Introduction to Condition Monitoring

From Imrtwiki

Jump to: navigation, search

Return to Condition Monitoring - Predictive Maintenance Techniques


Contents

1 Introduction

Condition Monitoring is regularly checking equipment symptoms that could predict future failure. It is also called Predictive Maintenance.

Figure 1.1 – Condition Monitoring Measurement & Trending

These symptom checks are best if recorded as numbers and trended. Trending is making a Run Chart, as above, so even small changes over time are made visual. This is called quantitative monitoring. Effective monitoring is also widely carried out with qualitative inspections, that is, using the human senses of look, listen, and touch (no numbers). The combination of quantitative monitoring and qualitative inspections together, can be a very powerful tool to understand equipment health. The key question is, “what are the checks & measurements that predict equipment health” and which are important enough to go to the trouble of regular measurement? The answer depends on what type of machine it is and the likely effects and frequency of unexpected failure.

An equipment health problem is called a ‘defect’. The hidden power of condition monitoring is that it helps identify the cause of equipment defects. If causes are acted upon, often the defect can be eliminated entirely. If not the time till failure may be significantly increases or at least the defect can be stopped from occurring in the future.

A personal View - A strong passion has been gaining knowledge of new condition monitoring measurement techniques. By finding a symptom that’s easy and cheap to measure an expensive unexpected equipment failure can be eliminated. This done widely will significantly increase the cost effectiveness of your business. Chapter 1 will describe some of the author’s key learning experiences with simple CM techniques and suggest where they could be applied. The following chapters give the detail required to understand and setup these simple monitoring systems.

The strong suggestion is that these simple condition monitoring (CM) techniques are suitable for use by the average maintenance person and can be integrated into existing equipment inspection systems. Often the complexity of some CM techniques encourages people to make CM a specialist area of maintenance, which is often viewed with some suspicion by front line maintenance. Basic CM can easily be made an integrated part of a maintenance group’s equipment inspection system. A responsible maintenance person or operator should be regularly collecting basic CM data on his/her equipment. Basic condition monitoring includes measurement, checking trends & alarms and simple diagnostics. These can be carried out on-site as a part of normal equipment inspections. Having a range of simple on-site CM techniques available makes this possible. To get big gains out of simple monitoring systems requires more rigour than is sometimes the case for in-house inspection systems. Also to gain the maximum benefits from such a system requires some knowledge of the process of CM (Page 30).

Higher Technology Condition Monitoring vs Simple Monitoring Systems –The above comments are not to put down the role of higher technology condition monitoring techniques. The author is a strong supporter for their use. But a CM system will be more successful if it has a significant involvement by front line maintenance and planning personnel. The author has seen many successful condition-monitoring systems fail and has heard of many more. More often than not, it is due to the loss of a key person involved with the monitoring system. This is even in large organisations that should be able to support changes in personnel who have specialist knowledge. When a monitoring system is based on simple techniques, the training of a new monitoring person can be counted in days or weeks, rather than months or years and diagnostic skills are more strongly based on the person’s existing equipment and trades knowledge.

One way to minimise the problem of handover is to use an external condition monitoring contractor, who hopefully can manage these training and handover issues with people changes. The challenge with this approach is having internal maintenance people who understand and can use the output from a specialist CM group. Another of the problems with this approach is that it can limit the cost effectiveness of CM and reduce the scope of what CM can be applied to. One common pattern is for a monitoring service to be setup and run successfully but with a change of management, the easily highlighted external cost becomes a key mark for a cost reduction purge. Having a simple front line monitoring system in place can reduce the required scope and frequency of an external monitoring service, thus significantly improving its cost effectiveness. Also the key internal monitoring person becomes the ideal person to assist when the external monitoring person comes onto the plant and is much more likely to fully understand the resulting CM report.

One of the key problems with using condition monitoring data is overreacting to a conservative interpretation of the CM data. But on the other hand it is important to be able to immediately recognise an urgent defect when it does occur. A regular user of simple CM techniques is much more likely to be able to make these judgements confidently as he/she has techniques that can verify the defect severity.

One final benefit of simple condition monitoring techniques is the ability to obtain an instant on-site output from the monitoring. This allows on-site diagnostics to occur while the monitoring person is still at the machine. This requires the ability to trend the monitoring parameters as they are being measured. One of the big frustrations when the author was doing front line CM was getting back to the office to look at trends and identifying a possible equipment condition issue. There would then be the urge to immediately go back to the equipment to do the other simple on-site checks to help verify the defect. If measurements, trending and diagnostics can be a one step on-site process, this substantially improves its cost effectiveness. The more cost effective your CM is the more widely it can be applied. There are a number of ways to achieve trending and diagnostics on-site but the simplest way is the use of manual paper trending sheets.

