the onset of a fault, and provides the possibility of preventive automatic shutdown."
"Hydroelectric turbine-generators are subject to forces and operating conditions unique to their operation and configuration.  They typically operate at low rotational speeds.  Their physical mass and slow rotational speeds give rise to large vibration amplitudes and low vibration frequencies.  This requires a monitoring system with special low frequency response capabilities."
In the study it was noted that "Vibration analysis was typically performed by a mechanic or the operator by observing a dial indicator.  This is still the only method in older facilities.  Recent developments in vibration sensor, data acquisition, and analysis technologies, however, are making vibration analysis cheaper, easier, and more widely available."

One special innovation of interest is the use of "Models [to] create virtual sensors where physical sensors are not able to be installed.  An example is where real data from physical sensors mounted on the bearings at the shaft ends, is used to create a virtual sensor for mid-span vibration."

Monitor placement, measurement, and output are discussed for hydro turbines and generators, and the importance of integrating their output with the rest of the facility's instruments and controls for ease of reference and integrated use.

The study concludes, "The best way to gain the benefits of a monitoring system is to take advantage of the economic opportunities offered by various modernization, refurbishment, and new projects to introduce the system and to adapt maintenance practices accordingly.  The monitoring system is a major input to a condition-based maintenance program and is a key contributor to capitalizing on high market prices."

"The cost of the monitoring system is low compared with the cost of a new power plant.  A new plant should automatically be equipped with a monitoring system to minimize maintenance outage periods and to help the unit owner stay well-informed of the condition of the equipment."

Photo by Wikisanchez (Own work) [Public domain], via Wikimedia Commons

Many Mills Now Make Most of Predictive and Preventive Maintenance

Founded in 1915 as the Technical Association of the Pulp and Paper Industry, TAPPI is today "the world's largest professional association serving the pulp/paper, nonwovens, converting, and packaging industries."  From 2001 to 2006, TAPPI published Solutions!, a slightly overenthusiastic-sounding monthly journal on all things pulp and paper.  Its August 202 issue included a five-page study "Benchmarking Maintenance Practices at North American Paper Mills," the joint effort of a team of seasoned maintenance professionals drawn from five different states.

The article is an eye-opening look at the state of maintenance at the turn of the 21st century, drawn from an industry-wide survey begun in 1997 of 571 U.S. and Canadian pulp and paper mills.  Of these, 141 (25 percent) took part.  Mills were of all types, and "ranged in tonnages from less than ten tonnes per year to one million tonnes per year, and from less than 100 employees to greater than 1,200 employees."

Annual maintenance budgets at 86 percent of these mills ran from $1 million to more than $25 million.  Much of that cost was labor, accounting for "30 to 40 percent of total maintenance budgets at more than two-thirds of...mills."  In fact, "Respondents reported that the number of maintenance employees equaled an average of about 20 percent of total mill employment," which actually was a drop of two to three percent since the mid-1980s.

"Since earlier paper industry surveys, mills have made substantial progress in the areas of preventive maintenance (PM) and predictive maintenance (PdM).  [Circa 1990], preventive maintenance was becoming was becoming a recognized best-practice, and predictive maintenance was seen by most as 'a nice theory.'  In fact, the earlier surveys did not even mention PdM.  However, with advances in sensor technologies and reduced costs of measuring and diagnostic equipment, PdM has become practical and cost-effective."

Mill budgets for PM and PdM soared accordingly.  Whereas "preventive maintenance had ranged from nine to 22 percent of the total maintenance budget" in the mid-1980s, "a few of the recently surveyed mills indicated budgets for both PM and PdM covered up to 100 percent of the total maintenance budget, with an average of nearly 40 percent."  Earlier surveys from 1993 to 2001 also showed the 85 to 90 percent of mills "rely on Computerized Maintenance Management Systems (CMMS) to track machine repair histories, schedule PM tasks, and provide a means of cost control."

"Most mills ranked quite high in their ability to control unscheduled downtime, with 72 percent...having less than five percent unscheduled downtime and 51 percent of the total reporting less than three percent.