In each of the following sections some of the authors experiences with specific simple useful condition monitoring techniques are documented with suggestions on how they might be used. It has already been suggested that to get the full benefits from CM, a deeper understanding of the CM process is desirable. Understanding of the CM process has evolved over the years and the last section documents some current views.

2 RMS Velocity Vibration

The first significant condition monitoring system that the author setup was in 1981 in what is now the Energy Services Department of Bluescope Steel (BSL) at the Port Kembla Steelworks in Australia. The CM system was based on simple vibration monitoring measurements of RMS Velocity and RMS Acceleration, all recorded onto manual paper trending sheets. It did not use one of the vibration monitoring data collectors that are widely used now because they were not available in 1981. The system monitored about 240 machines, from 30-megawatt turbines to small pumps and had one full time person collecting the data. The author setup the systems and was then involved part time for about seven years in managing it and carrying out diagnostics on defects found. If more powerful vibration diagnostics methods, technical support or base-lining of machines were required, the steelworks Noise and Vibration Group would be called upon. They used large analogue tape recorders and laboratory spectrum analysers to further analyse vibration issues.

This system of front line monitoring, although very simple, was extremely effective. It helped the department manage substantially more of its maintenance using known equipment condition. This was one of the systems along with other maintenance planning tools that allowed substantial reductions in the maintenance department’s workforce during this period. Historically rotating equipment maintenance had been based on preventive equipment strip-downs for inspection. The simple CM system was able to help eliminate this general practice. It also successfully detected major equipment defects and enabled repairs to be carried out in a planned manner. An example of this was detection & planned repair of a gear pump drive defects in large turbo blower used for oxygen making. An in-service failure on this one machine would have cost many times the cost of the total monitoring programme for the seven years the author was involved with it. The monitoring program also gave the ability to make regular checks to decide if a machine had to be shut down due to a serious defect or if it could make it to the next planned maintenance outage.

The recording system for the vibration data was setup using a one-measurement location per page manual paper trending system. The system used three vibration measurement locations per machine shaft, one radial for each bearings and one axial location on the bearing that took axial thrust. This means that there was a RMS velocity and RMS acceleration vibration per sheet and for a typical pump or fan, six sheets were needed. The trend graphs were on a log vertical scale so that there were no issues of setting up scales. The log scale graphing worked well. To ensure their longevity these trending sheets were placed in plastic sleeves.

The system involved a monitoring scheduling system that used the running hours collected on each machine by the operators to schedule the measurements at a frequency of 4 weeks operation. The scheduling system printed the due inspection sheets each week, which were used to take on-site to collect the data. The measurements were then manually transferred to the trending sheets and the trends graphed (Figure 1.6).

This system worked well but there were a number of issues that could have been improved. The first was the one measurement location per page made it more difficult to compare measurement patterns across a machine, which can be an important part of diagnostics. The current suggestion is to record three separate graphs per A4 page and with 2 pages facing each other in book style, which will allow the vibration data for a full simple machine to be viewed in one go.

Another major issue is the inefficiency of recording data on an inspection sheet, then going back to the office to copy the measurements to a table and then graph the result.. It took just as much time to do the office work as it took to do the on-site measurements and as indicated, changes in measurements levels were not discovered until back in the office. One solution to this problem is to use a more robust paper material and to do the trending of measurements while on-site so that immediate diagnostic checks can be made. The paper used has to be robust enough to survive being taken to site around 30 times and still be usable. One paper technology that will be suitable for these applications is synthetic paper and usually requires the use of special pens. Inspection systems generally tend to have a problem of getting continuity with previous inspection information, especially if different people are involved. Reusable inspection sheets give the ability to overcome this issue and should be considered when setting up any inspection sheet system.

There are electronic solutions to many of these data collection issues, such as using a specialised CM data collector. Another is using standard palm type PDA’s with standard spreadsheet or other programs to record inspections, monitoring data and trending the results. These systems replace paper inspection sheets and have the advantage of computer backup the information. These electronic options often do not have the flexibility or ease of use of a paper system. Also paper system can be backed up using a photocopier, which probably should be done every 3 to 6 months. As well as the above approaches, more standard vibration data collector systems have been used around the world for many years and should be considered as well. RMS Velocity vibration is the simple parameter most useful to understand current condition for higher speed rotating equipment (>600rpm). There is an ISO standard which allows a ballpark estimate of equipment condition (Figure 1.7). The velocity values in this table are an excellent guide for commissioning new equipment when there is no previous history. The acceleration values are not as useful. After this the best indication of condition is given by trending. Stable values indicate no significant increase in clearances, looseness or forces. 

Figure 1.7 - ISO Velocity Vibration Standard

Below is a diagnostic method for an increase in RMS Velocity vibration.

  • Check the pattern of increase across the machine and find the location with greatest increase
    • Axial increase may indicate alignment or other axial forces
    • An increase each side of a coupling may indication misalignment or a coupling defect (same for V belts)
    • Radial locations next to an impellor may indicate balance
  • Check if there is any evidence that machine elements have been changed or adjustments made
    • May be alignment or mounting issues
  • Carry out standard external looseness checks
  • Check around problem area for higher RMS velocity vibrations
    • Ensure there is not an external source for the vibration
  • Use chosen listening technique for bearings with an increase, comparing to other bearings and other similar machines
  • Measure the bearing and motor temperatures and compare to historic levels if baseline temperatures are available
  • Has this or similar machines shown this symptom pattern before?
  • Set severity for the defect (eg. Ok, Moderate, Warning & Severe)
  • Determine if the next stage in diagnostics is required and what the next step will be
    • Quick off-line alignment check
    • Quick off-line internal looseness, runout & bent shaft check (check with dial indicator)
    • Motor (or other) soft foot check
    • Used grease sample sparkle check
    • Motor rundown check
    • Fan impellor clean
    • Fan impellor balance (usually contractor)
    • Get in a vibration specialist to carry out an analysis
  • If possible, determine the cause of the defect and what actions could eliminate the cause in the future
  • Can the deterioration rate be reduced by some simple actions, eg. more lubricant or force reduction (eg reduce V belt tension)?
  • Assess if a repair is likely to be required within the next 6 to 12 months from the detected defect (If yes do an initial best guess of time to failure [prognosis] and initiate any early planning if required [spares & procedures available?])
  • If defect is serious, inform others that require the information

3 High Frequency Vibration And Lubrication

Figue 1.8 High Frequency Vibration Analysis

In the Steelworks Power Station monitoring system RMS acceleration vibration was uses as the parameter to focus on higher frequency vibration defects. This parameter was successful at detecting gearing and early bearing defects. After leaving the power station, one of the projects the author was involved with was to implement vibration data collector monitoring broadly across a large number of steelworks departments. The monitoring systems setup for equipment included both simple parameters as well as the more complex techniques such as vibration spectrum analysis. The simple vibration parameters chosen for monitoring were RMS velocity, RMS acceleration and Spike Energy. Spike Energy is a proprietary parameter on the IRD equipment that measures very high frequency vibration and most vibration equipment suppliers have equivalent parameters. One of the key lessons learnt from this project was the power of trending simple very high frequency vibration parameters to detect lubrication and early bearing defects. Significantly more than half of the defects located in the program were from this parameter and most highlighted poor lubrication issues. Indication of lubrication and early bearing defects give powerful information for use in eliminating the root cause of many equipment failures.

This particular vibration data collector implementation project was focused on giving vibration monitoring and analysis tools to front line maintenance planning teams. Many of the maintenance groups who took up the use of the instrumentation and software had excellent success with their monitoring systems. One of the characteristics of these maintenance groups use of these systems was that virtually none used the more sophisticated vibration information available for diagnostics. They only tended to use the three simple vibration parameters, even though they were given some training in the more powerful techniques. When they needed a more detailed analysis of a defect they went to the specialist CM group but most of the time this was not required. This reinforced that condition monitoring could be successfully carried out by front line maintenance people using only simple parameters. Because of the power of the more sophisticated CM techniques, CM specialist groups or contractors should back up this basic CM effort. The role of the CM specialist would be to monitor more complex and critical equipment and give assistance with diagnostics. Everyone can benefit by having an expert available to call upon when they run out of options.

Figure 1.9– Example of High Frequency Bearing Vibration Monitoring

Another interesting characteristic of the maintenance planners who were most successful in getting benefits from the vibration data collector system, was that they tended to do things rigorously. Once they were sold on the possible benefits of monitoring, they ensured the data collection and first level analysis was carried out every month without fail. The people who were collecting the data were usually called equipment inspectors and some were enthusiastic about being involved with monitoring (it tended to reduce the number of late night callouts to failures) and others were not. Still the long-term success of the system was more related to its rigour than by the enthusiasm of a particular person involved at any one time. Since that time the author has experienced use of other types of simple very high frequency vibration instruments. The type in particular that has made an impression is the contact ultrasonic vibration instrument (Example shown in figure 1.12). These use similar electronics to the Spike Energy parameter discussed above but also has a system that allows the very high frequency vibration to be heard by your ears. This listening system is a very powerful qualitative diagnostic tool similar to the listening rods and electronic stethoscopes that will be discussed below. The key difference is the person using it can also hear lubrication related issues as well as the other lower frequency sounds. A key application for this technology is for low speed rolling element bearings.

A method for diagnostics of machine defects using a very high frequency (ultrasonic) vibration parameter is given below.