Asked to rank 19 maintenance practices according to which gave the best value, vibration monitoring ranked far above the rest at 97 percent, followed by lube oil and wear analysis (64 percent), walk-down inspections (61 percent), alignment checks (54 percent), and temperature inspections (52 percent).

Even with near-unanimous agreement that vibration monitoring was valuable, actual practices varied wildly.  Ninety mills measured vibration at an average of 1,939 points in the paper machine area, but others monitored as few as ten points and others still up to 11,730 points.  Most mills did the monitoring on a monthly basis (54 mills) or a weekly basis (27 mills), but some monitored daily (17 mills), others annually (9 mills). And, 42 percent of mills reported that they outsourced the gathering of vibration data altogether.

Finally, mills were asked to report the effect of predictive and preventive maintenance on reducing unscheduled downtime to service mechanical and electrical problems.  Some 70 to 81 percent reported that PM and PdM "significantly" reduced downtime, 18 to 28 percent said that PM and PdM activities and programs had only a minimal effect, and only one to two percent felt they had no role whatsoever in reducing unscheduled downtime.

Predictive Maintenance Becomes a Bigger Bargin and a Better Bet

Pumps at a wastewater treatment plant.

Maintenance Strategies: Predictive Maintenance vs. Run to Fail was a five-page paper on water and wastewater treatment published five years ago by Multitrode, "a total solution provider for the municipal water and wastewater pump station industry worldwide."  This firm of some 60 employees, founded in 1986 in Melbourne, Australia, was acquired last month by Xylem Inc., "a leading global water technology company," itself formed in ITT Corporation's spinoff of its water and wastewater business in 2011.

The unattributed paper begins, "One of the most common questions from water and wastewater utilities when they are asked about their maintenance practices in lift stations is 'What do other organizations do?'....Maintenance practices for utilities include run to fail, preventive maintenance on a schedule of time or time in service, and predictive maintenance by monitoring leading indicators of pump and motor problems."

As the title suggests, the study discards at its outset the long-cherished concept of preventive maintenance, using the critique in Thompson and Granger's monograph What Price Preventive Maintenance?, published in 2004:
"For years companies have been performing the preventive maintenance recommended by manufacturers without question...reasoning...that manufacturers have done all the research needed to ensure their equipment will operate properly in any environment.  Most companies also want to ensure that equipment warranties are maintained during the initial install period.  Once these maintenance actions are entered into the CMMS [computerized maintenance management system] or work routine, no one challenges their validity or periodicity because 'that is the way we have always done it.'  This can be a very costly way of doing business."
The Multitrode study found that "across the majority of utilities in the U.S. and Australia the most common methodology was run to fail.  Manpower shortages and/or lack of a budget for better monitoring and control were by far the most common explanation for this choice..."

"The major problem with run to fail is that it is a choice to fly blind as to the state of the assets.  While there are many reasons why pumps and motors can deteriorate faster than historical data shows, there isn't space in this article to detail them all.  However, an excellent example is given in article in Pumps & Systems ["Unbalanced Voltages and Electric Motors," July 2008] by Thomas H. Bishop."

Bishop looked at the effects of slight voltage imbalances--"around 10 percent below nominal"--on the life expectancy of pumps in the United Kingdom "at a large regional municipal utility that introduced a predictive maintenance into the lift stations..."

"[B]ecause the 3-phase supply wasn't remotely monitored, no one was aware of it.  The supply reduction was causing higher running currents, but not enough to trip any panel components or the pump thermistor.  Still, the pumps were running too hot, causing a big reduction in lifetime."  Pumps that should have lasted 25 years were burned out after less than a third of that--only seven to eight years--because of a seemingly trivial voltage shortfall.

Not only can miscalculations like the preceding make "run to fail" extremely costly, but at the same time, predictive maintenance has become much more affordable, practical, and easy to install and operate; that is, cost-effective.  Once regarded skeptically as a cross between sorcery and algebra, predictive maintenance is now making a home where pumps, motors, compressors, fans, and any other rotating machinery is in operation.

Photo by Christopher Shannon/KCF Technologies.  All rights reserved.

KCF Making the Things You Work with Smarter