  • Identify location with a significant increase in the ultrasonic vibration parameter
  • Find the maximum point of ultrasonic vibration on the bearing housing. Confirm that the bearing loaded area is the source. (If not, locate the likely friction/noise location [seal area, impellor rub, flow turbulence, electrical noise])
  • Compare earphone sound to other locations on the machine and to other similar machines
  • If the defect seems to be bearing related apply grease to the bearing (If greasing achieves a significant reduction in the parameter & the level stays reduced from the previous level for more than one day, then the defect was poor lubrication)
  • If there was absolutely no change of sound with greasing, ensure that grease is actually getting to the bearing
  • If the ultrasonic parameter does not significantly reduce or increases again soon after greasing, then it is likely that loading or early bearing defect is the problem
  • Carry out listening rod checks (obvious noise pattern indicates higher severity) and temperature measurements (trending up sometimes indicates imminent failure but there can be a number of other explanations)
  • See other relevant off-line diagnostic checks and follow-on actions in section 2 above

The recommendation for a very simple but effective vibration monitoring program is to use a simple velocity vibration instrument such as a vibration pen (or similar) to measure RMS velocity (many preference an instrument with a separate magnetic mounted transducer) and to use an ultrasonic vibration instrument with headphones to measure a high frequency vibration parameter. The simplest method for recording and trending is to use paper based system. Trending and assessment on-site gives maximum efficiency for monitoring and allows diagnostics to be commenced immediately if a possible defect is detected. To ensure the paperwork can last a number of years it is recommended to use synthetic paper. The same ultrasonic instrument that is used for monitoring very high frequency vibration can also be used for a number of other monitoring and diagnostic applications, some of which are listed below.

  • Gas leak detection such as air, CO2 or nitrogen systems
  • Vacuum system leaks
  • Detection of steam trap leakage
  • Detection of internal leakage of valves in pipework networks (Can confirm no leakage for a pipework safety isolation)
  • Leakage past hydraulic or air system cylinder pistons
  • Detection of electrical arcing, tracking or corona (A good technique to use in parallel to electrical thermal imaging)

4 Listening Techniques & Grease Analysis for Low Speed Bearings

Listening to machines with a screwdriver to the ear or other listening device is an age-old and proven condition monitoring technique. The author has had a number of experiences where the human ear has been able to detect machine defect sound patterns where the most advanced vibration analysis equipment does not. Our ears are designed to detect small sound pattern differences. As it does not take any extra time, it is a huge advantage to be able to listen to machine noises at the same time as making vibration measurements. A very small clicking or scraping sound in the background may not be obvious in any vibration data collected but it may be the first sign of a serious defect. One challenging area for vibration monitoring is on low speed rolling element bearings (<120rpm), as the rotational forces are extremely small compared to higher speed shafts. There is an advanced vibration analysis technique, usually called demodulated acceleration spectrums. This is used by most modern vibration data collectors and has proved to be powerful in detecting slow speed bearing defects. Even with this advanced technique it is not unusual to hear of cases where failures have been missed, though it is always difficult to know if it is a limit of the technology or in its application. A number of years ago the author read a case study that has since resonated with his own experience. It was from a US paper mill where they were using standard vibration analysis on lower speed bearing but missing a significant number of failures. They started using the demodulated acceleration technique and significantly reduced the number of failures they missed. The final improvement in their monitoring system came when a listening technique was added to their monitoring, taking a signal from the accelerometer they were using and amplifying into a set of head phones. This they claimed enabled them to catch all failures.

Another experience related to low speed bearing monitoring occurred when the author was in Japan, visiting one of the NSC steelworks. They showed our group the method they were using to monitor their critical low speed bearings, which they claimed gave adequate warning for all their failures. Their technique was to use simple listening rods. They tended to use a 1-meter long rod for accessible bearings and a 2.5-meter long rod for less accessible bearings. The rod design was a metal tube with a steel spike brazed into one end and a ball bearing sized to fit neatly in the ear, brazed into the other end. After this experience the author made listening rods to their design and had the same positive experience of being able to easily identify defects in large low speed bearings in the steel rolling mill environment.

In 2004 the author had the opportunity to visit Mt Tom Price Mine in WA and met Shane New, a reliability engineer on the site. Mt Tom Price had had a major problem with failures of their many conveyor main pulley bearings. Shane personally initiated a successful monitoring program by personally monitoring all these bearings. His tools were a long and a short listening rod, effectively identical to the Japanese technique. As well as eliminating the in-service bearing failure problems he did bearing post mortems and was able to identify and help eliminate the main bearing failure causes. These were contamination, lubrication and fitting issues.

When using long listening rods, thought has to be given to ensuring no safety risk is taken when using them on rotating equipment, especially if it is within a guarded area. The recommended approach is by using a written ‘Take 5’ type process to ensure all risks are managed. You need to be very sure that the listening rod does not become a spear by coming into contact with a rotating surface.

In 2002 the author started a CM project in Newstan Colliery near Newcastle. They had a long-term problem with failures with the main pulley bearings on the drift conveyor that brought coal to the surface from the underground mine. The pulleys were large dead shaft type and the bearing failures seemed to be by a wear mode. Various CM suppliers had set up vibration monitoring on the pullies but the mine still seemed to get at least one unexpected failure a year. A slow wear defect in a large bearing can be difficult to detect using vibration monitoring, as the roller path can sometimes be worn smooth, producing little of the friction and impact noise that vibration analysis uses for detection. Due to the critical nature of this conveyor and the numerous long-term problems, the use of multiple monitoring techniques was specified. They were temperature, ultrasonic listening, vibration analysis (including demodulated acceleration), used grease analysis and seal clearance monitoring.

The technique that ended up being most successful was the used grease analysis. A number of the oil analysis laboratories can carry out analysis on grease in a very similar way to which they analyse oil. The difficulty in this application was getting a practical way to obtain a used grease sample while the equipment was in-service (ideally a grease sampling system needs to be set-up at design or manufacturing stage). Samples had to be taken while the conveyor was out-of-service for maintenance. The outside bearing sealing area on the dead shaft pulleys were cleaned and new grease was pumped in until some used grease was purged out through the seals. This grease was put into a standard oil-sampling jar. The two main drive pullyes on this conveyor had more conventional plummer block bearings, which could be sampled on-line due to multiple grease tapping points. The two monitoring parameters of interest from the grease analysis report were PQ index and Iron ppm. When there is a PQ index of around 2,000 it indicates there is a lot of large steel particles in the grease. In the time the author was involved, the monitoring work initiated two bearing changes and identified a number of early defect symptoms. One interesting issues with these pulley bearing defects was that there was no evidence that the bearing wear defects were being caused by contamination, as might be expected. The most likely failure cause ended up being the bearing tapered sleeves not being fitted to specification.

Another outcome of this monitoring program was that for all the serious bearing defects detected, the tradesman doing the grease sampling knew there was a problem before sending off the grease samples. The two symptoms he looked at were grease colour and visible steel particles in the grease. If there was serious wear occurring in these large bearings, there must have been enough localised temperature at the roller contacts to affect the grease colour.

Talking to a number of the lubrication technicians working on conveyor equipment, the ones that were most switched on to his/her job all seemed to do visual inspection of the used grease for bearing wear particles. Their standard terminology for this was doing a “Sparkle Test”. The standard way of doing this test is to thinly smear a small amount of the used grease onto a light coloured rag and holding the rag up the sunlight. If there is bearing particles in the grease they will “sparkle”, as the bearing wear particles are bright. It does not matter if the used grease has become somewhat contaminated before testing as none of the other contaminants are likely to be bright.

Another simple technique used to try to detect lower level defects is to put a larger quantity of used grease in a jar with kerosene in it. Then a smallish magnet is placed in the jar sealed and shaken. The magnet will tend to pick up magnetic debris as it helps mix the kero and grease and can be extracted and the debris observed. A portable 30X or 100X miniature microscope is used to do more detailed observations of particles. Again, thought has to be given to ensuring no safety risk is taken in getting grease samples while equipment is operating, especially if it is within a guarded area. This could be done using a simple written Take 5 process to assess risk.

A combination of listening techniques and grease analysis gives a powerful method of handling reliability problems in low speed bearings. Using an ultrasonic listing instrument gives the added advantages of being able to record very high frequency vibration parameter that can be trended to indicate lubrication and early bearing defects. The grease analysis can be a simple 5 senses analysis or for more critical equipment, can use laboratory analysis as well. The other obvious techniques are bearing temperature and vibration measured by simple contact by your hand to detect lower frequency vibration issues. These can be backed up by vibration analysis, including demodulated acceleration, for critical equipment.

5 Magnetic Chip Collectors for Oil Lubricated Systems

Many years ago the author saw a condition monitoring presentation from QANTAS. One of their key CM techniques they use is removable magnetic plugs (also called Magnetic Chip Collectors [MCC’s]) placed in their jet engines bearing oil return lines (Figure 1.17). This design allows the removal of the magnet from the oil line without loss of oil. QANTAS’s diagnostics of the bearing debris on a plug is so advanced they can tell exactly the level of debris collected that can allow an engine to fly back from London to Australia without the bearing failing. the author set up a number of these aircraft type removable magnetic plugs in recirculating oil systems yielding useful results but not enough to score a major win, as the range of equipment most suited to the plugs was limited.

The other equipment type where the use of magnetic chip collectors is well established is large mining mobile equipment. Large earth-moving trucks have magnets placed on the internal surface of the sump plugs of engines and transmissions and are checked on oil changes. The OEM’s supply wall charts to help diagnose specific failure modes. Gearboxes are a special area of interest as they can be complex and time consuming items to monitor with vibration. They have large number of bearings in close proximity and often very noisy gears in the same area. The other issue is some of the bearings and gears are often low speed, giving only small vibration patterns indicating a defect in the generally noisy environment. It is difficult to fully monitor gearboxes with just simple vibration techniques. Oil wear debris analysis is the major condition monitoring opportunity with gearboxes as it works independently from shaft and gear speeds.

One key opportunity is in retrofitting some type of magnetic chip collector to standard splash gearboxes. The author had seen designs to apply aircraft type magnetic plugs to gearboxes but these required on-site drilling into the box casing below the oil level. This did not seem worth the risk or the effort. An opportunity to develop a better system arose while helping setup a condition monitoring systems in the then new Millmerran power station in Queensland. One of the requirements of the job was to setup formal oil sampling systems on their gearboxes and oil recirculating systems. An oil sampling tube system was designed that could be easily be retrofitted into existing tapped holes in a gearbox and also included to a magnet chip collector as part of the design (Figure 1.18 above).

This design attempted to:

  • Eliminate the need for drilling but could still sample oil from an optimum location
  • Enable use of a standard oil-sampling pump
  • Take an oil sample or inspect the magnetic chip collector during normal operation, even if the gearbox was not fully accessible
  • Be made from readily accessible components

One application that the sampling tube was applied to was the gearboxes from the 72 large aircooled condensers at Millmerran station. On start-up of this plant there was a number of reliability problems with the low speed bearings in these gearboxes. In a typical CM system these failure modes would be monitored with PQ Index and Iron ppm with laboratory oil analysis. The lab oil analysis did not give a good indication of these defects. This was likely because of a small recirculating pump and filter in each gearbox that was continually removing bearing wear debris from the oil. It was very quickly found that the magnetic plug was capturing the debris from problem bearings and could easily be diagnosed. As in the jet engine bearing application it gave an excellent indication of when failure would occur. In this specific gearbox application the local maintenance personnel quickly took control of the magnetic chip collector monitoring. They had no problem interpreting the debris on the magnetic plugs.

Typically magnetic chip collectors are monitored and diagnosed by a purely qualitative visual analysis but there are laboratory instruments that can quantify the amount of magnetic debris removed from a MCC magnet. The author has previously purchased one of these laboratory instruments for a monitoring system to enable trending of the quantity of magnetic debris collected. The author has developed a simple on-site instrumentto readily quantify magnetic debris. It uses the principle of measuring the gap at which a magnet will hold a collected sample of magnet debris. A routine monitoring system would require plugs to be removed and debris collected at a set frequency of calendar time or service hours.

Use of MCC’s will not typically replace oil sampling for laboratory analysis but in most situations it could be used to significantly reduce the frequency at which it is done. Laboratory oil analysis would then focus more on oil condition and oil cleanliness than monitoring of steel bearing and gear components, as these will be monitored by the MCC’s. It is strongly recommended to investigate the use of monitoring with magnetic chip collectors if you are responsible for any critical gearboxes or bearing recirculating oil systems. It really is easy, cheap and very effective.

6 Temperature

Temperature monitoring has been mentioned in passing a few times already in this chapter. It is an obvious monitoring parameter for a large range of equipment. There is a wide range of lower cost portable temperature measurement instruments available. The type most used in CM is the non-contact type and can be purchased for a little as $50AUD.

One technology that is proven to be a powerful technique is thermal imaging or thermography. This is often considered an advanced condition monitoring technique but the analysis of thermal images relies more on knowledge of equipment than on any advanced analysis technique. As thermal imaging camera systems are becoming much cheaper, then larger organizations can easily justify buying one of these cheaper devices as long as the commitment to training and setting up routes is made. Lower end thermal imaging cameras can be purchased for as little as $6,000AUD. They are a useful monitoring and diagnostic tool for both electrical and mechanical equipment. The important issues is, if an investment of this type of equipment is made, systems must be put in place to ensure that it is not another expensive bit of gear that is purchased and spends 99% of the time in someone’s cupboard.

7 Thickness Measurement & Non-Destructive Testing

Non-destructive testing is a group of condition monitoring techniques that mostly focus on measuring cracks and other defects in materials. One simple NDT technique that is widely applicable is ultrasonic thickness measurement. It is most widely used for measuring the wall thickness of pipes, liners or other non-porus wearing components. The technique allows measurement of internal wear of equipment without disassembly.

Ultrasonic thickness measurement instruments vary in complexity but at its simplest has a simple digital readout of thickness. The technique is identical to sonar or radar. The instrument has a sensor that generates a very high frequency sound that is transmitted through the material. When the sound hits the opposite side of the object being measured it bounces back and is detected by the sensor. The time the sound takes to bounce back determines the thickness of the item. As well a being used for measuring component wear and corrosion, these instruments are also often used to detect significant cracks in shafts and bolts.

8 Defect Elimination & Root Cause Analysis (RCA)

The largest benefit from use of condition monitoring occurs when it is used to help understand the causes of equipment defects, so they can be eliminated. Unless the root cause of a defect is identified and eliminated then the problem may reoccur again and again. The strong relationship between Defect Elimination (DE) and Condition Monitoring (CM) is because evidence is required to understand what is causing a defect and CM is usually one of the main sources of that evidence. For many types of equipment there are readily available solutions for a wide range of failure root causes. Examples are improved shaft alignment, reducing oil contamination levels and improved bearing fitting techniques. Where solutions are not obvious or readily available, then ‘Root Cause Analysis’ is a process that can be used to help find solutions.

9 The Different Maintenance Approaches

There are four strategies for maintenance of equipment and assets.

Reactive Maintenance (RM) – Also called ‘Breakdown’ or ‘Operate to Failure’ maintenance. In this approach you do not perform any maintenance work on equipment until it fails or is about to fail. This approach is used where the effect or consequence of failure is small.

Predictive Maintenance (PdM)-Also called ‘Condition Based Maintenance’. This includes Condition Monitoring (CM), NDT and inspection activities and tries to detect and fix equipment defects before failure. Detecting failures early minimises the consequences of failures but it does not by itself reduce the number of failures.

Preventive Maintenance (PM) – This is servicing, component changeouts, equipment changeouts or repairs done on a fixed time or usage schedule. Assumes equipment deterioration occurs with time or usage.

Proactive Maintenance (PAM) – Also called ‘Root Cause Analysis’ & ‘Defect Elimination’. This includes any activity that tries to eliminate the reasons for defects and failures. This approach is key to improving equipment reliability.

10 Benefits of Condition Monitoring & Defect Elimination

  • Defect Elimination
  • Reduce Equipment Breakdowns
  • Quality Control Of Repairs and Overhauls
  • Extend Time Between Repairs and Overhauls
  • Improved Maintenance Planning

Defect Elimination - Condition monitoring is a very good technique for focusing people’s attention on the causes of machinery defects. Often a large percentage of machinery defects have a few common causes. People fix the same or similar defects over and over again. If solutions can be found to reduce or eliminate these defects, large reliability improvements can be achieved. Examples of this are improving coupling alignment standards and improving methods for bearing installation. Carrying out Post Mortems on the suspect components during & after repair work is key to understanding causes of equipment defects. Using a Root Cause Analysis (RCA) process is also recommended. Reduce Equipment Breakdowns - This is the gain that will be the easiest to achieve, especially if there are many reliability problems in your plant. Failures that without condition monitoring appear sudden can often be detected months or even years before. For rotating equipment reductions of breakdowns of from 50% to 80% are not unusual.

Quality Control Of Repairs and Overhauls - Anyone who has had any significant dealings with maintenance will have seen maintenance-induced failures. It has been estimated that in some environments they can represent over 50% of the machine breakdowns. An example of a maintenance-induced failure is poor alignment of a motor coupling causing bearing overloading and premature failure. Condition monitoring enables a check to be made on many of these defects during machine testing or start-up. If detected at this early stage the defects usually can be easily fixed or controlled.

Extend Time Between Repairs and Overhauls - Condition Monitoring can often detect defects that can be fixed without repair such as lubrication or looseness defects. A 20% to 40% increase in equipment life is not unusual for rotating equipment. With the introduction of Condition Monitoring many usage based Preventive Maintenance jobs (PM’s) can be eliminated. An excellent examples of the use of this technique comes from the aircraft industry. The DC-8 initially had about 200 Usage Based overhaul tasks (hours run, # of landings or flight time). Compare that to a more modern and larger aircraft like the 747, which uses a reliability history analysis and CM approach and has less than 10 Usage Based overhaul tasks. This was achieved with increased levels of reliability.

Improved Maintenance Planning - Another major benefit of Condition Monitoring is that it assists detailed maintenance planning by providing warning time before a defect becomes serious. Unhurried detailed planning allows:

  • Repair to be scheduled at most convenient time
  • Reduction of the repair or overhaul scope (which components are damaged is usually known)
  • Correct spares & other resources to be assembled prior to repair
  • Detailed repair, quality assurance (QA) and post mortem procedures to be produced prior to maintenance
  • Reduces the requirement to carry large stocks of spare parts in stores. The materials can be ordered Just in Time (JIT)

Summary of Condition Monitoring and Defect Elimination Benefits - A study of 500 companies from around the world who implements significant condition monitoring programs showed the following benefits:

  • Maintenance costs reduced by 50-80%
  • Machine breakdown reduced by 50-80%
  • Spare parts inventory reduced by 20-30%
  • Machine downtime reduced by 50-80%
  • Overtime premiums reduced by 20-50%
  • Machine life increase by 20- 40%
  • Productivity increased by 20-30%
  • Profit increase by 25-60%

11 Condition Monitoring & Defect Elimination Process

The 6 steps shown below describe the process of condition monitoring (CM) and defect elimination (DE). The rest of this section expands on these 6 steps. Figure 1.23 below tries to illustrate the process.

  • Setting condition monitoring strategy
  • Trend
  • Diagnose
  • Plan & schedule repair
  • Repair
  • Trend normally again 
  Figure 1.23 – Illustration of the CM and DE Process

Setting Condition Monitoring Strategy

  • Select appropriate CM technique/s
  • Set up monitoring systems
  • Initiate monitoring

Trend

  • Rigorous, repeatable & repetitive measurements
  • Graphical review for deviations from normal

Diagnose

  • Significant deviation detected
  • Diagnose defect & prognosis of failure
  • Diagnose cause/s & mitigate deterioration

Plan and Schedule Repair

  • Plan repair & post mortem
  • Failure prognosis reviewed regularly
  • Repair and post mortem scheduled

Repair

  • Precision repair & QA carried out
  • Monitoring checks to confirm repair success
  • Failure mode post mortem & RCA if required
  • History documented & CM strategy reviewed

Trend Normally Again

Condition Monitoring and Defect Elimination Process in Greater Level of Detail

11.1 Setting Condition Monitoring Strategy

1. Select appropriate CM technique/s

  • Select appropriate monitoring techniques for your business
  • Determine implementation scope including measurement frequency (Industry standards, expert input and FMECA)
    • Ensure measurement parameters focus on valid failure modes
    • Use strategy matrix if available (Chapter 9)
  • Estimate total cost of implementation
  • Determine payback from monitoring
  • Get approval for an implementation project

2. Set up monitoring systems

  • Purchase required instrumentation
  • Setup measurement points and routes
  • Setup monitoring recording and trending system
  • Setup monitoring system KPI’s (Key Performance Indicators)
    • Trend of MS Excel 100% stacked column of condition classification (is equip. condition getting better or worse)
    • Monitoring & inspection route schedule achievement
  • Train personnel in monitoring & simple diagnostics

3. Initiate monitoring

  • Carry out ‘Baseline Measurements’
  • Carry out initial defect diagnostics

11.2 Trend

1. Rigorous, repeatable & repetitive measurements

  • System operates to schedule measurement route collection
  • Specific person should be given responsibility for routes
  • System should not rely totally on one person
  • Measurement point identification to be maintained
  • Route completion report to be generated (preformatted)
    • Each equipment to have its condition classified (4 levels – Ok, Moderate, Warning & Severe)
    • Completed & required actions documented

2. Graphical review for deviations from normal

  • Review trends at the time of measurement if possible
  • Determine if there is an statistically significant change
  • Determine if there is an adverse trend
  • Determine if there is a alarm breached

11.3 Diagnose

1. Possible defect detected

  • Collect extra simple defect correlation data
  • Match symptoms with known defect patterns

2. Diagnose defect & prognosis of failure

  • Diagnose defect
    • 2nd level of diagnostics if required
  • Diagnose cause/s
    • Basic RCA (use generic causal tree if available)
  • Estimate remaining life

3. If practical, eliminate defect with an immediate fix or take actions that will slow the failure process (mitigate deterioration)

  • Grease, top up oil level, reduce forces, tighten looseness, remove blockage, deduce contamination, fix leak etc
  • Initiate ‘defect elimination’ workorder if required

11.4 Plan and Schedule repair

1. Plan repair & post mortem

  • Determine scope of repair & post mortem
  • Repair materials identified & long lead-time items ordered
  • Repair procedures, post mortem checks and QA checks (Inspection & Test Plan) identified/developed

2. Failure prognosis reviewed regularly

  • Failure prognosis updated & priority reviewed (Maximise life with minimum risk)

3. Repair and post mortem scheduled

  • Appropriate repair opportunity identified
  • Outstanding repair materials ordered
  • Repair and post mortem scheduled and responsible personnel identified

11.5 Repair

1. Precision repair & QA carried out

  • Repair materials packaged
  • Repair personnel review and confirm plan
  • Precision repair carried out
  • QA information documented, ITP (Inspection & Test Plan)

2. Monitoring checks to confirm repair success

  • Start-up checks to confirm repair success
  • Trend and baseline checks carried out

3. Failure mode post mortem & RCA completed if required. Use what is learned to:

  • Improve future diagnostics
  • Improve defect elimination for this and other equipment
  • Improve repair Quality Assurance (ITP)

4. History documented & CM strategy reviewed

11.6 Trend Normally Again

Figure 1.24 - Set Goals to Increase your Predictive Maintenance

Retrieved from "http://www.sirfrt.com.au/wikis/imrt/index.php/Introduction_to_Condition_Monitoring"